POPULARITY
Categories
19 episodes with Smaller
7 episodes with Smaller
9 episodes with Smaller
7 episodes with Smaller
8 episodes with Smaller
12 episodes with Smaller
14 episodes with Smaller
18 episodes with Smaller
12 episodes with Smaller
9 episodes with Smaller
5 episodes with Smaller
5 episodes with Smaller
15 episodes with Smaller
12 episodes with Smaller
11 episodes with Smaller
12 episodes with Smaller
5 episodes with Smaller
5 episodes with Smaller
5 episodes with Smaller
4 episodes with Smaller
8 episodes with Smaller
14 episodes with Smaller
5 episodes with Smaller
3 episodes with Smaller
5 episodes with Smaller
9 episodes with Smaller
4 episodes with Smaller
7 episodes with Smaller
4 episodes with Smaller
4 episodes with Smaller
9 episodes with Smaller
4 episodes with Smaller
4 episodes with Smaller
6 episodes with Smaller
4 episodes with Smaller
10 episodes with Smaller
9 episodes with Smaller
4 episodes with Smaller
6 episodes with Smaller
8 episodes with Smaller
7 episodes with Smaller
3 episodes with Smaller
10 episodes with Smaller
2 episodes with Smaller
3 episodes with Smaller
8 episodes with Smaller
4 episodes with Smaller
16 episodes with Smaller
3 episodes with Smaller
3 episodes with Smaller
5 episodes with Smaller
12 episodes with Smaller
5 episodes with Smaller
4 episodes with Smaller
3 episodes with Smaller
2 episodes with Smaller
3 episodes with Smaller
4 episodes with Smaller
3 episodes with Smaller
3 episodes with Smaller
6 episodes with Smaller
3 episodes with Smaller
4 episodes with Smaller
4 episodes with Smaller
3 episodes with Smaller
4 episodes with Smaller
5 episodes with Smaller
4 episodes with Smaller
4 episodes with Smaller
4 episodes with Smaller
3 episodes with Smaller
3 episodes with Smaller
4 episodes with Smaller
3 episodes with Smaller
3 episodes with Smaller
5 episodes with Smaller
3 episodes with Smaller
3 episodes with Smaller
5 episodes with Smaller
5 episodes with Smaller
4 episodes with Smaller
4 episodes with Smaller
6 episodes with Smaller
4 episodes with Smaller
3 episodes with Smaller
5 episodes with Smaller
3 episodes with Smaller
4 episodes with Smaller
3 episodes with Smaller
3 episodes with Smaller
3 episodes with Smaller
3 episodes with Smaller
4 episodes with Smaller
4 episodes with Smaller
3 episodes with Smaller
5 episodes with Smaller
6 episodes with Smaller
5 episodes with Smaller
4 episodes with Smaller
3 episodes with Smaller
7 episodes with Smaller
6 episodes with Smaller
4 episodes with Smaller
3 episodes with Smaller
4 episodes with Smaller
4 episodes with Smaller
2 episodes with Smaller
3 episodes with Smaller
2 episodes with Smaller
3 episodes with Smaller
3 episodes with Smaller
5 episodes with Smaller
5 episodes with Smaller
3 episodes with Smaller
4 episodes with Smaller
2 episodes with Smaller
3 episodes with Smaller
3 episodes with Smaller
3 episodes with Smaller
2 episodes with Smaller
4 episodes with Smaller
2 episodes with Smaller
3 episodes with Smaller
4 episodes with Smaller
3 episodes with Smaller
7 episodes with Smaller
Can you help me make more podcasts? Consider supporting me on Patreon as the service is 100% funded by you: https://EVne.ws/patreon You can read all the latest news on the blog here: https://EVne.ws/blog Subscribe for free and listen to the podcast on audio platforms:➤ Apple: https://EVne.ws/apple➤ YouTube Music: https://EVne.ws/youtubemusic➤ Spotify: https://EVne.ws/spotify➤ TuneIn: https://EVne.ws/tunein➤ iHeart: https://EVne.ws/iheart CATL OPENS CELL STORE TO SMALLER BUYERS https://evne.ws/ady4s BMW SHOWS CHINA-ONLY LONG-WHEELBASE I3 https://evne.ws/b7nab CHINA'S CHARGING NETWORK FOLLOWS NEV SALES https://evne.ws/sb9dn VOLKSWAGEN LAUNCHES ID. ERA 5S IN CHINA https://evne.ws/rkc68 CHINA'S NEV EXPORTS MOVE BEYOND PRICE https://evne.ws/mmc2q NIO FOUNDER PUTS PROFIT FIRST https://evne.ws/h4xbz XIAOMI SETS 2027 FOR NEW DRIVING CHIP https://evne.ws/ezx2y AVATR 06T TESTS UK WATERS https://evne.ws/3ib83 TRUEEV XPENG LAWSUIT DISMISSED https://evne.ws/mr66l
In this short podcast episode, Bryan and Roman Baugh talk about indoor air quality and microbial control, specifically MERV 15 filter physics and how it all works when bringing in outdoor air. We filter indoor air to keep dust and particles off the equipment and keep them from circulating through the air. However, we also need to filter fresh air. While indoor air has several pollutants, including dust and VOCs, that make outdoor air dilution a sensible solution, outdoor air also has spores and other pollutants we want to filter out, and its quality can be hard to evaluate. Roman had an ERV that fed his VRF system, and it had significant microbial growth inside; he had some MERV 6 filtration on it, but that didn't capture particles effectively enough. MERV ratings measure the ability to filter particles out, and we use a unit called microns to measure particle size; higher ratings indicate a greater filter surface area and a better ability to filter out smaller particles. ASHRAE 52.2 divides particles into three categories: E1 (ultra-fine particles between 0.3 and 1 micron in diameter), E2 (fine particles between 1 and 3 microns), and E3 (coarse particles between 3 and 10 microns in diameter). MERV 15 filters have a 90% capture rate of E2 particles and a 90%+ capture rate of E3 particles. Fungal spores tend to fall in the E2 or E3 category, and filtration can keep those from coming into the home with fresh air; bringing in fresh air to dilute indoor pollutants is good, but we need a nuanced approach, and good filtration (without bypass) is part of that equation. Filters can also capture much smaller particles due to the way those particles move. Smaller particles move via Brownian motion, which makes them move erratically and more likely to collide with filter media. However, filter use and design, as well as fresh air intake location, can also affect filter efficacy. Filters that don't fit well in the slot enable bypass, and some filters might be restrictive. Exhaust ventilation outlets also play a role, as wind can bring those pollutants back to fresh air intakes. Have a question that you want us to answer on the podcast? Submit your questions at https://www.speakpipe.com/hvacschool Purchase your tickets or learn more about the 8th Annual HVACR Training Symposium at https://hvacrschool.com/symposium. Subscribe to our podcast on your iPhone or Android. Subscribe to our YouTube channel. Check out our handy calculators here or on the HVAC School Mobile App for Apple and Android.
Website
In This Hour:-- This is the best year in history for the Second Amendment. Noted scholar Stephen Halbrook explains why.-- Smaller pistols may not be as powerful, but many have gravitated to them because they are easy to carry.-- Despite the dire warnings from the gun-ban lobby, the FBI reports that murders are at an all-time low.Gun Talk 08.23.26 Hour 3Become a supporter of this podcast: https://www.spreaker.com/podcast/gun-talk--6185159/support.
Best Buy's incoming CEO says the retailer's path to growth starts with smaller-format stores that can reach markets its traditional big-box locations cannot. Chris Walton and Jenn Hahn debate whether smaller stores can improve productivity and expand Best Buy's footprint or whether the math makes meaningful long-term growth too difficult. ▶️ Watch the full Fast Five episode here: https://youtu.be/VL3Q373gkfY
The episode identifies a structural shift within the IT services market, highlighting a bifurcation between two distinct economic models in the channel: the advisory economy, paid upfront for transformation and integration, and the operational economy, paid on the backend for managed outcomes and recurring support. Techaisle's 2026 Global Channel Partners Survey, referenced by Anurag Agrawal, underscores that most vendors operate single partner programs that implicitly favor one of these models, often without recognizing the divergence. This mechanism exposes gaps in vendor strategies and underscores uneven access to resources and incentives across partner segments. Data from Techaisle's study involving 5,450 partner firms in 24 countries illustrates the impact of these structural choices. Firms under $10 million in revenue project just 8.4% growth, while partners over $500 million forecast 16.8% growth, with 41% of the largest landing in top-tier vendor programs versus only 2% of smaller firms. Anurag Agrawal contends that allocation decisions—such as capital, leads, and support—by vendors drive part of this gap, independently of partner capabilities. The allocation process forms a closed loop, where larger partners consistently receive and convert the best leads, reinforcing their tier status. Furthermore, most vendor incentive spend lands at deal close, benefiting partners focused on new transactions over those delivering ongoing operational value. Supporting developments include evidence that smaller MSPs face higher customer acquisition costs (absorbing 31% of first-year deal value for contracts under $25,000) and operate with little error margin, as opposed to larger firms with more resilient economics. The transcript points out that tier progression within most vendor programs primarily reflects transaction volume and headcount, not actual customer outcomes or quality—making tiers unreliable as indicators of partner value. Additionally, practical AI deployments are now accelerating infrastructure refresh cycles and shifting the center of gravity for services revenue from break-fix to consulting and integration, further complicating the operational landscape for SMB-focused providers. For MSPs and IT service leaders, these findings imply increased dependency on vendor program design and expose operational risk due to imbalanced allocation of leads and support. Smaller providers should expect continued pressure on margins and incentives unless vendors alter their models to recognize operational contributions beyond new logo acquisition. Specialization—vertical or workload-focused—is suggested as a cost-control mechanism, while pricing and packaging transformation work around a recurring services base could mitigate risk. Governance challenges posed by AI adoption, such as managing large numbers of intelligent agents, call for enhanced identity, entitlement, and monitoring capabilities as table stakes for ongoing operational relevance. Supported by: ScalePadProofpoint
Fitter does not have to mean smaller and if your pants are getting snug while you're strength training, that might be the best sign you're actually doing it right. Heather Yancey delivers a blunt midlife reset on what women have been taught to celebrate, and why it's time to stop treating clothing size like a fitness scorecard.Using a client story about building muscle, getting stronger, and protecting bone health, Heather breaks down the real reason your legs or glutes may take up more space when your training is working. This is the conversation women rarely get: the difference between shrinking and becoming capable. You'll discover:Why tighter pants can signal muscle gain, not failureHow progressive overload changes body composition without relying on the scaleWhy glutes, quads, and other lower-body muscles matter for long-term independenceHow resistance training supports bone density, mobility, and longevityWhy muscle is metabolically useful, especially in midlifeHeather also challenges the old fitness script built around “tone,” “lean out,” and “lose the last 10 pounds,” replacing it with a more powerful question: are you smaller, or are you stronger? She reframes the goal from disappearing to building the kind of body that can carry groceries, climb stairs, hike farther, travel, lift, and keep saying yes to life.If you're tired of measuring progress by the waistband and want a smarter, stronger standard for fitness in your 40s, 50s, 60s, and beyond, this one will hit hard. Essential listening for any woman ready to build muscle, protect her bones, and stop apologizing for taking up space.Thanks for listening whether you were folding laundry, going for a walk or whatever other multi-tasking you were getting after. I am having so much fun sharing and connecting with you, badass! Be sure to hit subscribe and get notified of the next impactful episode of The Badass Reset Club which drops every other Thursday.Curious about how AlignSmart can help you boost performance, get out of pain and fix your posture? Book a free call to learn how! Join The Menopause Strength Society Follow me on Instagram Ladies, join our private facebook for menopause support and more! If you want to watch the podcast to see if I actually did something with my hair, find us here: https://www.youtube.com/@heatheryanceyfitnessWanna get STRONG? Grab my free 4 week Strength Training program! In 1 month, you will feel stronger, more confident and badass again! https://www.heatheryanceyfitness.com/opt-in
Bigger weddings can mean bigger revenue—but they can also come with bigger expenses.Studio rent. Coolers. Ladders. Lifts. Rental inventory. Delivery vehicles. Equipment. Staffing.And if you are not careful, all of those expenses can start eating away at the profit you were hoping to make by taking on larger events.In this episode, Jeni shares some of the practical hacks she has learned after more than 20 years in the floral industry for scaling into larger weddings without feeling like you need to buy everything first.She starts by challenging the idea that a successful florist needs a beautiful commercial studio. Jeni has operated as a home-based florist throughout her career and explains why keeping overhead low has allowed her to protect profit while still building a workspace capable of handling large-scale events.She also talks about one investment she does believe can dramatically change the experience of taking on bigger weddings: a cooler. Having proper flower storage can reduce stress, protect delicate product, and give you more flexibility during busy wedding weeks.From there, Jeni dives into renting instead of buying.Need a 12-foot ladder for one installation? Rent it.Need a lift for a massive hanging installation? Rent it—and understand exactly what kind of lift the venue has before assuming it will work.Need Harlow stands or specialty event equipment that you may only use once? Check with event rental companies before adding more inventory to your studio.Jeni also shares why she continues to rent U-Hauls instead of investing in an expensive branded cargo van. For her business, paying for transportation only when she needs it makes more financial sense than adding another major fixed expense.The goal is not to look like the biggest floral company.The goal is to build a profitable business with the tools and resources that make the work easier—without unnecessarily increasing your overhead.In this episode, Jeni talks about:Why bigger weddings can quickly eat away at your profit marginWhy you do not need a commercial studio to have a successful floral businessThe financial reality of adding thousands of dollars in monthly rentHow a home-based studio can support a large wedding businessWhy having a cooler can become a game changer as you scaleFinding more affordable used or commercial coolersWhy you should consider renting equipment before buying itRenting ladders for large installationsWhat to ask venues about their lifts before wedding dayHow the wrong lift turned one installation into a seven-hour projectWhy renting specialty event equipment can make more sense than owning itAsking rental companies about wedding-vendor discountsWhy you do not necessarily need an expensive branded delivery vanHow Jeni uses U-Haul rentals for large wedding weekendsWhy transportation costs need to be built into delivery, setup, and teardown feesHow controlling overhead helps protect profit as your business growsKey takeawayYou do not have to own every resource your business might someday need.Before adding another major expense, ask:Do I actually need to own this—or do I simply need access to it?Scaling does not have to mean piling on overhead.Sometimes the smartest way to grow is to rent, borrow, optimize what you already have, and spend money only where it genuinely makes your business easier and more profitable.
What evidence would convince you that an AI agent is ready to make decisions involving employment, money, healthcare, or legal rights? In this episode of Tech Talks Daily, I speak with Vin Sharma, founder and CEO of Vijil, about the trust gap preventing many enterprise AI agents from progressing beyond proof of concept. Vin has spent approximately 30 years building software across security, operating systems, open source, cloud computing, machine learning, and AI. His previous work includes leading engineering at Amazon SageMaker and helping develop 11 AWS AI services. He argues that AI agents differ from conventional software because they combine autonomy with agency. They can interpret an objective, make decisions under ambiguous conditions, and take action. This raises a deeper question than whether an agent can complete a demonstration successfully: will it remain loyal to the interests of the person or business delegating the task? Trust is also specific to the job. Vin uses a simple analogy. You may trust a gardener to care for your lawn, but that does not automatically make the same person suitable to babysit your child. An AI agent must therefore be evaluated within the context of its users, task, operating conditions, authority, and potential consequences. Vin proposes testing three areas. Reliability asks whether the agent can perform its assigned task. Security examines whether it maintains its integrity when facing hostile or noisy conditions. Safety considers what happens when the agent fails and whether the resulting damage remains contained. This evaluation cannot end when the agent enters production. Models, integrations, data, users, and external conditions change. An agent may drift away from its original purpose, which means businesses need continuous monitoring, testing, and updating across the full AI agent lifecycle. We discuss how established security practices can be applied to this problem. Trusted execution environments, containment, least privilege, limited-duration access, and bounded models can reduce exposure. Smaller language models may also be better suited to narrow, high-risk tasks than a general model with broad permissions. Vin offers a three-part framework for governance: personas, purpose, and policy. Personas describe the people and attackers who may interact with the agent. Purpose defines the legitimate task. Policy sets the boundaries between permitted and prohibited behavior. For high-risk systems, his recommended starting position is that any action not explicitly permitted should be prohibited. A natural-language policy can then be converted into deterministic rules and controls governing the agent's behavior. Vin's most direct advice concerns evidence. Vibes, demonstrations, and benchmark scores do not prove that an agent is safe for a particular business process. A CISO should expect a complete risk assessment, while a business owner should receive proof that the agent will serve the organization's interests. His bridge analogy captures the issue perfectly. Engineers do not claim a bridge is safe because it looks impressive during a demonstration. They calculate load, tolerance, failure conditions, and provide test evidence. AI agents acting in consequential workflows deserve a comparable engineering discipline. If an agent developer asked you to trust their system today, would they be able to provide evidence of reliability, security, safety, loyalty, and contained failure? Listen to the episode and share your thoughts with me.
Seattle police had a smaller footprint at this year’s Bite of Seattle than last year. Bob Ferguson’s office delayed a climate data correction. Pramila Jayapal demanded Republicans denounce Nancy Mace. A viral post humiliated her. Seattle area housing market slows as tech layoffs create uncertainty, according to a new report. // Big Local: Mount Vernon is the latest city to approve a data center moratorium. Snohomish County mayors revolt, demanding sheriff justify 130% jail booking fee hike. The City of Tacoma seeks to deal with pet overpopulation. // Fridays with Jake Skorheim on lazy millennials.
Could your company be paying suppliers earlier than its competitors and unintentionally financing their advantage? In this episode of Tech Talks Daily, I welcome back Oliver Belin, co-founder and CEO of Calculum. Our previous conversation took place around ten years ago when Oliver was working with the Marco Polo Network and blockchain was attracting attention across trade finance. His latest venture concentrates on working capital, payment terms, and the role of AI in supplier negotiations. Oliver explains why working capital has moved higher on the agenda for procurement, treasury, and finance leaders. Companies can generate cash through sales, borrowing, inventory efficiency, faster customer collections, or changes to supplier payment terms. With borrowing costs higher and sales growth difficult in many markets, businesses are examining the cash already tied up within their operations. The difficulty is that procurement teams usually know their own supplier data but lack reliable information about the terms those suppliers accept from other customers. Negotiating without market benchmarks can lead to blunt policies, such as extending every supplier to 90 days. Oliver warns that indiscriminate extensions can create serious consequences. Smaller suppliers may experience cash flow pressure, increase their prices, reduce service, or direct capacity toward customers offering better terms. The buyer may improve its balance sheet while weakening an important part of its supply chain. Calculum uses transactional benchmark data to compare existing payment terms with the wider market. According to Oliver, the platform can show how frequently a supplier appears in its dataset, which terms it accepts elsewhere, and the probability that it will agree to a proposed change. AI and predictive analytics can then help companies concentrate on the suppliers where an adjustment would create the greatest financial impact and carry a higher probability of acceptance. This is particularly useful when an enterprise has tens of thousands of suppliers and procurement teams can only negotiate directly with a small proportion of them. Oliver says Calculum typically identifies free cash flow opportunities equivalent to approximately 8% to 11% of the spend analyzed. The amount identified does not automatically become realized cash. Procurement teams need targets, internal ownership, supplier conversations, and financing options to turn recommendations into results. He shares the example of an unnamed Fortune 500 pharmaceutical company that generated $227 million in free cash flow over 16 months. The program combined market-aligned payment terms with Supply Chain Finance, allowing participating suppliers to receive early payment in exchange for a discount based on the buyer's financial strength. Another UK company with approximately 4,000 suppliers generated €3 million in free cash flow within two months. Oliver attributes the speed partly to knowing which suppliers to approach first rather than attempting a broad, manual campaign. We also discuss supplier protection. Calculum identifies whether a business is a small or medium-sized enterprise, examines ultimate ownership, and considers financial strength. A financially vulnerable supplier may need early payment support rather than longer terms. Oliver's wider point is that AI cannot create reliable benchmarks from nothing. Useful predictions require traceable transactional data, clear objectives, and people prepared to act. Could better payment term intelligence improve your cash position while creating fairer, better-informed supplier relationships? Listen to the episode and share your thoughts with me.
Grab the King Closer Blueprint: My Step by Step Sales Process for closing over 2,000 deals (Only $27): https://www.titaniumu.com/blueprintWant to work directly with me to close more deals? Go Here: https://www.titaniumu.comWant the Closer's Formula sales process I've used to close 2,000+ deals (FREE) Go Here: https://www.kingclosersformula.com/closeIf you're new to my channel my name is RJ Bates III. Myself and my partner Cassi DeHaas are the founders of Titanium Investments.We are nationwide virtual wholesalers and on this channel we share EVERYTHING that we do inside our business. So if you're looking to close more deals - at higher assignments - anywhere in the country… You're in the right place.Who is Titanium Investments and What Have We Accomplished?Over 10 years in the real estate investing businessClosed deals in all 50 statesOwned rentals in 12 statesFlipped houses in 11 statesClosed on over 2,000 properties125 contracts in 50 days (all live on YouTube)Back to back Closers Olympics ChampionTrained thousands of wholesalers to close more deals_________________________________With over 4,000 Videos, this is the #1 channel on YouTube for all things Virtual Wholesaling. SUBSCRIBE NOW! https://www.youtube.com/@RJBatesIII_________________________________RESOURCES FOR YOU:If you want my team and I to walk you through how to build or scale your virtual wholesaling business from A to Z, click here to learn more about Titanium University: https://www.titaniumu.com(FREE) If you want to learn how to close deals just like me, The King Closer, then download the free King Closer Formula PDF: https://www.kingclosersformula.com/closeGrab Titanium Profits: Our exact system we use to comp and underwrite deals in only 4 minutes. (Only $99) https://www.kingclosersformula.com/titaniumprofitsSupport the show
watch this episode for free here //////// join the patreon for more eps herehere is another episode of bst with your pals tom (big) and demi (loud) getting ready to go compleeeeetely off about whatever they have going through their brains! AGH AGH AGH AGHAG HAGH ! (I'm barking at you)we cover miniature horse news and also discuss asses we've seen. unfortunately that's an accurate summation of the episode Hosted on Acast. See acast.com/privacy for more information.
Angeliki Galanopoulou is Founder and Host of "the adaptive times" podcast, meanwhile taking a purposeful career break after more than seven years at Google. Angeliki discusses how AI is transforming work, job security, and career opportunities, particularly for millennials. She shares how burnout, loss, and rapid advances in AI prompted her to reassess work, wellbeing, and career priorities. Angeliki discusses why human connection, empathy, self-awareness, and continuous learning matter to stay relevant and thrive. She explains how millennials can build adaptability and agency to redesign work, explore entrepreneurship, and navigate increasingly uncertain careers. KEY TAKEAWAYS [01:34] Angeliki studies management science and technology while her career direction is unclear. [02:36] Digital marketing fascinates Angeliki combining technology and customers decision-making. [03:54] Understanding how the brain works - motivations, fears, and goals - improve sales. [04:98] Empathy, energy, and memorable relationships can produce stronger business results. [05:56] Burnout and coaching experience reshapes Angeliki's approach to skills, mindset, and work. [06:36] Relentless high performance becomes unsustainable without reassessing personal priorities. [07:05] Questioning what she wants, Angeliki identifies what gives her energy and does not. [07:48] Self-awareness matters to help optimise how work is done and reduce burnout risk. [09:05] Grief creates profound recalibration that cannot truly be prepared for beforehand. [09:35] Previous burnout awareness can help protect energy during difficult periods, including grief. [10:21] Major life events provoke deeper questions about choices, regrets, and careers. [13:11] Traditional long-term job security is disappearing as careers become less predictable. [13:42] Staying relevant now requires individuals to be 'selfish' about their professional growth. [14:35] People staying in stable corporate jobs must keep pace with AI to stay relevant and improve their wellbeing. [15:05] Other people leave their corporate jobs since AI lowers entrepreneurial barriers. [17:04] Time spent AI upskilling is limited inside companies as people don't feel they have time. [18:48] AI upskilling requires deliberately protecting learning time like formal education. [19:46] Individuals must actively practice new tools rather than await organizational guidance. [21:11] AI-prompted internal moves closer to technology for AI influence or to revenue for more perceived safety. [22:00] Communication and influence skills become increasingly important as technology advances. [22:15] Side hustles allow experimentation while maintaining the stability of corporate income. [23:35] Adaptability begins with comfortably admitting uncertainty and remaining genuinely curious. [26:06] Reframing is powerful, e.g. using frustrating situations to improve learning, communication, and personal agency. [28:38] Effective communication starts with deep listening beyond another person's spoken words. [29:31] Self-awareness of strengths and weaknesses reduces defensiveness and improves communication. [30:23] Professional positioning should prioritize audience value while being authentic personally. [32:18] Angeliki is advised to interview externally each year to test her market relevance and polish her positioning. [33:55] Internal mobility helps employees pursue evolving interests while remaining organisationally relevant. [34:54] Organizations should create protected time and resources for continuous AI learning. [35:34] Smaller companies are likely more human-centred than larger companies to keep talent. [37:30] Agency grows by noticing triggers, reframing reactions, and choosing constructive responses. [38:46] IMMEDIATE ACTION TIP: Practice communication, storytelling, positioning, and visible courage to build future readiness. RESOURCES Angeliki Galanopoulou on LinkedIn the adaptive times podcast on YouTube the adaptive times website QUOTES "The safety doesn't really exist, the job safety. Which means that it's on us selfishly and personally to arm ourselves and do what we can to stay relevant and be on top and succeed and thrive in the rest of our careers as millennials." "It's your responsibility to keep up with AI for two reasons. Either to be super up-to-date, relevant and always top of mind, when it comes to doing your job more effectively and, and performing highly. Second, to actually improve your own wellbeing." "The barrier to entry in anything you want to build is minimal. Anyone can build and launch anything with almost minimal cost." "You build it [a side hustle] while you have the stable income. So I like to talk about corporate not necessarily as safe, but as stable." "We should all be adaptable. That's the skill we should all build." "See what triggers in you... And ask yourself, 'Is there a better way that will not ruin my mood for the day that I can look at it?' And then, 'Can I do something about it?'"
Today on the pod we have, Abbey Roberts, the Founder of Fork Diet Culture. Sam and Abbey talk all about their journeys becoming anti-diet dietitians, anti-fat bias, stronger not smaller, and losing weight for health.Resources mentioned in today's episode:Sam's Thoughts on ‘Stronger Not Smaller': https://www.instagram.com/p/DYcM9i1hhjp/ Abbey's Thoughts on ‘Stronger Not Smaller”: https://www.instagram.com/reels/DYiTKHjSfpa/ Abbey's IG Paige | Fork Diet Culture: https://www.instagram.com/fork.diet.culture/ Social Determinants of Health: https://www.goinvo.com/vision/determinants-of-health ALL things Find Food Freedom®:Get your Insurance Benefits Checked: https://bit.ly/FFFinsurance Instagram: @find.food.freedom TikTok: @findfoodfreedom Website:https://find-foodfreedom.com/ Join the FFF Monthly Membership here: https://findfoodfreedommembership.com and use the code 'IWANTFOODFREEDOM' for 3 months completely FREE!
Have you ever felt like you'd like to get sponsorships for your podcast...BUT your audience is too small? This episode is for YOU Pod Pals! I'm tagging in podcast sponsorship guru Justin Moore to teach us how even podcasters with smaller audiences can become sponsor magnets! Connect with Justin here: https://www.creatorwizard.com Links mentioned in this episode: Want more strategies for podcast launch, growth, and how to leverage your podcast to get more clients? Get my FREE weekly Podcasting Business School Newsletter every Thursday! https://www.podcastingbusiness.school/news Join my FREE community on Skool: https://www.skool.com/podcastingbusinessschool/about *********************
Ahead of the upcoming elections, the Taub Center has published a paper pointing to a sharp rise in the threshold for entering the Knesset, driven primarily by the growth in the voting-age population since the 2022 elections. Prof. Alex Weinreb, Head of the Demography Program at the Taub Center, warns of the risk of another major loss of votes similar to 2022, when some 420,000 votes for parties that did not pass the electoral threshold – representing 8.8 percent of the total vote -- were effectively wasted. He told KAN reporter Naomi Segal that leaders of small parties who cannot ensure the support of at least 200,000 voters, should reconsider their course of action. (Photo: File. Yonatan Sindel/Flash90)See omnystudio.com/listener for privacy information.
In this episode of the Cannabis Accounting Podcast, host Raymond Guns sits down with Mitch Osak, Founder and President of Quanta Consulting, to unpack why small cannabis operators are outperforming the giants, and what it actually takes to raise and deploy capital the right way in this industry.Mitch has been in cannabis since 2016, working with over 200 companies across the globe, from Canada's earliest licensed producers to today's US and European operators navigating exports, M&A, and Schedule III.Mitch breaks down:
Providing quality healthcare for athletes isn't just about having more resources. It's about building stronger systems, creating meaningful partnerships, and making sure every athlete receives the care they deserve. HBCU athletic programs often face significant gaps in funding, staffing, facilities, nutrition, rehabilitation, and access to specialty care. These challenges can shape daily medical operations, but they should never lead to lower standards of care. Meeting these needs requires creativity, collaboration, and a commitment to finding solutions. In this episode of Time Out with the Sports Doctor, Dr. Derrick Burgess explores the realities of caring for HBCU athletes and what it takes to build a stronger sports medicine system. He highlights the importance of representation and trust, mentorship and career development, coordinated medical care, and partnerships with health systems, specialists, trainers, donors, and communities. The episode offers an important lesson: closing the resource gap isn't about lowering expectations, but about building the networks and systems necessary to help athletes thrive. "Smaller budgets do not mean you have less talented athletes, and it definitely does not mean you have less serious medical needs." – Dr. Derrick Burgess Topics Covered: 00:00:00 – Introduction 00:01:18 – HBCU healthcare gaps 00:03:06 – The power of mentorship 00:05:06 – The opportunity that started it all 00:08:19 – The orthopedic surgeon shortage 00:09:37 – The role of a team physician 00:11:17 – How funding affects athlete care 00:13:56 – Community impact 00:14:16 – Partnerships that close the gap 00:16:12 – Partnerships that close the gap 00:17:54 – Building the medical pipeline 00:18:49 – Team physicians as leaders 00:19:52 – Representation builds trust Key Takeaways: "We have to be able to take good care of our athletes." "We should be doing more protecting our athletes and taking care of our athletes." "Being in the community is very hard to take care of a team by yourself if you do not have these resources." "Representation can improve trust. If they don't trust you, they will never talk to you." "The best team physicians build networks." Connect with Dr. Derrick Burgess: Website: https://www.drderrickthesportsdr.com/ Instagram: https://www.instagram.com/drderrickthesportsdr/ Facebook: https://www.facebook.com/TimeOut.SportsDr LinkedIn: https://www.linkedin.com/in/derrick-burgess-72047b246/ YouTube: https://www.youtube.com/@dr.derrickburgess243 Email: thesportsdoctr@gmail.com Other Links: https://www.hbcuendzone.org/about Scrubs to Skylines: https://www.instagram.com/scrubstoskylines/ This episode of TimeOut with the SportsDr. is produced by Podcast VAs Philippines - the team that helps podcasters effectively launch and manage their podcasts, so we don't have to. Record, share, and repeat! Podcast VAs PH gives me back my time, so I can focus on the core functions of my business. Need expert help with your podcast? Go to www.podcastvasph.com.
On today's show we take a look at your family room and decide whether a 7.2 system is worth it over 5.1. We also read your emails and catch up on the week's news. News: How Valuable Is TV to Smaller, Independent Cable Operators? DirecTV and EverPass reach new deal to keep NFL Sunday Ticket in bars after threatening split Other: Rage Against the Garage Door Opener (ratgdo) Family Room - 5.1 vs 7.2 what makes sense? Most of us don't have a fully dedicated theater room to fully deck out with a 7.2.4 Atmos system. Most of us have family rooms and would like some sort of a surround sound experience beyond a low end sound bar. The question we get asked a lot is whether a 7.2 system will provide a noticeably better experience than a 5.1 system. So today we answer that question. Spoiler, in most family living rooms, a well-set-up 5.1 system is the smarter choice. A 7.2 system only becomes clearly better if you have good space behind the seating and can place the rear speakers properly. Let's get into it. Pros of Going 7.2 Better immersion - The two extra rear speakers create a more continuous sound field. Sounds that move from front to back (or circle around you) feel smoother and more realistic. Improved envelopment - You get a stronger sense of being "inside" the movie instead of just having sound coming from the sides. Smoother bass (the .2) - Two subwoofers reduce bass peaks and nulls. This is often the biggest real-world improvement over a single sub. More seating flexibility - Multiple people can enjoy better surround effects even if they're not in the exact sweet spot. Future-proofing - Many modern receivers can use the extra channels for Atmos height speakers later if you decide to expand. Cons of Going 7.2 Space requirements - You need room behind the main seats for the rear speakers. If your sofa is against the back wall (very common), the rear speakers end up too close and the benefit largely disappears. Higher cost - Two extra speakers + a second subwoofer + possibly a more expensive receiver adds significant cost. Harder placement - Getting the rear speakers at the right height, distance, and angle is more difficult in a shared family room. Diminishing returns in smaller rooms - In rooms under about 15x15–16x18 feet, the extra speakers can actually make the sound less coherent if they're too close together. More clutter & wiring - Extra speakers and cables are harder to hide in a multi-purpose family room. Content reality - Most streaming content is still mixed in 5.1. True discrete 7.1 tracks are less common outside Blu-rays Rules of Thumb for a Family Room If your sofa is against the back wall you are realistically looking at a 5.1 system since the rear speakers have nowhere to go. If you have 3 feet or more behind you, 7.2 may be worthwhile. If you can't fully afford a proper 7.2 system you will be better served with better speakers and a great subwoofer. Bottom Line For a typical family room: A high-quality 5.1 almost always delivers more enjoyment for the money and effort. Only go 7.2 if you have decent space behind the seating and are willing to spend more for the extra immersion and smoother bass. Next week should we discuss if Atmos is worth it in a family room?
What if the smartest way to grow in a market full of closures is to keep every restaurant small?Adam Weisblatt co-founded Last Word Hospitality and turned Found Oyster, a 777-square-foot oyster bar, into a seven-restaurant group that keeps expanding while famous names around it close.In this conversation, we get into why a tiny footprint takes the pressure off your rent and your margins, why he builds every concept around talented people instead of clever ideas, and why making your operators owners is the only way to grow without burning out.If you're convinced you need a bigger room to make real money, this one will change how you think about scale.That's Adam Weisblatt. To pre-order his upcoming book, For the Love of Restaurants, visit https://a.co/d/0cGZmQ5t._________________________________________________________Free 5-Day Restaurant Marketing Masterclass – This is a live training where you'll learn the exact campaigns Josh has built and tested in real restaurants to attract new guests, increase visit frequency, and generate sales on demand. Save your spot at restaurantbusinessschool.com
Did I make the title a joke about a game from 2005? Yes. Will anyone get it/see it/care? No. But you gotta do what brings you joy. Sometimes that's about making a little joke for you. Sometimes it's about rippin' lines in a grungy club bathroom while kids in bad fashion call you chopped. You do you. I'm not here to judge. I am here to tell you that the industry continues its march through rough times. In this week's episode we talk about EA going private. We wonder if that will lead to a bigger trend of publishers taking themselves private, or perhaps new publishers staying private. Matt seems to think so. I'm not so sure. The temptations and benefits of going public are too great for the corporate stooge. What does it mean for gamers if there are more private publishers? Smaller games. Less triple A hyper realism. Maybe cheaper games. Definitely more digital releases and less physical releases moving forward (and that is a good thing for smaller companies). A future populated with private companies also means more competition, and thus more effort to grab our attention and dollars. Hell, we could even see a return of E3 or similar events. Though likely smaller in scale.
Zander Krause explains why Jalen Hurts must elevate a less proven Eagles offense during a demanding transition.Privacy & Opt-Out: https://redcircle.com/privacy
Progressive candidate Abdul El-Sayed narrowly beat the more moderate Congresswoman Haley Stevens (D-MI) in Michigan's Democratic Senate primary, despite several outside groups reportedly spending over $60 million to support Stevens. Political strategist Lucy Caldwell joined "Forbes Newsroom" to discuss the results, the progressive movement, and the upcoming midterms. Learn more about your ad choices. Visit megaphone.fm/adchoices
The world just got SMALLER!The Wright Brother - Wilbur & Orville Wright, built and flew the first successful, engine-powered airplane. Location: Kill Devil Hills, Kitty Hawk, North Carolina, December 17, 1903 To help support this Podcast & get exclusive videos every week sign up to Neil Oliver on Patreon.comhttps://www.patreon.com/neiloliver To Donate,go to Neil's Website:https://www.neiloliver.com Gold Bullion Partners,for more info about buying gold & silver go to this affiliate link,https://goldbullionpartners.co.uk/download-our-complimentary-guide-neil-oliver/ Shop:https://neil-oliver.creator-spring.com Neil Oliver YouTube Channel:https://www.youtube.com/@Neil-Oliver Rumble site – Neil Oliver Official:https://rumble.com/c/c-6293844 Instagram - NeilOliverLoveLetter:https://www.instagram.com/neiloliverloveletter Podcasts:Neil Oliver: News Comment HistoryNeil Oliver: HistoryNeil Oliver: InterviewsAvailable on all the usual providershttps://podcasts.apple.com/gb/podcast/neil-oliver-news-comment-history/id1513737418https://podcasts.apple.com/gb/podcast/neil-oliver-history/id1871225730https://podcasts.apple.com/gb/podcast/neil-oliver-interviews/id1869660872 #NeilOliver #1903 #flight #firstflight #TheWrightBrothers #OrvilleWright #WiburWright #NorthCarolina #KillDevilHills #KittyHawk #history #travel #culture #ancient #historyfact #explore Hosted on Acast. See acast.com/privacy for more information.
GET 10% OFF TRANSPARENT LABS (use code GENIRON10): https://transparentlabs.sjv.io/7mvQQQ Buy two months of BlueChew Gold, you get the third for FREE (+10% off and free shipping on your first order) with promo code GENERATIONIRON: https://bluechew.com/ GET FREE SHIPPING & 365 DAY RETURN ON QUINCE CLOTHING: http://quince.com/generationiron Watch the video podcast of this episode here: https://generationiron.com/victor-martinez-bigger-muscle-smaller-junk/ Visit the Generation Iron official website for exclusive video content, feature films, and more: https://generationiron.com/ Follow us on Instagram: https://www.instagram.com/generationiron/ Follow us on Facebook: https://www.facebook.com/GenerationIron/ Follow us on Twitter: https://twitter.com/GenerationIron
With the backing of Utah's governor and congressional delegation, President Trump has once again shrunk two of the state's national monuments. It's only the latest move in a game of political football over Utah's public lands.
Want the playbook, not just the conversation? Subscribe for deep-dive, actionable breakdowns from every episode at unchurned.substack.com.Simon Farthing, VP of Customer Success at Bloomreach, joins Josh Schachter to discuss what it actually takes to lead a customer success organization in the AI era. Drawing from his experience leading a 100-person CS and services team, Simon shares the lessons he's learned from experimenting with AI tools at scale, building internal tools, rethinking team structure, and preparing for what's next. It's a candid conversation about moving beyond AI hype and turning it into measurable operational impact.---What You'll Learn- Lessons from testing 60+ AI tools across a 100-person CS organization- How Simon built a personal AI command center in Claude- A practical framework for build vs. buy vs. build-on-top decisions- Why AI accuracy becomes non-negotiable at the executive level- Why "tool migration" is becoming a core professional skill- How Bloomreach rebalanced CSM compensation around GRR and NRR- Why Forward Deployed Engineers are emerging and where they fit---Timestamps0:00 - Preview & Intro1:28 - Meet Simon Farthing, BloomReach overview2:48 - 60+ AI tools tested, tool proliferation problem6:00 - Building "Mission Control" in Claude8:57 - Bottoms-up AI leadership structure11:00 - The accuracy problem: when AI slides hit the boardroom13:52 - Tool migration as a critical human skill15:05 - Build vs. buy vs. build-on-top framework17:00 - CSM comp redesign18:33 - The rise of the Forward Deployed Engineer (FDE)21:10 - Smaller customers and customer marketing, agentic AI---Josh is writing a book on building customer relationships. Follow his journey and insights at www.joshschachter.com---Where to Find the GuestSimon Farthing: https://www.linkedin.com/in/simon-farthing/---Where to Find the Hosts: Josh's LinkedIn: https://www.linkedin.com/in/jschachter/Unchurned Substack: https://unchurned.substack.com/
Tiny coastal town was once known for the oldest privately owned aquarium in U.S.; it's a popular place for stormwatchers. (For text and pictures, see https://offbeatoregon.com/1012c_depoe-bay-worlds-smallest-harbor.html)
Keith is joined by Jim Sheils, a seasoned real estate investor and builder who specializes in new construction and co-living properties. Together they explore why traditional long-term rentals are struggling to cash flow and how co-living—renting individual rooms in purpose-built homes—can dramatically boost returns. Jim breaks down how the model works, who the typical tenants are, and why platforms like PadSplit are essential for management, compliance, and steady occupancy. Their discussion highlights how co-living can simultaneously address the affordable housing shortage and today's "cash flow crisis" for real estate investors. Episode Page: GetRichEducation.com/617 For access to properties or free help with a GRE Investment Coach, start here: GREmarketplace.com GRE Free Investment Coaching: GREinvestmentcoach.com Get mortgage loans for investment property: RidgeLendingGroup.com or call 855-74-RIDGE or e-mail: info@RidgeLendingGroup.com Invest with Freedom Family Investments. For predictable 10-12% quarterly returns, visit FreedomFamilyInvestments.com/GRE or text FAMILY to 66866 Unlock truly passive real estate income—visit flockhomes.com/GRE today to see if your properties qualify for a 721 exchange with Flock Homes. To get in the best physical, mental, and professional shape of your life, go to DanielThomasHind.com and apply for Daniel's intensive 1-on-1 coaching for burnt-out entrepreneurs and executives. Will you please leave a review for the show? I'd be grateful. Search "how to leave an Apple Podcasts review" For advertising inquiries, visit: GetRichEducation.com/ad Best Financial Education: GetRichEducation.com Get our wealth-building newsletter free— GREletter.com Our YouTube Channel: www.youtube.com/c/GetRichEducation Follow us on Instagram: @getricheducation Complete episode transcript: Keith Weinhold 0:01 Welcome to GRE. I'm your host Keith Weinhold. For the first time ever on the show, we're talking about what some call the greatest real estate cash flow strategy today: co-living. Learn about what it is, what it is not, the pitfalls to avoid, and just how terrifically profitable co-living property can be today on Get Rich Education. What if you got your mortgage loans the same place I get mine? You sure can at Ridge Lending Group NMLS 42056. They provided GRE listeners with more loans than anyone because Ridge specializes in investment property. They'll help you build a long-term plan for growing your real estate empire with leverage. Start your prequal and even chat directly with President Caeli Ridge. While it's on your mind, start at ridgelendinggroup.com. That's ridgelendinggroup.com. Speaker 1 0:59 You're listening to the show that has created more financial freedom than nearly any show in the world. This is Get Rich Education. Keith Weinhold 1:15 Welcome to GRE from Dover, Idaho, to Dover, Delaware, and across 188 world nations. You're inside Get Rich Education nation. I'm your host Keith Weinhold. For about five years now, it's been harder to make the cash flow numbers work on long-term rentals, and that's because sure rents are up, but not as much as expenses are, and that's why for your regular income properties, builders buy down your mortgage rate for you so that it works. But now enter co-living properties, and your cash flow can be multiples higher. In fact, today we're going to talk about a model that has seven times the cash flow of a regular long-term rental, for example, instead of renting a detached single-family home to one family, if instead you divide it up into six bedrooms and rent each one of the six bedrooms to an individual, you will drive substantially more income. You've got six rent checks instead of one. That's what a co-living property is in general, and you're usually renting it to tenants that are working a lot. They're away from the property. It's often run through pad split. You'll learn more about what that means. And though co-living is lucrative, you still need the right market, the right layout, the management, which is really key here, and the right operating model. Today, we'll talk to a GRE Marketplace operator that provides co-living properties to investors like us with a management solution. And as you'll see, it gets even better than that. Although they serve just one geographic area in the U.S. which happens to be an investor advantaged area, as you'll see, this is conducive to out of state investors. If you want to own there, we're talking about co living income property today. Next week, there's something vital I want to tell you about, and I can't wait to do that. It's about the way that I talk when I meet a 25-year-old, and I learn that their only source of income is as an employee at a job. And no matter what age you are, what I'm going to share with you is going to apply to you too, and it's pretty transformative. That's next week here on the show. As for today, let's learn about co living. Jim Shields is here. Jim, welcome back to the show. Jim Sheils 3:56 Keith, good to be here. Thanks for having me. Keith Weinhold 3:58 Well, Jim, we saw each other in person a few months ago at a conference. I wanted to have you back here today to discuss co living since you're involved in it and you help other investors learn about it. And now you even started providing properties for co living specifically for them. And you know, Jim, with co living, I have seen models in the past where, oh, a single-family home it might be retrofitted, renovated to say have six bedrooms and three bathrooms, and that way you, as the owner, you could rent it to six tenants, really who are each only renting a room rather than renting the whole place to one family. In this way, tenants get cheaper rent because they're only renting one bedroom, and then for the owner, this gives them stronger cash flow because they're renting it to six different parties and not just one. So, with affordable housing really being a struggle for so many, you know, co-living is. Really taking off. So tell us more about what co living is and what it isn't, Jim. Jim Sheils 5:05 Yeah, co living was something that was brought to us by one of the owners of Pad Split. Actually, they had met me and worked with me on other new construction projects, and and kind of my evolution, as you know, Keith was I I went from doing a ton of fixer uppers for many of years to new construction, and both can get you where you want to go. But I like new constructions. But meeting some of the owners of PadSplits, I found that they were starting to have the same struggles that I was when I was rehabbing a lot of homes. You know, it's not easy to take a three bedroom or four bedroom home and turn it into an eight bedroom or a seven bedroom, and so you know, starting from scratch with new construction, we're able to set up for what they're trying to create, and what they're really trying to create with co living is that affordable housing crisis. There's a lot more single individuals out there not looking for you know whole, not needing whole family dwellings to live. They're more at a basic income level that you know keeps them quite a bit below being able to rent. Like here in Jacksonville, the average price of a one-bedroom apartment is just outside of their qualifying range. So they might have a good job and decent credit, but they can't qualify. So what we've seen co-living do, and again, working through our our management partner of Padsplit, is they are bringing in this system of putting people together in one home, still with the screening, still with certain amenities that you know really help the property get seen, wanted, and rented, and then certain managerial things, where you're providing good options for living in areas that need more affordable living, and for landlord investors like ourselves, you're providing now a new opportunity to beat what I've called the cash flow crisis. You know, we have an affordability crisis, but that's also created a cash flow crisis, Keith. As you know, and the numbers that we've been able to see are quite advantageous for both the renter for the amount they pay and for the owner for the amount they're going to cash flow. And that's what I'm seeing this new co-living movement is about. It's really landlords taking a new risk on a certain type of property and tenants getting a type of property that fits their income and their needs. Keith Weinhold 7:27 You know, Jim, philosophically, I'm thinking about assisted living homes, and when society changed sometime last century, assisted living homes became more of a trend where a lot of times that's where the elderly people went. Now we have co-living, and you kind of wish the world would be a place like where you know someone freshly out of high school or college could be married and have children, and you know just one or two jobs could float and support that family, and they would be able to afford their own home, either to own or rent. But increasingly, that's just not the world that we're living in. It hearkens back to multi-time GRE guest and legendary investor Jim Rogers. To paraphrase Jim Rogers, Jim Rogers said, "I need to invest in the world with the way that it is, not the way that I want the world to be. That's one thought that keeps coming to mind with co-living. Jim Sheils 8:29 Yeah, for me too. Honestly, Keith, when I first heard about co-living, it was a few years ago. Someone saw me speak at a mutual event similar to one we last saw each other at. I got off stage, and they came up to me and they mentioned to me we're doing these things called co-living. Keith Weinhold 8:44 Yeah. Jim Sheils 8:44 And I was kind of blinded. I was looking backwards instead of forwards, and I said, "Oh, that doesn't sound right. That doesn't sound like it could work, you know. And I had a lot of what ifs, and how do you handle this? Well, you know, as the niches really started to solidify and show real wind at its back, all those questions and doubts I had have been answered, and I've kind of become a believer in why it's working and also the results. You know, for us, build right, finance right, manage right. That's always been our model. That's what's going to help us and our investors succeed. Building it right and managing it right. We had to figure out, but figuring that out-that's key. And it's very cool to see investors today breaking the norms of saying, "Well, you know, we can't have cash flow anymore on a nice new construction property in a growth market like Jacksonville, Florida. You can't have good cash flow. Well, that's just not true with co living coming on because again you're answering the call of a forgotten tenant and their needs. Keith Weinhold 9:46 Yeah, I just think for any thoughtful investor, it's got to give them pause. But this is where society has gone. Well, we're going to talk about how profitable co living is for investors later. But first, tell us more about the nuts and bolts of how co. Jim Sheils 10:01 Yeah. So the way that it works is, first of all, you start, you know, again going with our model, build right, finance right, manage right, building it right. When you're starting with new construction, you're able to go into the property with all eyes open, without a lot of surprises, and you're able to build it right to the design that a co-living property of success would entail. You know, we're doing seven bedrooms, seven baths. We're doing 14 bedrooms, 14 baths. We're doing 20 bedrooms, 20 baths. These are all things that we worked with the owners of Padsplit as they've worked and researched areas all over the country, starting here in Northeast Florida. But what we're doing, we figured out exactly what type of build you do you need to do that's going to work? Smaller living areas. There's only one utility box. There's not you know multiple utilities. It's not a multi-unit building. So there's one utility box. You build it to that specimen of either we build anywhere from seven bedroom, seven baths, right up to 20 bedroom, 20 baths, and that's the build part. That way, you're getting into it and not having to, you know, kind of. It's really tough, Keith. I don't need to tell you with your experience to turn a smaller house into that bigger house and starting from scratch with new constructions. Great. The second thing we found was finance. Right, as you know, we do our own in-house financing. Well, a lot of the co-living since the banks didn't understand it. Just like you know, it's become a newer niche. They were locking in at 8% You know, with our in-house financing, we were able to get deals down to five and a half percent. So right there, that helps with the rate, the long-term rate, the cash flow, and then management. Again, we've managed 1000s of properties, but we've really teamed up with PadSplit to help us manage these. This is what they specialize in, not only for attracting investors into their organization, but also for managing the properties and screening the tenants, getting them in, and their management system combined with our Build Right Finance Right has been a great combination. Keith Weinhold 11:57 Pad Split, somewhat of a platform like Airbnb, but it's for co-living type properties. We'll talk more about pad split shortly. But yes, you are making these more efficient for co-living because right from the beginning, you are building them new construction specifically for co-living type of arrangements, rather than that six-bed, three-bath retrofit example I brought up near the beginning of our chat, but tell us more about who actually lives in co-living homes. What's the tenant profile like? Jim Sheils 12:29 Yeah, let's go to the opposite end. When I first heard about this years ago, Keith, I said, "Oh my gosh, this is going to be really unqualified, seedy people not working, getting into trouble, and that's just not true. Again, a lot of these people are hardworking, but they can't afford the $1395 for a one bedroom apartment, but they can afford an $825 a month room. And a lot of these people might be working at a local warehouse, at a hospital. They're very localized, blue collar, or some in training jobs, you know, extra five $600 a month makes all the difference. Where it's not going out to rent, not even including utilities, it gives them a nice place to live. So it's really entry level replaceable income people, maybe single people that work at a nearby restaurant. But again, they're trying to save more money in their pocket and not spend it on that higher expense, which the competition would be a one-bedroom apartment. Keith Weinhold 13:31 Now, the tenancy durations here are shorter than what you're going to have in long-term rentals, of course, because one part of what you do, Jim, is for years you have helped GRE followers with build-to-rent long-term rentals. The resident does get more. They're going to get a furnished room with utilities in co-living arrangements, and they're also probably going to have their utilities bundled as well. So tell us about the typical tenancy duration, and then what all the resident gets. Jim Sheils 14:00 Yeah. So the residency is going to receive all the things that they have to be turnkey. That's the bed, the desk, the dresser, the closet, the bathroom. Everything is set up there for them, and so they just move in. They're not going to have to pay for electric or water or internet. All that stuff's going to be included. There's a washer dryer normally there. Sometimes they're coin operated, other times they're just included, but that way they're not trying to take out a utility in their name or set up internet. Also, for the investor, that's a good thing because you have one master lease. If you start to do a bunch of leases, well, you could get in trouble with the rules of your community, probably of having multiple leases on one property. It's not a multi-unit building, so you can't do that. So one master lease with one utility and all utilities included allows you to do that. This makes it very easy to move in and out. And the average Tennessee might only be six months, but again, the way that it's set up, what we like about PadSplit is they. Have a very good marketing and screening process, so they're constantly marketing and screening to people in the area, and then they have their own private community with investors. So we'll build it and finance it. Our people will move over to their community for management, so the owners can speak together, and then the tenants, though they're coming through pad split system, and so what I like about that is if they try to not treat the property well, well, they're not allowed back into any pad split properties anywhere within the city. So that's really good for co living protection. So we just see that turnkey approach, and you know you've been preaching turnkey real estate for a long time. Yeah, this is a turnkey room where they're able to move in, they're able to move out easily. They can transfer to another co-living property, and by doing that, you can get people in and out very quickly, which keeps the vacancies low, even with shorter tenancy. Keith Weinhold 15:57 All right, so an average tenancy duration of about six months, and for you, the prospective investor, as you're trying to understand co-living, maybe think of it as like when you check into a hotel. Co-living residents stay longer than you stay in a hotel, but as far as all the utilities are in the room combined, all into one charge with your WiFi, your water, your electricity, your natural gas, and the room is already furnished. Just one all-inclusive payment, making it easier for that co-living tenant. And Jim, you've been talking about pad split, where you're partnering directly with them, and that's the management part of this. Of course, this is more management intensive than a long-term rental. So, tell us more about Pad Split and how it works. Sort of like an Airbnb platform, but yet for longer-term, affordable room rentals that has the property management infrastructure somewhat already built into Pad Split. Jim Sheils 16:56 Yeah, I think your comparison to Airbnb is very accurate, Keith. But it's more of a community for both tenants and investors. Airbnb, you know, I have short-term rentals and I use Airbnb. But what I've seen with Padsplit, which I think they've done a good job with, it has a community feel sharing for the investors who are a part of the Padsplit community, and so there's extra communication on your management, how properties are going, what areas are doing that. You know, great source of communication on the ongoing management. But for the tenants too, they know right where to go to find these types of properties, and the tenants also have to join this community. So there's a joining where if you want to rent a pad split, they have to join the community. They have to go through all the approvals and such, so that's pretty much how it's set up. It's seen to be very effective. Again, what held me back from this probably for about two years, Keith, was the management piece. You know, we've always managed our properties. I said, well, this is not our niche. Just like we don't manage short-term rentals, we only rent manage long-term rentals, and so I really had to watch and survey a lot of the existing investors and how they were doing, but it's nice to see someone with a good managerial system for both the tenants and the investors to work together in. Keith Weinhold 18:12 Does PadSplit handle everything like marketing and tenant screening and rent collection? Jim Sheils 18:19 So what they handle is the way that pad split works, like here in Jacksonville, where's our main market where we've started building co living properties. Is they will work hand in hand like an Airbnb, but underneath them they'll have preferred property managers that they'll work hand in hand. It so they'll handle certain things of the marketing, the tenant screening, and then some more of the mechanical PM pieces, the property management pieces. There'll be an assigned property manager that the client will be working with, the investor will be working with, and the tenant will be working with. So it's kind of a two-tiered approach. Like right now, my I'll use Airbnb, but I have a property manager for my short-term rentals. Same thing here, but we actually have the preferred management list and approvals through Padsplit. Keith Weinhold 19:06 Okay, so much of this is handled through Pad Split, but not everything. There's a second tier where they partner with local property managers in that area to, for example, help with tenant turns or help with maintenance requests. Jim Sheils 19:20 Yep, absolutely, absolutely. Keith Weinhold 19:23 Now, short-term rental hosts are used to Airbnb fees. What are Pad Split fees like? Jim Sheils 19:29 There's a monthly fee for belonging to PadSplit to keep your property occupied in there, and I don't remember exactly what it is for the tenants, but I know that they keep it affordable. So overall, you're going to be paying a little bit less than an Airbnb property that you would for you know if you use Airbnb and use a short-term property manager, it can be quite expensive. We've seen the pad splits come in below that and still achieve the goals of a good short-term rental. You know we're seeing. I know we're getting into this later, but taking those fees out and doing quite a bit of contingency because of the move and move outs, we were still seeing like a seven bedroom, seven bath based around the same price of one of our single family homes. The cash flow can be about seven times higher. Wow, on that thing, so it it really does answer the call for higher cash flow by bringing that more affordable rental in. Keith Weinhold 20:25 We're going to talk more about just how profitable it is for investors, and more about co living somewhat nascent model, a model that's actually been around for quite a while, but it's really gaining traction in making things more affordable for tenants and making things more profitable for investors, we're back with more shortly. I'm your host Keith Weinhold. This is Get Rich Education. Flock Homes helps you retire from real estate and landlording, whether it's one problem property or your whole portfolio, through a 721 exchange, deferring your capital gains tax and depreciation recapture, it's a strategy long used by the ultra wealthy. Now, mom and pop landlords can 721 the residential real estate request your initial valuation. See if your properties qualify at flockhomes.com/gre. That's flockehomes.com/gre. Let me ask you something: If you've worked hard to build wealth, is your money positioned to actually support your goals? A lot of accredited investors leave capital sitting in cash because it feels safe, but inflation and missed income opportunities can quietly erode its value. Freedom Family Investments offers freedom notes for investors seeking structured income backed by real estate. It's a straightforward approach built on real assets, not speculation. And full disclosure, I'm an investor myself. What I like is that their team walks you through how it all works, so you can decide if it aligns with your portfolio and income goals. Every investment carries risk, and nothing is guaranteed. But with a track record of consistent, on-time investor payouts, they built real credibility. Keith Weinhold 22:11 Go to freedomfamilyinvestments.com to book a clarity call, or text family to 66866. That's family to 66866. Hey, it's corporate direct Ted Sutton. Listen to Get Rich Education with Keith Weinhold, and don't quit your daydream. Welcome back to Get Research Education. I'm your host Keith Weinhold. We're talking about co-living, specifically how smart it is to get this right from the beginning and build a new construction single-family home, or something that could be larger than a single-family home with seven bedrooms, seven bathrooms, or 14 and 14, or 20 and 20, and when we think about this gym physically, the footprint. What is parking like at a seven-bedroom home or larger? And did this entail any zoning hurdles? Jim Sheils 23:15 A couple of things. It can entail zoning hurdles, so I would make sure, especially if you're doing new construction, you have someone who is a builder in the area that knows how to work with the county department or city department of how to get these approved. That's a very important things for how you submit them and how you make sure you're staying compliant. I think compliance is very important, obviously, for a rental and and parking. What we've seen right now is we look for about 50% So when I say 50% of bedrooms, so a seven-bedroom house, we're going to have three or four parking sites. There's a large majority of people that don't have a car, and that's why a lot of these are built near public transportation. But we still try to always serve, you know, based off of what we saw, the needs to stay in compliance, also working with Pad Split on lots of their designs. 50% of parking spots is usually good for the amount of bedrooms that you have. Keith Weinhold 24:08 What about zoning hurdles? What had to be met there? Jim Sheils 24:12 It's going to be a trial and error for a new builder or a person who's rehabbing a property. You know, we were not the first to come in and do this, we had watched Pad Split do it for two years and more with rehab properties, but just starting new construction, we knew Northeast Florida and what needs to be done very well, and so it's all in how you present it. Again, if you go in there with multiple utility boxes and such, you're going to be in quite an issue. So we found that we didn't have any issue with that. Also, we own some of these in some of our quad communities, so these were for larger. You know, you've worked with our quads before; they're yeah, you know, larger buildings. Well, we've been able to turn those into, for example, 20 bedroom, 20 baths, and so they operate like a small. Building, but they're in our quad communities where we wrote the HOA, and so they're allowed. And there's a plethora of parking there, so that keeps us in compliance. So really, compliance is knowing what your local city or county is going to require and knowing how to get that approved. That's very important. Keith Weinhold 25:18 Okay, so these can thrive in sort of duplex and fourplex type neighborhoods. That's where they're being built, and you have a good bit of control over that with you having your own de facto HOA there as well. And Jim, I looked at several of the co-living properties and the footprints and the floor plans, and you know I was really encouraged to learn that really the cost for one of these brand new build seven bath seven bed co living properties really isn't that much more than a regular single family home designed for one family. So talk to us about the pricing. Jim Sheils 25:57 Yeah, in Jacksonville, for example, seven bedroom seven baths, going between 325 and 345, which is very similar to single-family homes. And then our pricing for our quads, like a 2020, would be in the low nine hundreds, and that's the same as a quad as well. So we've been able to match our pricing to, and we have duplex ones. The 1414 would be mid five hundreds, which would be about the same as a Jacksonville duplex. So you know, I'm going off Jacksonville pricing right now, just as a sample. Sure. Okay. Keith Weinhold 26:30 The model I looked at was seven bed, seven bath. It had two stories. The price was 325k, and it had 1800 66 square feet. Talk to us about just how profitable that it is for an investor on a pro forma income and expenses basis, Jim. Jim Sheils 26:49 From what I've seen is if the average single family home was bringing in about $3,500, and that's a net on the year. So we're saying net, you know, cash flow on the year. Let's say it was about $3,500 on the year of a single family home that you bought. Well, your pad split property would be more closer to about 25,000 for the year. Wow! So that's a big jump. Now, with that comes a lot more rent collected, Keith. But also, and again, we can't do it here, but on the performa, there's more expenses because you have pad split. You have just very similar again. I think your analogy of a short-term rental for anyone who owns those very similar. There's more components. There's more pieces to it. There's more management. But what I did like about the performas that were first brought to us, large amount of rent collected, but also a large amount of contingency and expenses calculated in to get that number. So those are what I liked again seeing these is there's a lot of expense taken out to still reach that you know higher cash flow, but you know it's something that you will want to account for as well. You know what keeps people safe? Do your numbers. Do your numbers real. You know talk to people who have owned co-living properties, make sure that you're accounting for the extra expenses of tenants turning over more and and just more management involvement. Keith Weinhold 28:10 Okay, so on that comparison where you likened it to a long-term rental versus this model, it's about $25,000 of annual cash flow, which is about $2,000 of monthly cash flow, and was that for the seven bed, seven bath model? Jim Sheils 28:26 Yeah, and it would go up from there for the larger models. Keith Weinhold 28:29 Yeah. Now, are tenants paying in advance by the week or by the month, or how does that work? Jim Sheils 28:34 That's a key thing. They pay by the week, which I didn't realize how important that was for affordability and staying current on your rents, but from the things that I've been shown on working with Padsplit, they showed that when they charge people by the week, it's much more affordable. A lot of these people are paid weekly, so it really keeps them in good rhythm, and so they go on a weekly process. Keith Weinhold 28:57 And is this just as conducive to out of area investors like long term turnkey rentals are? Jim Sheils 29:03 Yeah. Well, that's why you always want to build right, finance right, manage right. You have lots of great connections. You can find a great builder that's willing to build them. You get good financing and then good management. So you don't have to live in the area, but you want to make sure someone is taking care of that for you. Again, I am going to say I know some people like to manage their own properties from afar. That can work with long-term rentals. I've done it with a few of mine when I left California and came here. From what I've seen, though, for co-living, the involvement-if you are from out of area-I would highly, highly recommend that you follow a manager process and work with a manager. Keith Weinhold 29:43 Meaning that you would use one of PadSplits recommended managers. Jim Sheils 29:47 I would use PadSplit with one of the preferred managers that we know well, and actually one of the managers that they highly recommend worked for our company for five years, and she's great. And for what we're. Doing, I can put a stamp of approval on it, and again, I think just seeing the involvement, these can work really well. But you want to have just like any time, but even more importantly, on these ones, you want to have management in place, especially if you're afar. Keith Weinhold 30:14 Sure. So, what could the involvement realistically look like, Jim, if that out-of-state investor is using Pad Split and using Pad Split's recommended manager. What might that investor have to do remotely? And maybe that's just on an email basis. Jim Sheils 30:32 You know, a lot can be done by email. Again, Keith, our goal-I don't think you should be spending more than two hours a month on managing your property manager. So again, just because there's more involvement, what I like is it doesn't mean there's more involvement for you. You're paying someone to set that up to handle it. You'll have to be involved somewhat. You are a property owner, but again, I don't see that you have to get involved with every little thing. In fact, I like to step back and not get too involved in my properties. I find like I just kind of go and stir things up. Let my manager do their thing. I'll manage certain big picture managerial things and in communication and overall just directionals. But you should not be getting overly involved. That's what their job is. You should not be doing that. Keith Weinhold 31:18 Now I'm a turnkey real estate investor myself, as you know, with multiple properties in various states and places, and I'm used to getting monthly emails from my manager in those markets, and that is what my owner statement looks like: income and expenses and anything that's going on with the property. But that's just on a monthly basis. Are there weekly statements for co-living managers and owners? Jim Sheils 31:42 Still rents are collected weekly, but statements still come out monthly. Keith Weinhold 31:46 Because— Jim Sheils 31:47 You want to, what you do is you collect all of the rents, but then again, it's easier to reconcile with all expenses and such that come out on a monthly basis. So it's done monthly. Keith Weinhold 31:58 Are they writing common mistakes to avoid that are developing in the space, like an investor that gets in and buys their first co living property, and then they think, "Oh gosh, I wish I would have known about this thing sooner that I didn't think about because I'm only used to long term rentals. Any common mistakes to avoid pitfalls like that with co living, Jim? Jim Sheils 32:17 Yeah, a couple of things. First, again, I can't stress enough that co living, from what I've seen and experienced, they do make money. But on the build it right, you know, going back to our build right, finance right, manage right. Just know if you're going to try one your on your own and you want to convert a home that's three bedrooms into a seven bedrooms, it is a much more tedious, involved process. Where again. Keith, like you said, are you getting things approved with the county or city? Just know that you really want to be in the know, and that's going to take some involvement. So that's my warning on if you're going to use co living and rehab your own property, financing wise. The thing that the mistakes I've seen made is people didn't see that they had higher interest rates, so you want to do that in your numbers. You know, working like with us, we're we have our own in-house financing. We're able to get it down to five and a half percent, but some of these co living banks are looking at it differently, and you might be more around 8% So you want to just do that in your numbers, and then the third pitfall, which we've you know hit on, and this is one of your real foundational rules: is management is key. And on these, where I see the biggest harm, once you get it built right and financed right, it could all fall apart if you don't have management in place. So I would just be you know some of you do it yourselfers. I have some things that I like to do it myself too, but on a more technical type of property like this, I highly encourage you get management in place to do their thing. Keith Weinhold 33:47 For sure, it is easy to make the case that the management is even more important than the property itself. One of the things that I like about what you do there, Jim, is you're so forward-thinking and you are so into making the experience as turnkey for that investor as it can possibly be, and one of those things is as you rolled this out, you had a lot of the financing hurdles rolled out right with it. Tell us more about those in-house financing options you have that you touched on specifically for co-living, and especially I'm thinking through the lens of like that seven bed, seven bath, 325k co-living property that I saw. Jim Sheils 34:28 Yeah, I mean it's nice with getting to be a builder of our size with a really good balance sheet. We're able to work with banks and buy large tranches of money or slate large tranches of money at cheaper rates than available to the public. You know, for our normal long-term rentals, we can get down to 3.75. That's for normal houses. You can't get that low because of risk factor. Banks still they like working with us, but we have gotten it down to five and a half percent. So right now, our most popular program is a 30-year fixed five and a half percent for co. You know, and again, a lot of people that have come to have said, "Holy moly, we were locked in at 8.15. We are getting it not as low as our lowest rate, but at a really good rate now. You know, below what's normally offered out there. So that's normal qualifying that you would have to do with any bank loan through us. And our counselors are happy to talk more about that. That's our most popular program for the co living for the seven seven. You're looking at a five and a half percent interest rate, 20 to 25% down. That is a super attractive rate. Is that something that the home builder helps participate in buying down discount points, or that the buyer is asked to do in order to get down to that rate? Working with us, well, it could go either way out in the marketplace, Keith. When working with us, we pay all of those required fees and points to get the lower rate. That's on us, not on the buyer. Keith Weinhold 35:50 All right, we're talking about co-living properties today-a way to supercharge your cash flow. This is one of the greatest cash flow strategies in all of residential real estate today, Jim. Is there any last thing you have to tell us about co-living? Perhaps something that I did not think about asking you that I should have. Jim Sheils 36:09 I think that's probably just what you and I talked about, like what Jim Rogers said. It's how not how I want things, but how things are. Yeah. And so for me, I held back for a few years, even when some of the founders right here in my own backyard have pad split to team up with, but I think that I couldn't picture Keith 10 years ago having a short term rental and like wait a minute I'm going to have them stay there every week and I'm going to have to furnish it you know is from the old guy doing long term rentals so I think for people just read up on it it's becoming a very interesting trend there's some very interesting statistics of why this is working, how management can be handled effectively, how to stay in compliance with your city. There's a lot of good information. This is a great starting point to our conversation today, but I don't think this niche should be ignored because the track record is already there and it is answering a need of both investors and tenants, which is pretty cool. Keith Weinhold 37:04 It has been super intriguing to learn more about this. Jim represents one of our GRE Marketplace providers. It's been great having you back on the show. Jim Sheils 37:13 Now, thanks for having me, Keith. Keith Weinhold 37:20 Yeah, a really informative episode today. If you want to learn more about co-living properties, you can do so at gremarketplace.com/co-living. That's where you'll get the investment report, floor plans, financial projections, see the exact pricing, and learn more about pad split there. I've been around this space for a while, and I know investors that own co-living properties. Some other best practices that I've learned about are that you want to have limited visitation or a zero visitation policy for your co-living tenants. As we touched on, these properties can really make money, but don't try to manage them remotely. Make the house rules unusually specific. Address guests, quiet hours, smoking, drugs, pets, parking, food storage, shared bathrooms, thermostat settings, and cleaning and abandoned belongings, and enforce the rules consistently and quickly. Because one disruptive resident can cause several good residents to leave, and in co-living, retaining household harmony that is often more valuable than retaining one problem tenant. Provide professional common area cleaning. Don't expect seven or 14 unrelated adults to collectively develop some passion for wiping down the stove. That is not going to happen. Shared areas should be cleaned at least weekly. Install bedroom locks and then smart exterior locks. Give each resident private space, eliminate shared keys, and immediately revoke keys after move out. Cameras they should generally be limited to lawful exterior and entry locations, not in private spaces. Provide excellent internet in co-living, unreliable WiFi. That is practically a habitability crisis. Use business-grade equipment, strong coverage, and have a backup plan for internet. And as an investor, it's wise for you to maintain a larger repair and turnover reserve than you would for a normal long-term rental. Keep these things in mind, and it can keep seven times the cash flow from becoming seven times the headache. Again, you can get the investment report, floor plans for the very properties we discussed today, financial projections, pricing, and get more information about pad split all at. gremarketplace.com/co-living. That's gremarketplace.com/coliving. Until next week, I'm your host Keith Weinhold. Don't quit your daydream. Speaker 1 40:15 Nothing on this show should be considered specific, personal, or professional advice. Please consult an appropriate tax, legal, real estate, financial, or business professional for individualized advice. Opinions of guests are their own. Information is not guaranteed. All investment strategies have the potential for profit or loss. The host is operating on behalf of Get Rich Education LLC exclusively. Keith Weinhold 40:43 The preceding program was brought to you by your home for wealth building getricheduceducation.com
Watch the full episode on YouTube:We first covered Baseten last year when DeepSeek mania was at peak hype. Now they have raised a monster $13B round and become one of the new cohort of AI Infra decacorns that are (with Nvidia, Intel, and the semis complex) chief beneficiaries of the Inference Inflection. We return to Baseten at the peak of the 2026 edition of Open Weights debate. Ali has published a viral breakdown of Kimi K3:And since you last saw him, Philip has spoken at AI Engineer and written the definitive book on Inference Engineering spotted all over SF:Three years ago, inference engineering barely existed as a category.Today, it is one of the most critical disciplines in AI. Inference engineering inherently tackles a different question than standard model training: “How do you turn those weights from training into a product that is fast, reliable, and affordable at scale?” Focusing on these creates an entirely new optimization problem.In one recent GLM-5.2 experiment, quantizing more of the model actually preserved its benchmark quality while increasing throughput by 20%, because the errors introduced in different layers could cancel each other out.Inference is no longer just the final step after training. It is becoming its own engineering discipline, with its own research problems, infrastructure, and increasingly specialized roles.In this episode, Baseten's Philip Kiely and Ali Taha join swyx and Vibhu to explain what actually happens after a new open model is released and what it takes to turn “we generated a token” into a fast, reliable, production-ready API.We go deep on cache-aware routing, disaggregated prefill and decode, quantization, speculative decoding, KV-cache movement, model parallelism, GPU kernels, and the race to make frontier models up to 10× faster. Philip and Ali explain why inference optimizations can still produce gains of 20%, 100%, or even 200%; how quantization errors can cancel one another out; why identical weights can behave differently across clusters; and how Baseten grafted a Kimi vision encoder onto GLM-5.2 without changing the underlying language model.The conversation then expands beyond LLMs into NVIDIA Dynamo, mega kernels, Rubin, AI-specific chips, local inference, video generation, diffusion versus autoregressive models, and the enormous compute barrier to generating coherent long-form video. Finally, we explore the convergence of training and inference, continual learning through persistent KV cache, and the emerging loop where models help optimize the infrastructure that runs them.We discuss:* What happens when a 200,000-token request enters an inference system* Cache-aware routing and reusing previously computed KV cache* Why prefill and decode are increasingly handled by different GPUs* When dedicated deployments become cheaper and more reliable than shared APIs* How speculative decoding uses a smaller model to accelerate a larger one* Tool calling, structured outputs, and what LLMs actually do* What it takes to support a new open model on day zero* Grafting Kimi's vision encoder onto GLM-5.2* Retrofitting inefficient model layers with components from other architectures* Why models sometimes collapse into repeating the same token* How hardware, kernels, and race conditions create nondeterministic failures* Preserving model fidelity while making inference faster* How quantization errors can cancel each other out* Why inference optimizations still deliver gains of 20%, 100%, and 200%* How optimized serving can make a model up to 10× faster* NVIDIA Dynamo, KV-aware routing, and distributed model serving* Speculative decoding the speculative decoder* Why local AI is about making models less dumb while data-center AI is about making them less slow* Tensor, expert, and pipeline parallelism across GPUs* Hardware-aware model design, auto-tuning, and the case against mega kernels* Rubin and why inference is becoming a systems problem* Whether modern GPUs are evolving into programmable AI ASICs* Why enormous models like Kimi K3 require GB300-class hardware* Why open-source video generation still trails Veo, Kling, and other closed models* The quadratic attention bottleneck behind long-form AI video* Autoregressive video, real-time generation, and compounding quality drift* Why future video systems may combine autoregressive and diffusion architectures* Training for inference and inference for training* Continuous post-training, deployment, evaluation, and improvement loops* How GLM-5.2 helped optimize the kernels serving GLM-5.2 itself* Why faster networking could unlock dramatically faster decoding* Continual learning, KV-cache compaction, and persistent model memoryShow Notes* How to build a day-0 API for Kimi K3* 22580: From GPT2 to Kimi3, ExplainedPhilip Kiely* LinkedIn: https://www.linkedin.com/in/philipkiely* X: https://x.com/philipkiely* Inference Engineering: https://www.baseten.co/inference-engineering/Ali Taha* LinkedIn: https://www.linkedin.com/in/aliestaha/* X: https://x.com/waterloointernTimestamps00:00:00 Introduction and the 200K-Token Prompt00:03:18 Dedicated Deployments, Speculative Decoding, and Tool Calling00:11:26 Launching Production-Ready Open Models00:19:06 Model Retrofits, Failure Modes, and Nondeterminism00:28:22 Quantization and Canceling Errors00:32:15 The Race to 10× Faster Inference00:40:48 Dynamo, Speculation, and Local vs. Data-Center AI00:50:18 Model Parallelism, Auto-Tuning, and Mega Kernels01:00:55 Rubin, GPUs vs. ASICs, and Custom AI Chips01:10:03 Giant Models and the Limits of GPU Memory01:12:42 AI Video, Quadratic Attention, and Autoregressive Generation01:21:47 Audio, Images, and Diffusion Models01:27:32 Training, Self-Optimizing Models, and Continual Learning01:40:06 Closing ThoughtsTranscriptIntroduction: Baseten, Waterloo Intern, and Inference EngineeringSwyx [00:00:00]: Okay, we're here in the studio with Philip, old friend from Inference Engineering, the book, as well as Baseten and everything that you've done, you and I have done before, as well as Ali. Welcome.Ali [00:00:15]: Pleasure to meet you.Swyx [00:00:15]: Waterloo intern.Ali [00:00:16]: Waterloo intern, always.Swyx [00:00:17]: When did you get “Waterloo intern” as a handle?Ali [00:00:19]: As a handle? Oh.Ali [00:00:20]: I think the rebranding happened mid-March. When I saw it was open, I was like, “I have to take it. Up for grabs.”Philip [00:00:26]: The problem is that Ali is really good at his job and is not gonna be an intern much longer.Philip [00:00:30]: So we have to figure out who's gonna get the handle.Ali [00:00:33]: Well, I'll pass the torch over to the next intern.Swyx [00:00:34]: Oh, okay. It can be, like, you just pass it to another Waterloo grad.Ali [00:00:37]: To another Waterloo intern. No, bruh.Philip [00:00:39]: Yeah.Ali [00:00:39]: Intern.Swyx [00:00:40]: Intern, yeah.Ali [00:00:40]: And no.Philip [00:00:41]: You gotta get an intern from Waterloo.Ali [00:00:42]: Yeah, I've gotta get an intern from Waterloo.Swyx [00:00:44]: Right.Ali [00:00:44]: But they have to follow the path.Swyx [00:00:45]: Oh, it could, but it could come from Baseten, so it's like whoever Baseten gets from Waterloo.Ali [00:00:48]: Right.Swyx [00:00:49]: Has the title of Waterloo.Ali [00:00:50]: It stays in the ecosystem.Philip [00:00:51]: Exactly.Ali [00:00:52]: Halfway through the internship, you either get it or you're out.Philip [00:00:55]: You should also do, like, a big graduation ceremony where you change the handle.Ali [00:00:59]: Just say it.Philip [00:00:59]: For everybody.Swyx [00:01:00]: You guys are good at ceremonies, clearly. We had a nice launch of the book, very successful. But before we get into all that, I wanna start off with a fun question for you. Okay, you're an expert inference engineer. What happens when I send a long query, say two hundred thousand tokens into Baseten's inference? What's the process of query through GPU model routing, balancing, all that? What is all the stuff that we don't think about?Long Context Requests, KV Cache, and Cache-Aware RoutingPhilip [00:01:26]: With a long query specifically, the first thing that I'm gonna ask is, “Have you sent me this query before, or at least part of it?” and I really hope you have, because it's gonna be a lot easier for me and a lot cheaper for you. So the first thing that we're gonna look at is some cache-aware routing, where we're going to see, we probably have a number of instances, a number of replicas up serving whatever model you're hitting. We want to send this one to something with, number one, available prefill workers, and number two, ideally some cached input already there so that we can skip prefill on at least part of these two hundred thousand tokens. If you're doing two hundred thousand tokens, it's probably coding or a multi-turn agent or something where you would expect to have that cached. If you don't, we're gonna have to send it to a prefill worker. We've at least on certain models disaggregated prefill and decode, so you're going to have one set of GPUs that's solely going to process the input, create the KV cache, and get you your first token, and then that's going to be passed over to a separate set of GPUs, which is going to run decode. We're going to iteratively make those tokens. We're probably going to have some speculator model in front of that. I'm going to assume that you're doing coding, and because of that, our speculator model, which assumes you're doing coding, is gonna have a high draft token acceptance rate. If I'm wrong and you're asking me to summarize every Harry Potter book, it's gonna be slower. And then we stream that output to you and account for it, charge you, a couple of pennies and say, “Hey, would you like to send another one?”Swyx [00:03:04]: Except Baseten doesn't charge by pennies.Philip [00:03:07]: Well, yeah, we charge. I'm assuming that we're talking about the public model APIs. If you are setting up a dedicated deployment, then yeah, it's not pennies.Public APIs vs. Dedicated DeploymentsSwyx [00:03:18]: Yeah, one of the key differentiators when I was talking with Baseten initially was that people who want very high volume just need to rent by the box, ‘cause then it's up to you to figure out how to saturate the box.Ali [00:03:31]: And more often than not, it's, like, way cheaper if you're pushing, like, millions of tokens per hour, if you just pay per hour instead of pay per token.Philip [00:03:37]: Yeah, they do. I think that we've increasingly seen a lot of demand for the pay per token APIs, just because everyone wants to try open models, and then once they find a use case that's really sticky, then they move over to dedicated.Swyx [00:03:51]: Is there a best practice on when it's time to swap over?Philip [00:03:54]: Couple reasons. Yeah, reliability, that's a big one, right?Ali [00:03:57]: Like, if they have a very specific use case, they want you to train something specifically for them, like they want their own spec dec, for instance, for their own traffic.Swyx [00:04:04]: Spec dec is speculative decoding.Speculative Decoding and Custom SpeculatorsAli [00:04:05]: Speculative decoding, yeah.Swyx [00:04:07]: You have to explain.Ali [00:04:07]: Sorry. Like, speculative decoding is like, if you have a huge model, right? And so the model is going to be generating one token at a time every single turn, every single forward pass. So we attach, like, this little, like, parasite, like this layer that goes on top of the model, and this model just has to predict. It does three very fast autoregressive forward passes, and it will predict, like, three certain tokens, and then you do one forward stage over the entire original model in order to see if those predictions were correct or not, and then you accept them or you reject them. Now, this draft model is traffic specific, so if you, like, Philip said, if you're summarizing Harry Potter books, I can train exclusively that draft model on Harry Potter books, and I can guarantee you that I'm gonna accept the three tokens every single time. And so with that case, I increase your decode speed. I wouldn't be able to provide this to you if you're a shared endpointSwyx [00:04:53]: YeahAli [00:04:53]: ‘cause I have no idea if you're doing Harry Potter, if you're doing coding, if you're doing English. We don't know. Also, there was a thing in the book that mentioned that if they really cared about a specific threshold, chapter four, I think. Do you remember that?Philip [00:05:06]: Yeah. The things that you can do is you can set a specific, like, batch sizing, a specific, like, parallelism strategy if you're trying to optimize for, like, throughput versus latency. You can. Maybe a NVFP4 quant doesn't pass your benchmarks and you wanna run a model at higher precision, you could do that. There's just a bunch of reasons why you might wanna have your own endpoint and the biggest one, of course, just being, like, you don't have to deal with someone else doing a hundred million tokens of benchmarking traffic at the endpoint when you happen to be trying to serve your users.Swyx [00:05:40]: Yeah. I think one thing that is. That is a classic journey. Like, it's people is asking the, what happens when you type Google into the browser. Tool calling, is that just, you're generating JSON or is there more complication beyond that?Tool Calling, JSON, and Structured OutputsAli [00:05:58]: Certain customers that we have, they have their own post-trained models, and so they demand a tool calling that's not just, like parse a file or go find the weather. It's something that's very specific and you have to do post-training on this. And if the post-training on the model is not good or if the quantization after the post-training to get the inference to be fast, the model will struggle reading the JSON file and reading the tool calling. But it doesn't require its own like sandbox. It's not like it's going to use that tool calling to like escape a sandbox or like it doesn't have to be contained. It can just be a normal dedicated deployment. The challenge with tool calling more and more seems to be that the companies want certain tool calling which is a very sensitive thing to train. And because you're dealing with all of the JSON outputs, if it doesn't like close the end of the request in a very certain manner, you end up with a model that did the tool calling and like the thinking and so as a result of that, it didn't see the result and just hallucinated the result as it decoded. That seems to be the most challenging thing with tool calling, not really the sandboxes model.Philip [00:06:56]: Yeah, that's a challenge on the training side and then on the inference side, there's work that you can do to scope the possible output. So we published this at this point close to two years ago, the solution to this problem which is you make a state machine and you use that to constrain the output to a specific format. So this is the structured output problem. If you remember backSwyx [00:07:27]: Yeah, the specific grammar is,Philip [00:07:29]: Yeah, exactlySwyx [00:07:30]: GML had this thing.Philip [00:07:31]: Yeah. So it's like the old-school “make sure this is only JSON”, return only JSON orSwyx [00:07:38]: YeahPhilip [00:07:38]: Grandma's gonna die type of prompts.Swyx [00:07:39]: Is it BNF grammar? At some point OpenAI had released a thing that was like, yeah, if you want to constrain your output, write BNF grammar, back as NOR.Philip [00:07:47]: In our inference system, it's just a specified output format. And you get the guarantee that your output's gonna be structured along that format. And so applying that to tool calls can like help cut down on. You can still call the wrong tool or call no tool. It doesn't solve the certainty problem but it at least solves the output structuring problemSwyx [00:08:10]: YeahPhilip [00:08:10]: Within tool calls.Swyx [00:08:12]: And MCP is just another form of tool, right.Philip [00:08:14]: Yeah, exactly.Swyx [00:08:15]: As far as there's no special thing there.Philip [00:08:16]: The thing I'm always like explaining to people is the LLM is not capable of doing anything. It's only capable of making suggestions of what to do and then if those suggestions are formatted in a certain way and applied to a system that knows what to do with them, then an action occurs.Swyx [00:08:32]: Yeah. Part of the fun stuff is, this is solved outside of tool calling too. Like in an agent loop if the output is not correct or you're right, like reasoning, tool calling was done in the reasoning trace, just be like, “Oh, I don't know what to do. Let me just try again.” And it might get there after a few tries. And on your point of training, sometimes this is harder in smaller models, so you don't have the same exact quality outputAli [00:08:56]: Right.Swyx [00:08:57]: When you just swap from a big model, right?Ali [00:08:59]: Yeah. I will say that, before, I think we need to go back to inference engineering proper.Ali [00:09:04]: But, I had expected that something would replace JSON because it's hard to stream JSON ‘cause JSON must be complete and you must have open and close brackets and everything. So it's hard to parse something or validate something while it's being streamed. So people invented all sorts of things that are like, I forget the name of some of these alternatives, but it's something like TOML, something like YAML. But JSON seems to be dominant still.Philip [00:09:30]: The JSON outputs aren't that long, right? Like you could have a long-- ‘cause tool calls also contain the arguments in them and perhaps for a certain tool you might pass like a very long argument. But my impression of the median tool call is that it's a relatively small number of tokens, right? So I would expect that speculators are generally fairly good at something as formatted as JSON. And so you would have like a pretty fast decode step there and that the streaming wouldn't be as valuable, but maybe I'm wrong about that.Ali [00:10:02]: I think you're also bounded by the software or that the model is gonna integrate with if the software is built with JSON for the tool calls or if the company that you'- if your customer says that this is how our software works and our tools are interfaced with JSON, you can ask them to like, change their software and say like, “Yeah, this is gonna be better for the model.” but like with the right training shouldn't be that much of a difference. Also more profitable if it outputs more tokens probably.Swyx [00:10:25]: Depends on your business model.Swyx [00:10:27]: It really depends. But I will say that, as a writer with like experience a lot with generated output, I do try to move from text to JSON text which is very long JSON, right? Like there's paragraphs in every field because I'm trying to structure it, right?Philip [00:10:44]: Right.Swyx [00:10:44]: I want you to first make factual statements, then make opinions then make bullet point summaries, have dates, have entity references have your sources for references, all these things. Anyway, so these are things that like I think people who really experiment with structural output have to really care about. But, let's, let's recurse up the stack a little bit. Before we started recording, you mentioned something really cool, which is that there's a lot of engineering that-- inference engineering that goes on when a new model provider releases a new model, right? So let's call it GLM-5.2, Kimi K3. I had previously assumed, especially if it's like, well, GLM 5 to 5.1 to GLM-5.2, like that you've supported them before. Is it that much work?What It Takes to Support a New Open ModelAli [00:11:26]: It's a lot of work.Swyx [00:11:28]: Yeah. Okay. So like, a lot of people, all you guys, right whenever a new model launch like, people rush to say like, “Oh, Hugging Face supports this, Fireworks supports this, Spacetime supports this,” and I'm like, “Yeah, of course we support it.” But what goes into that? What goes intoPhilip [00:11:40]: I think it's more than just support it too, right? It benefits the consumer a lot. Like I think it was with Kimi K2.5 or GLM-5.2 the latest, there was an inference war, right? X provider is at 90 tokens a second. The next day we're at 150. The nextSwyx [00:11:55]: I kinda kicked that off with the GLM-5.2.Swyx [00:11:58]: I wrote a Twitter article about. It got like half a million views,Ali [00:12:02]: Based on being numberSwyx [00:12:03]: YeahAli [00:12:04]: Or it's for something else.Swyx [00:12:05]: Yeah. Which,Ali [00:12:06]: Oh my GodSwyx [00:12:07]: Which then got everyone really excited about, hey, how can we, bend tracks a little bit further and,Philip [00:12:14]: There's a difference between support the model, as in I can make a token out of this model, and support a model, as in I have a production-ready API from this model.Philip [00:12:26]: Getting to the point of I can make a token out of this model is not that hard because generally the, open source inference engines, vLLM, SGLang of the world oftentimes even receive weights ahead of time, maintainers do, or the people making the model merge PRs to ensure support. So you generally can, just get it working on the standard open source stack without too much pain in most cases. The challenge is, every inference company is gonna have own proprietary stack. Some open source components, some in-house stuff. And for any arbitrary model, there's going to be some new stuff. Sometimes you get lucky, like K, two five to two six was, like, pretty similar.Quantization, Speculators, and Production ReadinessAli [00:13:16]: Yeah. It was pure continued post-trainingPhilip [00:13:18]: YeahAli [00:13:18]: If I remember correctly.Philip [00:13:19]: Even in those cases, there's still stuff you have to do. You have to redo the quantization work. You're taking the model from. Generally, these models are not released in NVFP4, and we want them to be in NVFP4 for maximum Blackwell compatibility. So we have to perform that quantization, and, calibrate the quantization to make sure that we're not causing any regression in the model's intelligence. And then we also have to train the speculator, as we've talked about. Generally, we have. We have ZDR, zero data retention on our model APIs, so we don't know exactly the traffic that people are sending us, but we know what's popular. We know that coding use cases are popular. We know that agents, agentic use cases are popular. So we can get public data sets that are representative of that traffic and train general speculators. Now, with speculators today, you need to train the speculator using the base model itself because you're getting hidden states out of the model from running inference on these specific prompts, and that is the training data you use to create the speculator. So there's that process which you need the real model weights for. And then there's of course just the process of, standing up all the infrastructure behind it, loading all this stuff, testing it. And then when there's a new model with a newer architecture, I think that, like, the DeepSeek models tend to be the most challenging as they have, like, the most novel architectural stuff going on, model after model. But every new model has something. Kimi K2 had. Oh, sorry, GLM-5.2 hadAli [00:14:53]: Sparse attention.Philip [00:14:54]: Yeah,Ali [00:14:54]: YeahPhilip [00:14:54]: the DSA.Ali [00:14:55]: Right. Which is brought from DeepSeek.Philip [00:14:57]: Yeah. AndAli [00:14:59]: So you can copy-paste then?Philip [00:15:01]: It kindAli [00:15:01]: I don't know how this works.Philip [00:15:02]: So, like we had to, like, build support for that into our runtime. And you're right, like it is really interesting the way that all of these open source labs borrow from each other. For example, like GLM-5.2 doesn't have vision. So something that, Haley, a guy on our team, if we could take a look at this, he, like, grafted the Kimi vision encoder onto GLM-5.2.Retrofitting Vision into GLM-5.2Ali [00:15:27]: We'll be training the projector.Philip [00:15:28]: Exactly. So if you think about, like, the encoder, there's the encoder, which is the part that looks at the image and turns it into latent information, and then there's the projector which likeAli [00:15:38]: You can say latent space. It's okay.Philip [00:15:41]: And then there's the projector that maps it onto, the model itself, and then there's the model weights. You don't wanna mess with the model weights because you run a chance of making the model dumber at something else for the purpose of giving it vision. So instead, Haley started with just a projector, which is only a handful of millions of parameters.Ali [00:16:02]: That would be, yeah.Philip [00:16:02]: Yeah.Ali [00:16:03]: Can you show the training one?Ali [00:16:04]: Like the way it groksPhilip [00:16:05]: YeahAli [00:16:06]: Very interesting.Philip [00:16:06]: And maybeAli [00:16:07]: That right therePhilip [00:16:07]: Maybe Ali, you should take it from here. You've got a betterAli [00:16:10]: Ooh, double the sandPhilip [00:16:11]: Understanding of this than I do.Ali [00:16:11]: Yeah. You can see, like, he. The way he trained this is really cool. At the beginning, he was training it using just like, “Here's a picture of a mountain. Can you describe what's in this mountain?” And that caused it just like the first, learning walls. Like here you can see this all we're trying to teach it is to translate the encoded. Like it's already taken the encoder from Kimi K. It's taken the image. It'Philip [00:16:31]: Yeah. FrozenAli [00:16:31]: FrozenPhilip [00:16:32]: With adapter.Ali [00:16:32]: Exactly.Philip [00:16:33]: Yeah.Ali [00:16:33]: So the brain is frozen and the eyes are frozen. It's just we're tryingPhilip [00:16:37]: AlignAli [00:16:38]: Interconnect between the eye and the brain, right? So the projector. And so you take the tokens and then he's like, “Oh, can you describe what's in this image?” And he's like, “Oh, it's a mountain,” or it's a person or it's a human, whatever the case is. But that didn't cause complete understanding. So he changed it such that every image was associated with a data set of questions. Like, does this image have a white male? Does this image have birds in the top corner? Does this image have a scientist in it? All of that stuff. And it would have to answer questions correctly. And using not just training on describing an image, but being able to answer question, another question, answer over time. Like you can see the grokking, which is like genuinely insane, that retrofitting vision into a large LLM can learn to that extent. And even for images that it doesn't perform well on, for instance, if you ask it a picture of like Stephen Hawking, “Who is this?” Maybe it doesn't get it, but it will say something like, “This is Albert Einstein.” Like it still understandsPhilip [00:17:25]: Close enoughAli [00:17:26]: That this is a scientist who is a man who has, some significant achievements, all that stuff. So that's like really cool.Philip [00:17:32]: Yeah. So, we've covered Hao Tian before, who the author of the LLaVA paper that did this, a while ago. And I think that's very foundational work for anyone who hasn't done vision work before.Ali [00:17:41]: Same with the CLIP and MetaCLIP, where you go from just captioning to building out questionsPhilip [00:17:47]: RightAli [00:17:47]: Off the image and how much better you can get performance.Philip [00:17:50]: Right. Right. Right. Yeah. But what's, what's so exciting about this is if you look at a model like this. Now, this is a little bit more of a research project. It's not. It got to 56% on MMLU Pro, I think. So not quite frontier. But if you're running this model, you haven't suffered any loss on your GLM-5.2 quality. If you don't have an image, it'll just behave exactly the way it used to. And ultimatelyAli [00:18:14]: Which in the inference code you literally do not include the other part, right?Philip [00:18:18]: Yeah. You would just skip the encoder if you don't have an image input.Ali [00:18:22]: Okay.Philip [00:18:22]: Just confirming.Philip [00:18:23]: YeahAli [00:18:23]: Does it affect a lot on the overall inference side? Like you're not adding much, you're adding a very small vision encoder. These are typically likePhilip [00:18:30]: They're super fineAli [00:18:31]: Less than a billion parameters, right?Philip [00:18:32]: Yeah. It's, - There's a little bit less standardization among vision encodersSwyx [00:18:37]: YeahPhilip [00:18:37]: So the support matrix can be a little bit, sparser. But overall, yeah, it's a pretty, it's a pretty minor component of the overall system. And ultimately what you get out of the system is all of a sudden you have Kimi Vision, GLM weights, and DeepSeek attention all in one model.Open Source Model Grafting and Franken-MergesPhilip [00:18:56]: And that's, I think, a lot of the power and beauty of open source, is that you can take all of these different components and combine them together into a system that's better than anyoneSwyx [00:19:05]: YeahPhilip [00:19:05]: Can be individually.Swyx [00:19:06]: People used to say that you would also do Franken-merges where you would take likePhilip [00:19:10]: YeahSwyx [00:19:10]: Layers from each model.Swyx [00:19:11]: Does anyone do that anymore?Ali [00:19:13]: Well, to your point previously when you were mentioning like, the work that goes into supporting a model when it first comes out, like GLM-5.2 or MiniMax M3 or whatever the case is. Sometimes you do have to like, you do have to switch out some things. Like, for instance, the MiniMax M3 head uses full attention, and with full attention you end up with this like insane bottleneck in spec dec ‘cause you're doing auto-regressive token generation for three tokens, and you're doing this like N squared over all of the tokens that are in your sequence. Your KV cache is like very large because it's not sparse, it's not top K. So we find it better to like, okay, we're gonna replace this, we're gonna replace this layer with a layer from another model that's using like GQA, for instance. And then just with the right training, you can get it to have the same acceptance rate. So it is very possible to retrofit layers from other models and very much needed. If a layer is like inefficient, the training just becomes the challenge, like how do you ensure that you train it properly? Which again to your earlier point is like the mesh between training and inference. As in like you need very good training in order to do fast inference. That's like, I feel like more and more becoming true.Swyx [00:20:21]: Yeah. Anything else on the support side when you say like get it to fully production ready?Loop Detection, Race Conditions, and Non-DeterminismPhilip [00:20:26]: Yeah. I think that there's also a question of just, we can test a model to a pretty extensive degree, but we're trying to get it out quickly and then you see a bunch of other people test it and you get interesting results. There was an issue with, GLM briefly where we had some like mode collapses where it would just output the same token over and over again for certain prompts on certain temperatures. Like once you expose an endpoint to the real world, there's going to be, so many more varieties of things given to it that you're able to, discover and patch things. So it's not just a, day zero process, it's then like for the first week, for the first month, if a model remains popular, like how do you both fix bugs and then continue to push the envelope on performance?Ali [00:21:21]: What do you mean you don't want your model outputting S?Swyx [00:21:24]: Is there loop detection on that stuff, by the way? It still happens like quite a lot, which is surprising.Ali [00:21:30]: We have like we, in our endpoint, like if a model was to output the same token like four plus times, we just cut the generation. We say like, “Oh, sorry, this-- Like try again,” or like we will reprocess the request. ‘Cause we know then, like if it, like if, yeah, it's four times the same token, it's probably collapsed.Swyx [00:21:45]: Yeah. Is there a way to opt out in case I really want that?Ali [00:21:48]: You want that?Ali [00:21:50]: I think there's a way that we have to handle it. I'm not exactly certain, but I feel like in certain models, like when they output something like you can imagine, like a table for instance, and so they want, they wanna draw like 12 dashes and 12 dashes. Yeah, I think there's a way for that to happen. I think we only do it on certain tokens. Like we exclude certain special characters.Swyx [00:22:07]: Yeah.Ali [00:22:07]: So we only do it on like certain like S is the most common almost. GLM-5.2Swyx [00:22:11]: OhAli [00:22:11]: And I think it was DSV 4 as well. Like you'd just have like looping issues where like you literallySwyx [00:22:17]: ItAli [00:22:17]: Just have like S.Swyx [00:22:18]: Yeah. Is there a special, something special about S? No, just randomlyAli [00:22:21]: It just seems to be the one token involved.Swyx [00:22:23]: Yeah. And it'Philip [00:22:24]: Is thereSwyx [00:22:24]: And it's only temperature 0Ali [00:22:27]: NoSwyx [00:22:27]: Even at other temperaturesAli [00:22:27]: Even at like 0.9 or whatever, it will still, it will still collapse.Swyx [00:22:30]: That's weird, right?Ali [00:22:30]: It's, it is an inference problem to be honest, like a software problem. Like oftentimes, the image you run will-- like NVIDIA will release an image for instance, and if we will upstream the changes from their latest TensorRT-LLM image into our stack, we'll find that it fixes it. Or oftentimes this will only happen in an inference engine that you're using like SGLang. But if you were to switch to vLLM, that isn't the case. So it seems to be like an extremely like deterministic software issue and not really a model issue. It's not like a weights problem. Like I'- we'll say like, “Oh, it's a problem with the quant. We did PTQ wrong,” right? But that isn't, that doesn't make sense because the same weights used with a different inference engine does not repeat the problem. And sometimes it's, the kernels that are being used in the backend have like these very subtle sometimes race conditions, where if you were to use this model hosted on one cluster, you will never get this problem.Swyx [00:23:19]: Oh my God.Ali [00:23:19]: But if you host it on a different cluster, you will. And the reason is the KV cache transfer from a node to node in that one cluster is using a slower interconnect than the node to node in another cluster. So that exposes the race, whereas in another cluster it doesn't. So then you end up just like, okay, this model is not gonna be hosted on this cluster. We're gonna host it on, another cluster because that cluster exposed that problem. But then it ends up with like, okay, is it the software? Is it the model weights or is it the hardware?Swyx [00:23:42]: There is a thing about this with temperature 0 still not being deterministic, right?Ali [00:23:46]: Right.Swyx [00:23:46]: Mostly because of hardware. Even at temperature 0 same model, you won't always get the same output.Swyx [00:23:52]: Even-- But I'm surprised by the race condition one because, I thought PyTorch was a graph that like guarantees that you at least, execute things in the right order.Ali [00:24:02]: Well, yeah, true. Like I'm not, I'm not saying that there is. Like well, you have things like PTL optimizations where like you can start a kernel before the end of the previous kernel, and that's like ‘cause you want to do that because there'sSwyx [00:24:12]: It's like pipeliningAli [00:24:12]: Expense. Exactly.Swyx [00:24:13]: Yeah.Ali [00:24:13]: But it'- But you don't do it cleanly. Like you overlap a little bit of the execution. No, it is very possible that the kernel itself, like that one block that is supposed to be running in this instance of time, that kernel itself has a race condition. For instance, like a missing barrier. Like often if you're designing a kernel and you want it to make it to be very fast, if you don't test it extensively, you'll, you'll have certain threads access data points from registers before they've been written to by other threadsSwyx [00:24:36]: YeahAli [00:24:36]: For example, because like your barrier is wrong or your synchronization was wrong. But yeah, like the testing itself is very difficult in those like, andSwyx [00:24:42]: And there's no like borrow checkerAli [00:24:45]: What does that mean?Swyx [00:24:46]: Like Rust. Like the. If you're trying to have like memory safety It sounds like a comparable problem.Ali [00:24:52]: Well, yes, but you're working in CUDA, right, NVIDIA GPUs. Like- You just need a higher level language like modular Maybe that's what modular is supposed to do. I don't know.Quantization Quality and Vendor FidelityVibhu [00:25:00]: How do you see keeping quality of the model? So you talked about all these steps of, okay, you gotta do quantization, train your own speculative decoderAli [00:25:07]: RightVibhu [00:25:07]: Run on different hardware. Looking at other model providers, okay, you kicked off a inference speed race on the consumer end. What goes into keeping quality the same across them, right? Sure, you can run benchmarksAli [00:25:22]: YeahVibhu [00:25:22]: But, like, how do you determine how much quantization are there standards? What goes intoPhilip [00:25:27]: There's a few things on quality. Most inference optimizations are lossless. KV caching, for example. You are just recomputing or preventing recomputing the same values. Speculation, of course, if a draft token is wrong, it gets rejected. The main lossy optimization is quantization. And that really comes down to, number one, data format, number two, which parts of the model you choose to quantize, which layers, and number three, like doing a lot of calibration on the quantized weights, to ensure that you're preserving all the outliers. There's other tricks that you can do, though. A big one is long context, ‘cause one thing you asked at, right at the beginning is, “Oh, what's gonna happen if I send a 200,000 token request in?” So with a long input sequence, you need to, store a lot more information. You need to process a lot more tokens. And so even if a model has a context of a certain length, you might, as an inference provider, choose to build an API with a shorter context length, and of course a full length one as well. Because if someone doesn't need the full million token context, for example, you can get them better performance. I don't know if that's exactly like quality of the model. The way that I think about quality is to what degree are we faithfully serving the original model? If you think of a golden implementation of a model that performs exactly the way the model is designed to perform, I think of quality as how close are we getting to that, 100% fidelity of the model.Philip [00:27:13]: You can also, of course, think about quality from the training side and how do you push yourself past 100%. But when I think about purely inference optimizations, it's getting faster while staying as close to that 100% fidelity mark as possible. And certainly our standard internally is that, like you should not be able to tell the difference between our API and a, official API. I think Kimi in particular does a good job of vendor benchmarking hereAli [00:27:41]: YesPhilip [00:27:41]: Where they haveAli [00:27:42]: They released an actual vendor benchmark.Philip [00:27:43]: Exactly, yeah.Ali [00:27:44]: ‘Cause they accused, some people, Amazon? There was some provider that was not doing very well on Kimi's benchmark.Philip [00:27:50]: Yeah.Philip [00:27:51]: So, with Reflect we probablyVibhu [00:27:52]: This was a long time ago, right?Philip [00:27:54]: No.Ali [00:27:54]: Yeah, like threeVibhu [00:27:55]: They alsoAli [00:27:55]: Four, five months agoVibhu [00:27:57]: This also happened with, I don't remember which model, but they pulled out quite a few, and then they started a whole chart about this. It might have beenPhilip [00:28:03]: Kimi Vendor Verifier.Ali [00:28:04]: Yeah.Philip [00:28:05]: Yeah.Ali [00:28:05]: Yeah, ‘cause you, ‘cause you'd be pissed, right? Like if you'Philip [00:28:07]: Yeah.Ali [00:28:07]: If like if I'm a consumer and I'm using like Amazon's endpoint for instance, and I've used Kimi and I'm like, “Oh my God, like this is bad,” I'm not gonna say, “Oh, Amazon quantized the model in a bad way.” I'm gonna say, “Oh, Kimi sucks.” Right?Philip [00:28:17]: Yeah.Ali [00:28:17]: So it seems like that makes sense.Philip [00:28:19]: Yeah, they care. They care.Vibhu [00:28:21]: Justifiably.Ali [00:28:21]: Yeah, justifiably.Vibhu [00:28:22]: This is probably a stupid question, but just checking, has anything improved from main quantization?Philip [00:28:28]: Yeah.Vibhu [00:28:28]: Like, is quantization always strictly worse?Ali [00:28:30]: Well technicallyVibhu [00:28:32]: NoAli [00:28:32]: It's a lossy. QuantizationPhilip [00:28:33]: YeahAli [00:28:33]: Is a lossy, it's a lossy implementation.Philip [00:28:36]: Speed improvesVibhu [00:28:36]: Speed improves.Ali [00:28:37]: It the number, likeVibhu [00:28:38]: No, I' always look for inverse scaling laws.Philip [00:28:40]: Yeah.Ali [00:28:40]: Yeah.Vibhu [00:28:40]: This is something I learned from Noam Brown, where like things that normally act in one direction sometimes do.Philip [00:28:45]: Well, technically when you run a benchmark, because these models are deterministic, sometimes your,Ali [00:28:52]: YeahPhilip [00:28:52]: NVFP4 quant is like, two basis points higher than yourAli [00:28:56]: No, it's noise. It's noise.Philip [00:28:57]: Yeah, exactly. I'm like, yeah, it's, it's within. That's why I always say within margin of error.Philip [00:29:01]: And I stopped saying that because everyone assumes that what is, well, within some margin of error, we're barely inside of that to the worst, so we're saying. But yeah, sometimes it's just like, gives you a higher output score. But like Ali said, that's noise. To my knowledge, you're not necessarily making the results better. You're just trying to, again, like keep your fidelity as close to 100% to the original model.Layer Selection, KL Divergence, and Better QuantizationAli [00:29:27]: There is, to your point, research that we did on MP. I don't know if you are able to pullPhilip [00:29:31]: YeahAli [00:29:32]: A tweet we did. One of our research interns, Joshua, I think it's a tweet on how we have 20% better quantized GLM-5.2 than NVIDIA. Essentially what we found throughout like this month research is, okay, quantization is a lossy. It's. You're compressing the data from, occupying 16 bits to occupying, four bits, for instance. And so you're losing some information, and you're trying to minimize that. And so when I say that I'm gonna quantize the model, my job becomes how do I find the layers that I can quantize, and how to find the layers to not. For instance, with image models, I don't quantize modulation layers, and I don't quantize out projections because those two are. Like out projection is what you see as the user. Modulation is what the model sees or understands. Right, exactly. And so to his paper, do you have the. It doesn't have the. Yeah. It's a long paper. I don't know if I can findVibhu [00:30:25]: If there's a part to search or it's probably in the thread.Ali [00:30:28]: It's probably in the thread.Vibhu [00:30:29]: Yeah.Ali [00:30:29]: But the long and the short is it is very possible that quantizing more of the model makes the results. Like if I have a model that I quantize layers one, five, and 10, and another model where I only quantize layers one and It is possible that the model in which I quantized more information is going to perform better because the quantization errors have canceled out. And so what Joshua showed in his mathematical proof where he had like a verifier in, is that you can predict which layers are going to have quantization errors that will cancel out with each other, and you choose to quantize those layers. And so the result of doing this mathematical quantization is you end up with a model that's 20% more quantized than another provider, so you get 20% more throughput of it because there's more layers than running an NVFP4, and your quality is better than that other quant because the layers that you chose to quantize have their errors cancel out, like one layer skewed to the right one layer skewed to the left, one layer skewed to the right. Your final logits distribution is more similar to the original distribution of the model, so you have better fidelity. And so the way we proved this was with KL divergence. So instead of just scoring on the benchmarks, we scored the KL divergence between the logit distribution of the quantized model and the logit distribution of the original full precision model, and we showed that with this technique we get. If your probability distribution on the logits which token it wants to select is more of the same as the original model, you're probably gonna end up staying true to the original model. So yeah, so it seems like previously before this, it seemed like the industry was, well, the more you quantize, the worse it's gonna be, ‘cause the more loss you introduce. That's not exactly, not necessarily true. So yeah, doesn't improve it, but can cancel out.Philip [00:31:57]: I think it might be this, but reminds me a good bit about pruning where you can prune off certain layers.Philip [00:32:03]: But very interesting. Didn't know this was a whole paper you guys put out.Ali [00:32:06]: It's. Fun fact, it was originally 72 pages, this paper, and then we decidedPhilip [00:32:11]: WowAli [00:32:11]: We can't tell. We couldn't release it. So it's now 45.Swyx [00:32:15]: Still 39 pages, so very substantive. We talked about evals and all these things and, like what's possible in terms of speedup? Like it's like probably like the numberInference Speedups and BenchmarkingSwyx [00:32:25]: Thing that people do wanna care about, and it's something that you wrote about in your post. Like official API is 70 tokens per second, and you push it up to 90. Is that like a normal thing?Philip [00:32:36]: So what's cool about working in inference, the reason that I think inference is going to be a useful place to do engineering for a long time, is that if you look at highly optimized domains like, say, finance, if you're in finance, you measure how much better you got in basis points. It's like, “Oh, I got five basis points better, like twentieth of 1% better,” that's huge news because everything is so optimized. When we publish optimizations, it's 20%, it's 100% it's 200%. So there's still probably like a lot further to go, honestly. Like you'll, you'll know that inference is pretty much solved when researchers start publishing about how they got 1% faster at something.Swyx [00:33:19]: Which by the way, because I am from the finance background, in the ‘70s, that was the margin at the time. When you did quantitative finance research, you would findAli [00:33:27]: And like 20%, tens of percent.Swyx [00:33:29]: That's. Yes.Philip [00:33:29]: Yeah.Swyx [00:33:30]: And now it'Philip [00:33:31]: Tiny fractionsSwyx [00:33:32]: For those people interested, look up Andrew Lo's paper. He had a really interesting illustration of quant, stat arb, distribution, narrowing down from like those kinds of 20% differences in the ‘70s, down to nothing today, which is very cool.Philip [00:33:48]: Exactly, and we're at the beginning of the same type of thing. Now benchmarking is hard. I think anyone will tell you that, and benchmarking provider speeds is hard because there's so many variables that go into it. What hardware are you using? How much load do you have on the system? What's the exact nature of the prompts and input and output sequence lengths? All that stuff. But overall, when you start stacking these improvements, you're looking at multiples. You can look at it. The most common form, of course, is TPS, tokens per second, which is bad naming by us in the industry, ‘cause there's two tokens per second. There's tokens per second, the throughput number, and the latency number.Ali [00:34:31]: TTMT, yeah.Philip [00:34:32]: Like total tokens per second out of the, out of the GPU as a throughput number. Most people only care about tokens per second as the latency number, which we should call ITL, intertoken latency, but we don't.Philip [00:34:44]: Anyway, so you can imagine a standard API without many optimizations for a 1 trillion parameter model operating somewhere in the 30 to 50 tokens per second range for reasonable traffic profile. And we generally see the goal of, pushing to 10X that. But, not necessarily day zero, but by stacking enough optimizations, if you have, say like four optimizations, each of which doubles performance. Or sorry, three optimizations, each of which doubles performance, then you stack that up, that's an 8X gain. That's the order of magnitude that we're working with in this space. We're trying to make things substantially faster, not just go from like 70 to 90.Swyx [00:35:38]: Are you saying you've. You have done that?Philip [00:35:40]: So let's say you have as a reasonable baseline, 30 or 40 tokens per second. You can achieve 10X that. So like on GLM-5.2, if you run it unquantized, perhaps on H100s even, and you're just using an off-the-shelf inference engine with no particular optimizations, no speculator, nothing extra around like KV routing, no disaggregation, you're, you're probably, yeah, looking at that like 30 to 40. You think that's like a reasonable baseline?Swyx [00:36:12]: Right. Right.Philip [00:36:12]: To get to something like 10X, there's a lot of trade-offs that you're making. If we're running at more like a 300, 400 tokens per second range, you are using the best hardware possible. You have a optimized speculator. You have done all of your quantization work. You are Seeing a pretty high cache hit rate. You are running with a reasonably small batch size and a parallelism configuration that is tuned for latency versus throughput, but it is possible. So the spreads that you see if you, like, go on artificial analysis or you go on OpenRouter and you look at, the worst provider to the best provider, oftentimes can hit that range. 10X is of course very aggressive. It's oftentimes maybe more of a four to six times improvement. But that's the performance that makes us really excited, is when we can get these huge gains, not just go from 70 to 90 tokens.Stacking Optimizations: NVFP4, Speculation, and DisaggregationAli [00:37:19]: It's also, like, hardware dependent. Like, ifPhilip [00:37:20]: YeahAli [00:37:20]: If you have a thing where you're serving it on just, like, a node of H100s and then you throw, like, you shard the model across, like, four nodes of B200s. Like, you can definitely increase the speed with just throwing more hardware at it. Like, normalizing for the same exact hardware and the same number of GPUs.Philip [00:37:35]: Yeah. Then you're looking at, like, a two to 4X improvementAli [00:37:38]: Right. RightPhilip [00:37:38]: Depending on the inference optimizations. So yeah, it's. Some of it's, what's the call, and some of it's who's the driver.Vibhu [00:37:46]: If you break down the two to 4X, say the example is run GLM-5.2Ali [00:37:51]: YeahVibhu [00:37:51]: On B200sAli [00:37:53]: YeahVibhu [00:37:53]: Single node, right? What's, like, the cost trade-off for effort to get, like, the last bit of juice out versus what should people just think of, right?Ali [00:38:01]: Spectre quantization. Yeah.Vibhu [00:38:03]: Spectre quantization.Ali [00:38:04]: That's, that's, that's like 95%. LikeVibhu [00:38:06]: And how far does that get you? And how easy is that for the average person to do? So say right I wanna throw the weights of GLM-5.2 on a node of B200s, how easy is it to find speculative decoder- decoder model or already quantized model? How much work goes into it?Philip [00:38:23]: If you're doing it up front, it's quite a lot of work. If you're doing it today, there's going to be people who have published things that you can just, you can just grab some NVFP4 weights. You can grab a speculator. Yeah, if we're thinking about, like, what are the 2Xs we're stacking, going from, BF16 to NVFP4 is, it's not quite a 2X, right? It's like. I think it's about, like, 30 to 40%, from 16 to 8, and then another 30 to 40% multiplied from, 8 to 4. So that doesn't quite get you a 2X, but, like, roughly a 2X. Speculator, roughly a 2X. Disagg on top of that if you're able to get enough hardware and put enough traffic through it, another roughly a 2X. And then you add in some, double-digit percent increase from having just a better runtime with, the latest kernels and stuff behind it. And that's how it stacks up.Ali [00:39:21]: YeahPhilip [00:39:21]: So building each of those, like, building the, quantized weights is, for someone who really knows what they're doing, hours to days of work. Building the speculator, again, like, hours to days of work. And the, disagg setup, hours to days. Well okay, but like once you haveAli [00:39:39]: Once set up. Once set up. YeahPhilip [00:39:40]: Yeah, getting disagg working for the first time, I'm saying, of course, is very difficult.Philip [00:39:44]: The marginal implementationAli [00:39:48]: Like, if you're just grabbing, like if you are a person, like just a normal consumer who has access to, like, a node of B200s and you're wondering, “How can I just host it myself?” You don't need to quantize the model yourself. There's always gonna be, like, an open source quantized checkpoint. NVIDIA's gonna push one out if no one else does. You. Usually, the providers will have their own spec dec that they've trained as well. You don't need to train your own spec dec. You can just use that as well.Philip [00:40:09]: Yeah. Like, GLM-5.2 has its own MTP.Ali [00:40:13]: Right. Right.Vibhu [00:40:14]: What's multi token prediction?Philip [00:40:15]: Yes.Ali [00:40:16]: I'm justVibhu [00:40:16]: Can you explain that?Ali [00:40:16]: I'm just an expert.Ali [00:40:18]: I can do it for you in case I get it wrong?Vibhu [00:40:20]: No.Vibhu [00:40:21]: Yeah, you should correct if we're wrong, but their multi-token prediction can be used for self-speculative decoding.Ali [00:40:27]: I'm not sure. I'm not gonna correct that.Vibhu [00:40:28]: Okay. I'm semi-confident in thatAli [00:40:30]: Okay. YeahVibhu [00:40:30]: But someone can check. But it's useful to paint the story of, okay, not just the average person, but say a company wants to switch from serverless inference I wanna throw this up on. I wanna rent some GPUs, throw it up. These are the steps you take to do significantly faster than just put it behind vLLM.Ali [00:40:48]: Right.Vibhu [00:40:49]: I was waiting for a mention of Dynamo.Vibhu [00:40:51]: I feel like, that's supposed to be the baseline that you measure against.Dynamo, KV Routing, and Disaggregation ToolkitsPhilip [00:40:55]: I would think of Dynamo as less of a box system and more of a toolkit for building with. So when we talk about doing aware routing, when we talk about doing KV offloading, when we talk about doing, PD disaggregation, Dynamo fundamentally is. By the way, Dynamo is an open source library from NVIDIA.Ali [00:41:17]: We've done a pod with KylePhilip [00:41:18]: OkayAli [00:41:19]: Kyle Cranin.Philip [00:41:19]: Cool. So then your listeners know then that it supports all the different inference frameworks. And it is multi hardware, which is interesting.Ali [00:41:28]: But it's just a router, it's not like an optimizer layer.Philip [00:41:30]: Yeah. All it does, like, what Dynamo is good at, it is a library for moving information around your cluster, around your hardware. So if you have, KV cache on one place and you need it to be somewhere else, Dynamo coordinates NIXL for you to move that around.Philip [00:41:49]: That doesn't mean that, like, out of the box, you just say, “Pip install Dynamo,” and then you get, like, a massive performance speed up. It's more of a developer toolkit.Ali [00:42:01]: Yeah. I would have said it would. It comes with a set of defaults that you can then swap out.Philip [00:42:06]: It does. If the industry at large, I think, was, like, rolling out all of these deployments, standard, then I think it would be, like, a credible baseline. But, we've got to, we've got to benchmark against, like, what we're seeing in the wild.Speculative Decoding Methods: Medusa, EAGLE, n-Gram, and Spec-SpecVibhu [00:42:23]: I did wanna talk a little bit more about PD disagg, because that is probably, like, number three after quantized and speculative decoding. In your book though, I was just gonna pull out the book.Philip [00:42:31]: Yeah.Vibhu [00:42:32]: Like section 522 on Medusa, 523 on EAGLEPhilip [00:42:35]: YeahVibhu [00:42:36]: 524 on gram.Philip [00:42:37]: It's 55, would be disaggregationAli [00:42:42]: Yeah. Well, no, I just wanted to dwell a little bitPhilip [00:42:44]: YeahAli [00:42:44]: The other. Like, so what do you choose to include? What do you choose to not to include? Because there was all these other techniques.Philip [00:42:51]: Yeah.Ali [00:42:51]: Are these still relevant? Because I think they came out, like, a year and a half ago maybe.Vibhu [00:42:55]: Medusa is quite old.Philip [00:42:56]: Yeah, Medusa's old.Ali [00:42:58]: It was old.Vibhu [00:42:58]: But is it in the book as a good, here'sPhilip [00:43:01]: BaselineVibhu [00:43:01]: Baseline vanilla understand it?Philip [00:43:02]: Like you should know this.Vibhu [00:43:03]: Like I read the paper, I'm like, “ it makes so much sense.”Philip [00:43:05]: Yeah.Philip [00:43:05]: So with the book, I had a couple goals. One was to give people just a working vocabulary for the space as a whole, and the other was to give them some intuition about how each of these techniques works. As I mentioned in my AI Engineer talk, which is the first public addendum to this, the speculation space has moved much faster than everything else. So yeah, even at the time that I wrote the book Medusa, I very much included as a way for people to understand how the space evolved rather than what the most modern technique is. And now of course, there's DFlash, dSpark. There's, there's newer techniques even than EAGLE, although EAGLE is still very commonly used.Ali [00:43:51]: SpecSpecta.Philip [00:43:52]: Yes. Speculative decoding.Vibhu [00:43:54]: What canAli [00:43:56]: Oh, it's a paper by Tri Dao and it's like, it's doing speculative decodingVibhu [00:44:00]: HuhAli [00:44:01]: For the speculative decoder.Philip [00:44:02]: Oh, in spec- oh my God.Ali [00:44:02]: It's literally just an another. It's like, yeah, that's the most simple way to explain it, and it seems like he got trivial speed ups there. But it seems that the complexity with training, it's almost like in our mind at least, it's almost as complex as training GANs. Like it's like a very delicate balance and oftentimes you, it's just but yeah, it's literally speculative decoding on speculative decoding.Vibhu [00:44:21]: Speculative.Ali [00:44:22]: Yeah. We saw this paper.Vibhu [00:44:24]: It's interesting, right?Ali [00:44:24]: Yeah.Vibhu [00:44:24]: I wouldn't even expect it to be very particular to train, I wouldAli [00:44:29]: Right.Vibhu [00:44:29]: The naive part of me is like, okay, train speculative decoder.Ali [00:44:32]: But like, and it makes sense, like the whole idea of speculative decoding is you. It's like, it's like almost like the iPhone auto predict version but for a normal model, right? Like you're just, you're just, generating three tokens and you're like, okay, I'll do prefill on them. And so you save those three turns for your original model. Now your speculative decoder is doing three turns of auto regression, so why not just have an even smaller model?Ali [00:44:53]: The other question there is what are the size of speculators? So say forPhilip [00:44:58]: Right. It's like a billion parameters.Ali [00:45:01]: Like for MiniMax, it's. Yeah. It's like one layer. It's like one 60th of the original model usually.Philip [00:45:06]: Yeah. I think we should do a paper when we get back to the office.Philip [00:45:10]: SpeculativeAli [00:45:11]: SpeculativePhilip [00:45:11]: Decoding.Ali [00:45:13]: No, it's, it does seem like how, when do you stop? But then it also seems like if you're able to train spec-spec decode for instance, right? Like if you're able to have a small model that is accurately predicts what the intermediate speculator is gonna predict, that is able to predict what the original target model's gonna predict, then why not just use that smallest model directly, right?Vibhu [00:45:34]: Yeah. This isAli [00:45:35]: Like it seems likeVibhu [00:45:35]: Adjacent to the routing problem.Ali [00:45:36]: Right.Vibhu [00:45:36]: Yeah.Ali [00:45:36]: Right.Philip [00:45:37]: The thing with speculators is one of the practical constraints on using them is that you do have to run a small model on the same hardware that you're running the big model on. There is a orchestration and resource competition problem inherent in that, and that is one of the constraints on speculation in general, is that draft tokens cost resources to create and cost software complexity to manage. And so if you have like infinitely recursive speculators, you add in quite a bit of that complexity on the actual implementation within the inference engine as well, not just in the training process.Vibhu [00:46:17]: I was gonna say, I would wonder if you could do similar, like distillation and pruning of, it's the same thing, it's just a model. Can we not just distill a lot of the weights, quantize the speculator, out of my domain? The question that also comes up is, this is all for big server workloads, right? How much of this applies to, say I have this MacBook, I wanna run Gemma really efficiently. Similar problems, not the same?Local AI vs. Data Center InferencePhilip [00:46:45]: Pretty different. I talked to Selo, about this on his podcast a couple weeks ago. The difference between inference engineering for the data center and for production workloads versus inference engineering for local AI, is that we start with fundamentally like different constraints and different goals. With local AI, it's how do I fit this model onto my hardware and then make it less dumb? And with data center influence, it's how do I load this model and then make it less slow? And we care about less dumb, and they care about less slow. But the local AI inference engineering ecosystem, I think has a lot for us to learn from in the data center space. They are experts in various forms of quantization, including dynamic quantization that we just don't touch, in the pruning, in the distillation, in the, layer removal. There'Ali [00:47:42]: Layer removal matters less.Philip [00:47:43]: Yeah. There'Ali [00:47:44]: No one loves pruning really.Philip [00:47:45]: Yeah. Well, but the, but they doVibhu [00:47:46]: Which is surprising, right? But that's, that's a whole different thingPhilip [00:47:48]: Just to fit something on the laptop.Ali [00:47:50]: Right.Philip [00:47:50]: So yeah, it's a, it's an interesting, it's an interesting space. Not necessarily that like their techniques make sense for us to do in the data center, because we have different resources and different goals, but more that the process as well as the openness of that field is something to, admire.Ali [00:48:12]: Yeah. Like to your point, like, certain optimizations that would. Like for instance, Turbo Quantum Sharper, like it made such huge hype on that and we did like a whole deep dive on Twitter and like said, what is it? How does it work? Why is it good or not? And it took off and it was implemented on local devices because your memory bandwidth is so slow on like a MacBook, for instance. But try putting the same thing on like an NVIDIA GPU on a B200 Turbo quant would not be. Like, it would not be used. Like, NVIDIA - Like, NVIDIA made it clear that this is not a good optimization, and we've seen it firsthand where the overhead of doing dequantization, quantization of, in the kernel itself with turbo quant kernel, each end is much slower than the time that you save from doing the bandwidth. ‘Cause on the B200s, you have like 3.5 terabytes per second. You don't need decrease the storage that much. You don't need to do, FP4 KV cache. You don't need to use a requant. There's, there's, there's better optimizations to be made. But on Edge devices, it's extremely important, it's extremely useful. So, seems to be, like, different optimizations there, but then they're all uniquely combined with like all you wanna quantize the model, you wanna do speculative decoding, like certain common prefixes with bothPhilip [00:49:18]: Principles.Ali [00:49:19]: Yeah, exactly. Exactly. Exactly.Philip [00:49:20]: They also do a lot of work on, model parallelism, especially over, heterogeneous topology, where you have, some sparks and they are wired together with, Ethernet, DGX sparks.Ali [00:49:35]: Yeah, this is the Exo Labs guys.Philip [00:49:36]: Yeah. You have, a nu
This episode explores how subtle moments of acceptance shape our sense of belonging, especially for children and adults alike. Discover why creating spaces where others feel safe to be themselves is a gift that can transform lives with simplicity and kindness.Follow me on instagram for more inspiring, motivational and uplifting content. See you there❤️
"Ich seh nicht mehr so aus, aber im Herzen bin ich Punk" - Marcus Smaller von 3 Feet Smaller – P4Cast 153
The 365 Days of Astronomy, the daily podcast of the International Year of Astronomy 2009
From February 10, 2021. Hosted by Dr. Pamela L. Gay. Scientists collected fresh data on Orion's bright star Betelgeuse to try and understand this star that caused so much controversy. They found it's smaller than previously calculated, and last year's dimming was likely caused by dust, but it's also more complex than thought. Plus, galaxies, Earth's forests, ice on Mars, and Saturn's moon Rhea. We've added a new way to donate to 365 Days of Astronomy to support editing, hosting, and production costs. Just visit: https://www.patreon.com/365DaysOfAstronomy and donate as much as you can! Share the podcast with your friends and send the Patreon link to them too! Every bit helps! Thank you! ------------------------------------ Do go visit http://www.redbubble.com/people/CosmoQuestX/shop for cool Astronomy Cast and CosmoQuest t-shirts, coffee mugs and other awesomeness! http://cosmoquest.org/Donate This show is made possible through your donations. Thank you! (Haven't donated? It's not too late! Just click!) ------------------------------------ The 365 Days of Astronomy Podcast is produced by the Planetary Science Institute. http://www.psi.edu Visit us on the web at 365DaysOfAstronomy.org or email us at info@365DaysOfAstronomy.org.
Finishing up the listener Q+A questions.Chapters00:00 Client Shoutouts and Introductions01:27 Core Training Insights: Vacuums and Rib Pain08:34 Fasted Cardio vs. Weight Training: Understanding the Differences15:42 Breaking Free from Tracking Macros: Maintaining Your Physique19:33 Choosing the Right Coach: Finding Your Fit22:35 Nutrition and Activity Management Around Surgery26:07 Strength vs. Hypertrophy Training: Key Differences30:04 Addressing Disordered Eating and Compulsive Behaviors31:29 Correcting Muscle Imbalances: Single Leg Training Strategies38:41 Reverse Dieting Update: Progress and InsightsLinksApply for Coaching: https://form.typeform.com/to/ubUfJiEu?utm_source=podcastLiving Lean Podcast: https://www.buzzsprout.com/712032Follow Jeremiah on Instagram: https://www.instagram.com/jeremiahbair/Follow Andrea on Instagram: https://www.instagram.com/andirogersfit/Follow Natalie on Instagram: https://www.instagram.com/natalieatswell/Keywordsfitness, core training, fasted cardio, nutrition, coaching, surgery recovery, strength training, hypertrophy, muscle imbalancesTo Apply For Coaching With Our Team: CLICK HERE
Psalm 8 The post The Higher We Look the Smaller We Feel appeared first on Pillar Baptist Church.
In this episode of The Pharmacist's Voice Podcast, I talk with Ohio pharmacist Dan Krinsky about writing, co-authoring, and editing. This is Episode 22 in the Pharmacist Authors Series. Unlike most episodes in the series, our conversation does not focus on one specific book. Dan has written and edited many types of professional content, including continuing education and journal articles, newsletters, patient education materials, business proposals, and pharmacy reference resources. Dan first appeared on the podcast in Episode 133, when we discussed his pharmacy career and pharmacogenomics. This time, writing gets the spotlight. Dan shares practical advice for pharmacists, pharmacy students, pharmacy technicians, and pharmacy professors who want to begin writing or expand their writing experience. Dan's contact information, website, and social media links Dan Krinsky BS, MS, RPh, FAPhA on LinkedIn https://www.linkedin.com/in/dan-krinsky-bs-ms-rph-fapha-a8537a/ Dan's email dkrinsky@educare-4u.com EduCare4U, LLC website https://www.educare-4u.com PGx101, LLC on LinkedIn https://www.linkedin.com/company/pgx101 The Pharmacist's Voice Podcast Episode 133 featuring Dan Krinsky About Dan Krinsky B.S., M.S., R.Ph., FAPhA (July 2026) Dan is business owner, educator, entrepreneur and pharmacist who is passionate about advancing patient care and the pharmacy profession. He owns 2 businesses: EduCare4U, LLC, (focused on education and patient care) and PGx101, LLC (focused on pharmacogenomics education and practice implementation). Mr. Krinsky is currently a Professor at the LECOM School of Pharmacy, serving as a lecturer, committee member, and student advisor. He is a Fellow of the American Pharmacists Association and an active member of the Medication Management Special Interest Group. Other ventures include working with providers to optimize patients' drug therapy; professional writing; creation of certificate training programs, and new business development. Mr. Krinsky is Editor-in-Chief and Section Editor for the American Pharmacists Association publication 'Handbook of Nonprescription Drugs: An Interactive Approach to Self-Care.' Mr. Krinsky's areas of expertise include community pharmacy practice, drug information, patient counseling and education, OTCs, natural products, pharmacogenomics, and developing and implementing medication therapy and disease state management programs. His primary professional goal is to 'pay it forward', finding ways to make the profession better for the next generation of pharmacists. Questions and abbreviated answers from our conversation How did writing become part of your professional life? Dan's first major writing project was his master's thesis in pharmacokinetics. Writing did not come naturally at first, but he discovered that he enjoyed gathering information and turning it into something useful for an audience. How can someone decide whether writing is right for them? Start with a topic you care about. Identify the audience, look for gaps in available information, and find a mentor or support network. You do not need to work alone. Should new writers start small or go directly to writing a book? Dan recommends beginning with an article, newsletter, or another smaller project when possible. Smaller projects provide quicker feedback and help writers build skills and confidence. Someone who wants to start with a book should narrow the topic, identify the audience, create an outline, set deadlines, and seek feedback early. What has Dan written? His work includes newsletters, business proposals, patient advisory leaflets, consumer health information, journal and continuing education articles, and two co-authored point-of-care resources: the Natural Therapeutics Pocket Guide and the Drug-Induced Nutrient Depletion Handbook. He has also contributed to multiple editions of APhA's Handbook of Nonprescription Drugs as an author and editor. How are writing, co-authoring, and editing different? An author creates the content. Co-authors share responsibility and need clear roles, flexibility, and communication. An editor improves accuracy, organization, clarity, and usefulness while protecting the author's voice. When should an author involve an editor? As early as possible. An editor can help shape the audience, scope, outline, organization, and flow before the author invests months in the wrong direction. Early feedback is usually easier to handle than dozens of changes at the end. How can someone find a useful topic? Pay attention to recurring patient questions, conversations, media claims, and gaps in health information. The final product does not need to be a book. It could be an article, handout, script, social media post, or educational program. What keeps people from starting? Dan identifies three common barriers: waiting to become "the expert," fear of criticism, and lack of structure. Know what you know, recognize what you do not know, and build a network that can help fill the gaps. How can pharmacists develop a writing habit? Schedule writing time, set specific goals, work backward from deadlines, and use an accountability partner. A detailed outline makes a large project feel more manageable. Where can beginners find legitimate writing opportunities? Consider professional associations, colleges of pharmacy, health systems, community pharmacies, continuing education companies, consumer health websites, and digital health companies. Build a portfolio with writing samples, areas of expertise, testimonials, and other communication experience. How long should someone give writing before deciding it is not for them? There is no single timeline. A newsletter article may take a month; a book may take 12 to 18 months. Set realistic deadlines for each project, start small, and give yourself time to learn. What is Dan's final advice for aspiring authors? Find a mentor, be patient, expect revisions, learn from feedback, and enjoy the process. Writing may not be the right path for everyone, but it can be a meaningful way to educate patients, serve the profession, and create new opportunities. Pharmacist Authors Series Episode List 2025 Pharmacist Authors Series Part 1 of 3 (June) Episode 336 with Jerry Levin, PhD, LMT - Affirm Yourself: 15 Principles to Retrain the Voice in Your Head. Part 2 of 3 (July) Episode 340 with Ashley Walker, PharmD - Expanding Your Brilliance: Creating Effortless Abundance while Navigating Business and Motherhood https://amzn.to/418hdwx Part 3 of 3 (October) Episode 353 with Sandra O. Onye, PharmD - Winning Is in My DNA: 15 Minutes of Self-Reflection https://amzn.to/48KTtn3 2024 Pharmacist Authors Series Part 1 of 3 (June) Episode 282 with Steve Leuck, PharmD - A Pharmacist's Story An Authentic Tale of True Love, Family, Addiction, and the Practice of Pharmacy Part 2 of 3 (July) Episode 288 Helen Sairany, PharmD - The We You Don't See: Understanding the Long Shadows of Trauma Part 3 of 3 (August) Episode 293 Kim Newlove with Publishing in Doses Co-Founders Janan Sarwar, PharmD and Theary Chhim, PharmD plus audio engineer, Julie Walthers. = 2023 Pharmacist Authors Series (Part 1 of 15) June 5, Introduction to the Pharmacist Authors Series (Episode 220) (Part 2 of 15) June 9, The Pharmacist's Voice ® Podcast Episode 221: Interview with Salam Kabbani, PharmD about her book: COVID Long-Hauler: My Life Since COVID (Part 3 of 15) June 12, The Pharmacist's Voice ® Podcast Episode 222: Interview with audio engineer Julie Walthers from Whole Story Studio: https://www.wholestorystudio.com/ (Part 4 of 15) June 16, The Pharmacist's Voice ® Podcast Episode 223 Interview with Erin L. Albert, PharmD on her book The Life Science Lawyer Part 5 of 15) June 19, The Pharmacist's Voice ® Podcast Episode 224 Interview with Sue Ojageer, PharmD on her children's book The Pharma Heroes: The Power of Precision Medicine (Part 6 of 15) June 23, The Pharmacist's Voice ® Podcast Episode 225: Interview with Tony Guerra, PharmD about his Pharmacist Residency and Career Series (8 books) (Part 7 of 15) June 26, The Pharmacist's Voice ® Podcast Episode 226: Interview with Christina Fontana, PharmD about her book Moving Beyond the Counter: Elevating into Heart-Centered Health Care through Entrepreneurship (Part 8 of 15) June 30, The Pharmacist's Voice ® Podcast Episode 227: Interview with Jade L. Ranger, PharmD, about her book Mustard Seed Mentality: Unscripted Pearls of Wisdom from a Wife, Mother, and Entrepreneur (Part 9 of 15) July 7, The Pharmacist's Voice ® Podcast Episode 229: Interview with RDML Pam Schweitzer, PharmD and her daughter Amy Graves about their children's book Alice and Jack Hike the Grand Canyon (Part 10 of 15) July 10, The Pharmacist's Voice ® Podcast Episode 230: Interview with Cory Jenks, PharmD about his book Permission to Care: Building a Healthcare Culture That Thrives in Chaos (Part 11 of 15) July 14, The Pharmacist's Voice ® Podcast Episode 231: Interview with Donna Bartlett, PharmD about her book MedStrong: Shed Your Meds for a Better, Healthier You - Aging Well Through Deprescribing (Part 12 of 15) July 17, The Pharmacist's Voice ® Podcast Episode 232: Interview with Frieda Wiley, PharmD about her book Telecommuting Psychosis: From Surviving to Thriving While Working in Your Pajama Pants. Plus, we touch on her 3 children's books in development. (Part 13 of 15) July 21, The Pharmacist's Voice ® Podcast Episode 233: Interview with Tim Ulbrich, PharmD about his book Seven Figure Pharmacist: How to Maximize Your Income, Eliminate Debt, and Create Wealth (Part 14 of 15) July 24, The Pharmacist's Voice ® Podcast Episode 234: Interview with LaQuoia Johnson, PharmD about her book How Rxacism Manifests Inside the Small World of Pharmacy (Part 15 of 15) July 28, The Pharmacist's Voice ® Podcast Episode 235: Pharmacist Authors Series wrap-up (solo show) Kim's websites and social media links ✅ Monthly email newsletter sign-up link https://bit.ly/3AHJIaF ✅ LinkedIn Newsletter link https://bit.ly/40VmV5B ✅ Business website https://www.thepharmacistsvoice.com ✅ Buy my book on amazon.com https://amzn.to/4iAKNBs ✅ The Pharmacist's Voice ® Podcast https://www.thepharmacistsvoice.com/podcast ✅ Drug pronunciation course https://www.kimnewlove.com ✅ A Behind-the-scenes look at The Pharmacist's Voice ® Podcast © Online Course https://www.kimnewlove.com ✅ LinkedIn https://www.linkedin.com/in/kimnewlove ✅ Facebook https://www.facebook.com/kim.newlove.96 ✅ Twitter https://twitter.com/KimNewloveVO ✅ Instagram https://www.instagram.com/kimnewlovevo/ ✅ YouTube https://www.youtube.com/channel/UCA3UyhNBi9CCqIMP8t1wRZQ ✅ ACX (Audiobook Narrator Profile) https://www.acx.com/narrator?p=A10FSORRTANJ4Z ✅ Start a podcast with my coach, Dave Jackson from The School of Podcasting! Click my affiliate link: https://community.schoolofpodcasting.com/invitation?code=G43D3G *New 12-4-25* Thank you for listening to episode 372 of The Pharmacist's Voice ® Podcast. If you know someone who would like this episode, please share it with them!
innercityleft.com Support us at patreon.com/innercityleft Follow us on IG @InnerCityLeft Merch Store is open ICL Store Purchase Amoja's book of poetry here
Tall women are less likely to fit into the traditional image of femininity. It's common for us to be called masculine because of our height and naturally larger features, since being tall and bigger are things associated with masculinity. And then on top of that, it's a bit of a struggle to physically present ourselves in a feminine way, meaning when clothes like dresses, shoes, or even skirts aren't available in our size or length. But that's a topic for a different day. But a common thing that we hear in the tall girly community, whether or not they like their height, is that they want to be or feel smaller. Some outright say that they wish that they were shorter. Others say that they want to feel small and petite and protected next to their male partners. Either way, it all ties back to wanting to feel or be smaller. In today's episode, we're talking about the obsession with being smaller, how it ties into femininity, and why some tall women desire to be smaller when really, they desire to feel feminine. Tune in for more! Buy Me A Coffee: https://buymeacoffee.com/atallgirlspodcast Subscribe to A Tall Girl's Newsletter: https://atallgirlspodcast.beehiiv.com/subscribeLet's stay connected: https://beacons.ai/atallgirlspodcast Leave a review and let me know how tall you are: https://atallgirlspodcast.com/reviews
Today we're doing a throwback episode to one of our favorites from the early days of Stories Podcast. Smaller Than Ever! Rebekah lives in a house full of sisters and husbands and children and babies and the only place she can go for some peace and quiet is the temple. When she asks the Rabbi what to do about her tiny house, he gives her some really strange advice... Check out Stories RPG our new show where we play games like Starsworn with all your Max Goodname friends, and Gigacity Guardians featuring the brilliant firefly! https://link.chtbl.com/gigacity Draw us a picture of what you think any of the characters in this story look like, and then tag us in it on instagram @storiespodcast! We'd love to see your artwork and share it on our feed!! If you would like to support Stories Podcast, you can subscribe and give us a five star review on iTunes, check out our merch at storiespodcast.com/shop, follow us on Instagram @storiespodcast, or just tell your friends about us! Check out our new YouTube channel at youtube.com/storiespodcast. If you've ever wanted to read along with our stories, now you can! These read-along versions of our stories are great for early readers trying to improve their skills or even adults learning English for the first time. Check it out.
Send us Fan MailI pride myself on being a fat person who isn't afraid to take up space. But the more you take up space, the more people will take issue with it. Especially if you're visibly different — Fat, Trans, Queer, Neurodivergent. So I suffered a few setbacks over the past few years and I learned that if I don't censor or filter myself, chances are I would be punished and my words would be used against me. I learned there is a "right way" to do social justice, and if that means hiding parts of myself, then it's worth it for the greater good. Turns out that faking it doesn't work out much better, and liberation can only come when people are free to be themselves. Sounds pretty obvious when you say it out loud, but this wasn't an easy lesson for me to learn. Join me this week as I start taking up space my way, in spite of the potential future backlash. Don't forget that My Joints Hurt: A Practical Guide to Navigating Healthcare in a Fat Body is available to purchase on Amazon or read on Kindle Unlimited. Got a question for the next podcast? Let me know!Connect With MeBUY THE BOOK: Never suffer through another through another weight loss lecture from your doctor againJOIN THE NO WEIGH MOVEMENT: Get a free script when you sign upTHE WEIGHTING ROOM: Community with a neurodivergent flavour. **BOOK CLUB** exclusive to Weighting Room members. BOOK A CONSULTATION: For the ultimate transformation in your healthcare journeyEXPLORE THE MASTERCLASS LIBRARY: Become an expert in your condition and the weight inclusive ways to manage itFREE GUIDES:Evidence-based, not diet nonsenseFind me on Instagram, YouTube, and LinkedIn.
Machine learning teams are moving faster, but the hard part has not disappeared. The work is shifting from writing and debugging every line of code toward defining the right problem, setting requirements, reviewing outputs, and deciding what belongs in a durable platform.Niels Bantilan, Chief Machine Learning Engineer at Union AI, explains how machine learning work has changed, why coding agents are accelerating prototyping, and what engineers must consider when building infrastructure that supports many teams instead of optimizing one model. He also shares how customer needs become product decisions, why machine learning roles are becoming more specialized, and why measuring AI productivity remains difficult.Key Takeaways• Coding agents reduce time spent on implementation, debugging, and exploration, but engineers still need judgment around architecture, quality, and business value.• Platform teams must balance experimentation with stability by giving users freedom at the edges while protecting a reliable foundation.• Machine learning engineering now spans a wider range of skills, from low level performance work to customer empathy, education, documentation, and developer advocacy.• The best model for a task may depend on complexity. Smaller self hosted models can handle tightly scoped changes, while longer and more complex work may still require stronger hosted tools.Episode Highlights00:50 What Union AI means by an AI runtime for production02:10 How machine learning work has changed over the past five years10:40 The mindset shift from model building to platform engineering15:00 Turning customer problems into reusable product capabilities19:00 Why machine learning roles are becoming more specialized21:50 Using coding agents through specifications, tickets, and code review26:50 Token costs, productivity measurement, and choosing the right modelOne Line That Stuck“I'm still solving problems. It's just the level at which I'm doing it doesn't require me to necessarily get into the weeds of the implementation.”Follow The Tech Trek for more conversations on AI, data, engineering, product, and technical leadership.
Episode Summary Michelle opens with the question a lot of solo PR pros have been quietly asking themselves: if I call myself a communications generalist on my website right now, am I costing myself money? Karen's answer is immediate — not 'am I?' but 'you already are.' What follows is a data-driven, practically grounded conversation about the specialization economy: the growing body of research across the freelance and independent consulting world in 2026 that shows generalists getting squeezed and specialists pulling away. Karen and Michelle aren't just reporting a trend — they're translating it specifically for PR and communications practitioners who've never had anyone apply this research to their work. The episode covers the bimodal income distribution hiding inside freelance averages, the vertical-horizontal framework for finding your niche, four common objections to specializing (with honest answers to each), a three-question filter for identifying your niche, and the metric-capturing habit that makes specialization pay off over time. This is a conversation for the solo practitioner who has 'I do everything' on their website tonight — and might be ready to change it. Episode Highlights [00:03] The Opening Question That Frames Everything: Michelle opens before the intro music with a direct question to Karen: if she calls herself a communications generalist on her website right now, is she costing herself money? Karen's answer: not 'am I?' but 'you already are.' The episode's premise is immediate and personal — and Karen and Michelle make clear they're talking to themselves too. [01:05] The Specialization Economy: What the Data Shows: Freelance and independent consulting data in 2026 is pointing in the same direction across multiple sources: generalists are getting squeezed and specialists are pulling away. The average US freelancer earning rate hides what Karen calls 'a canyon' — generalist content and writing on the low end, specialists in high-demand niches billing well over $100 an hour on the high end. Almost nobody is actually earning the average. The floor is dropping for generalized skills; the ceiling is rising for specialized ones. Karen's framing: the middle — 'I'm pretty good at a lot of things' — is where people get stuck. Note: some figures referenced in this episode are still working through the show's verification process; sourcing details will be linked in the resources section as they are confirmed. [04:13] Why Specialization Wins: The Practical Case, Not the Philosophical One: The argument for specializing isn't philosophical — it's structural. A generalist PR consultant competes with an enormous pool of other generalist PR consultants. Someone who specifically handles crisis communications for mid-size healthcare systems competes with a much smaller, more identifiable group. Smaller pool, higher rates, and — critically — the client doesn't have to explain their industry from scratch. That last point is underrated: starting a client engagement already fluent in their world, their vocabulary, and their stakeholders is worth real money. Karen also flags a related shift: companies are increasingly requiring proof of impact before hiring specialists, not just portfolios. That proof is much easier to produce when you've done the same kind of work for the same kind of client repeatedly. [06:40] The Vertical-Horizontal Framework: What Niching Actually Means in Practice: Karen and Michelle push back on the idea that niching just means picking an industry. The framework showing up across freelance research: pick a vertical (the industry — healthcare, legal, fintech, sustainability, professional services) and a horizontal (the service — media relations, crisis management, thought leadership, internal comms, funding round communications). Your niche is the intersection. Examples drawn from recent guests: Sharon Toerek does IP and marketing law for independent agencies. Kara Ryan came up through healthcare communications and built an advisor-led, AI-powered practice on top of that. Both dialed in the vertical and the horizontal. The practical test: once you say your niche out loud, it should stop sounding like a limitation and start sounding like a positioning statement. [09:06] The Filtering Benefit Nobody Talks About Enough: When you're specific, the wrong-fit inquiries mostly stop coming in. You stop getting the 'can you also just quickly help with our internal newsletter' request from an industry you don't want to be in. Positioning does some of your qualifying for you before the discovery call even happens. Karen and Michelle note this is deeply connected to scope creep — a topic worth its own episode. [10:02] Specialization Is an Income Stability Conversation, Not Just a Rate Conversation: Once you're known for a specific thing, you stop pitching one-off projects and start getting asked to stay. Broader freelance data shows a large majority of hiring managers plan to lean more on freelance and fractional talent for ongoing work — and that shift toward retainers happens specifically because specialists make ongoing relationships easy to justify. Nobody keeps a generalist on retainer. The FinTech thought leadership expert stays. The practical consequence: niching is often what makes the retainer conversation possible in the first place, and retainers solve the feast-or-famine cycle that most solo practitioners experience. [12:41] Four Objections — With Honest Answers: Karen and Michelle work through the four most common pushbacks they hear in the Solo PR Pro community. One: I'll turn away good work and go broke. Honest answer — there is a real ramp-up period of roughly six to twelve months; don't torch your existing client base overnight, shift new business conversations while honoring existing relationships. Two: my market is too small. Counter with math — a few hundred mid-sized healthcare systems in the country, you only need a handful of retained clients for a full solo practice; smaller pool of competitors is not the same as a smaller pool of clients. Three: I'll get bored. Flips the other way — as a generalist, every new client is a cold start; as a specialist, the energy goes into strategy instead of orientation. Four (the quiet one): what if I pick the wrong niche? [16:16] The Three-Question Filter (Plus an Unofficial Fourth): A practical framework for identifying your niche. Question 1: Where do you already have an unfair advantage? Past industry experience, a network, credentials, lived experience — something that means you start ahead of a stranger walking in cold. Question 2: Where's the budget? You can be brilliant in a niche that simply doesn't spend on PR. Healthcare, legal, financial services, B2B tech consistently show up as categories with real comms budgets. Question 3: Can you say it in one sentence — and does that sentence make a stranger say 'I know exactly who needs you'? If it requires three qualifying clauses, it's not sharp enough yet. The unofficial fourth: are you willing to hold the line publicly? Your website, your LinkedIn, your pitch all need to stay consistent or the positioning won't do its work. [18:11] Running the Filter Live: The In-House Bank Example: Karen and Michelle run a hypothetical listener through the filter: six years in-house at a regional bank, now doing a bit of everything for small business clients. Unfair advantage: already fluent in financial services vocabulary, compliance, and regulatory relationships — most PR consultants would need a year to learn that. Budget: financial services and fintech are consistently well-funded for communications. One sentence: 'I help community banks and credit unions navigate media and regulatory communications.' Karen: say that at a conference and watch how fast someone says 'I know someone who needs that.' [20:28] How to Prove It's Working: The Metric-Capturing Habit: For every engagement going forward in your niche, capture one number — one sentence, one metric, one outcome. Not a dramatic case study, just: 'Positioned the founder as a category expert; three inbound press inquiries within a month of the first byline running.' Build a running document, whether in Notion, a notes app, or wherever your system lives. Michelle: build it into your closeout process for every wrapped engagement, and into every campaign, not just every full client relationship. Karen: future you will be very grateful. Related Episodes That Solo Life, Episode 343: Sharon Toerek on Legal Protection, IP, and Building a Specialized Practice That Solo Life, Episode 341: Kara Ryan on Going Solo After 20 Years in Healthcare Comms Resources & Additional Information Doers Circle: The Future of Freelancing in 2026: 8 Trends Solopreneurs Can't Ignore Venture Lab: 10 Freelancing Trends in 2026 (Rates, Niches, and Client Expectations) Solo PR Pro membership community: soloprpro.com That Solo Life podcast website: thatsololife.com Host & Show Info That Solo Life is a podcast created for public relations, communication, and marketing professionals who work as independent and small practitioners. Hosted by Karen Swim, APR, President of Solo PR Pro, and Michelle Kane, Principal of Voice Matters, the show delivers expert insights, encouragement, and practical advice for solo PR pros navigating today's dynamic professional landscape. Listen to all episodes and catch up on previous conversations at thatsololife.com. Did this episode inspire you? If you found value in this conversation, please take a moment to leave us a review on your favorite podcast platform. Your feedback helps us reach more solo pros just like you! Don't forget to subscribe so you never miss an episode.
Listen and subscribe to Money Making Conversations on iHeartRadio, Apple Podcasts, Spotify, www.moneymakingconversations.com/subscribe/ or wherever you listen to podcasts. New Money Making Conversations episodes drop daily. I want to alert you, so you don’t miss out on expert analysis and insider perspectives from my guests who provide tips that can help you uplift the community, improve your financial planning, motivation, or advice on how to be a successful entrepreneur. Keep winning! Two-time Emmy and Three-time NAACP Image Award-winning, television Executive Producer Rushion McDonald, interviewed Jocelyn Pearson. Purpose of the Interview The interview on Money Making Conversations Masterclass with Rushion McDonald and Jocelyn Pearson aimed to: Share Jocelyn’s journey of graduating debt-free by securing $126,350 in scholarships. Educate families on how to avoid student loan debt through her proven system, The Scholarship System. Dispel myths about scholarships and provide actionable steps for parents and students. Key Takeaways Scholarship System Approach Jocelyn developed a six-step process to simplify scholarship applications and avoid overwhelm. Focus on breaking the process into small, manageable steps rather than a vague “go get money” directive. Common Myths Debunked Too early or too late to apply: Start by junior year; it’s never too late—even college seniors can apply. Only perfect students or low-income families qualify: Many scholarships don’t require high GPA or athletic ability. All good scholarships are gone: Smaller, local scholarships ($500–$5,000) add up over time. It takes too much time: With a system and reusable materials, effort decreases each year. Avoiding Scholarship Scams Beware of “easy,” “enter to win,” or sweepstakes-style scholarships—they often sell personal data. Real scholarships require effort and personalization. Role of Parents Parents should help with planning and identifying legitimate scholarships but not complete applications for students. Committees can detect when parents write essays. AI in Scholarship Applications Jocelyn warns against copy-pasting AI-generated essays. Her platform introduced TESS, an AI assistant for ethical guidance and support. Financial Aid Basics Submit FAFSA even if you think you won’t qualify; some colleges and states require it. Combine all sources—government aid, institutional aid, and private scholarships. For Current College Students Check with financial aid offices, academic departments, and organizations for scholarships available after freshman year. Entrepreneurial Journey Jocelyn turned her passion into a business by starting with a book, building an email list, and launching webinars. She emphasizes persistence and ignoring naysayers. Notable Quotes “I had to accumulate my way to getting college paid for—the mere mortals’ way to going to college without tons of debt.” “Most families want scholarships, but they get stuck in the overwhelm.” “There’s no big red easy button—but with clear steps, it feels less daunting.” “We’re saying no to the broken system… It takes, on average, 21 years to pay off student loans.” “With great power comes great responsibility—AI can help, but only if used ethically.” #SHMS #STRAW #BESTSupport the show: https://www.steveharveyfm.com/See omnystudio.com/listener for privacy information.
Listen and subscribe to Money Making Conversations on iHeartRadio, Apple Podcasts, Spotify, www.moneymakingconversations.com/subscribe/ or wherever you listen to podcasts. New Money Making Conversations episodes drop daily. I want to alert you, so you don’t miss out on expert analysis and insider perspectives from my guests who provide tips that can help you uplift the community, improve your financial planning, motivation, or advice on how to be a successful entrepreneur. Keep winning! Two-time Emmy and Three-time NAACP Image Award-winning, television Executive Producer Rushion McDonald, interviewed Jocelyn Pearson. Purpose of the Interview The interview on Money Making Conversations Masterclass with Rushion McDonald and Jocelyn Pearson aimed to: Share Jocelyn’s journey of graduating debt-free by securing $126,350 in scholarships. Educate families on how to avoid student loan debt through her proven system, The Scholarship System. Dispel myths about scholarships and provide actionable steps for parents and students. Key Takeaways Scholarship System Approach Jocelyn developed a six-step process to simplify scholarship applications and avoid overwhelm. Focus on breaking the process into small, manageable steps rather than a vague “go get money” directive. Common Myths Debunked Too early or too late to apply: Start by junior year; it’s never too late—even college seniors can apply. Only perfect students or low-income families qualify: Many scholarships don’t require high GPA or athletic ability. All good scholarships are gone: Smaller, local scholarships ($500–$5,000) add up over time. It takes too much time: With a system and reusable materials, effort decreases each year. Avoiding Scholarship Scams Beware of “easy,” “enter to win,” or sweepstakes-style scholarships—they often sell personal data. Real scholarships require effort and personalization. Role of Parents Parents should help with planning and identifying legitimate scholarships but not complete applications for students. Committees can detect when parents write essays. AI in Scholarship Applications Jocelyn warns against copy-pasting AI-generated essays. Her platform introduced TESS, an AI assistant for ethical guidance and support. Financial Aid Basics Submit FAFSA even if you think you won’t qualify; some colleges and states require it. Combine all sources—government aid, institutional aid, and private scholarships. For Current College Students Check with financial aid offices, academic departments, and organizations for scholarships available after freshman year. Entrepreneurial Journey Jocelyn turned her passion into a business by starting with a book, building an email list, and launching webinars. She emphasizes persistence and ignoring naysayers. Notable Quotes “I had to accumulate my way to getting college paid for—the mere mortals’ way to going to college without tons of debt.” “Most families want scholarships, but they get stuck in the overwhelm.” “There’s no big red easy button—but with clear steps, it feels less daunting.” “We’re saying no to the broken system… It takes, on average, 21 years to pay off student loans.” “With great power comes great responsibility—AI can help, but only if used ethically.” #SHMS #STRAW #BESTSee omnystudio.com/listener for privacy information.
Private practices — ever feel like you can't win against the DSOs? Kiera talks about what your practice can continue to produce that DSOs, with their multiple locations and bigger budgets, can't replicate. Episode resources: Subscribe to The Dental A-Team podcast Schedule a Practice Assessment Leave us a review Transcript: Kiera Dent- Dental A Team (00:00) Hello, Dental A Team listeners. This is Kiera, and I am excited. Today's gonna be a fun rift of a podcast for you. It's gonna be like, can independent practices like private practice still beat DSOs if you want to? This isn't a rag on DSOs, it's not a rag on private practices. It's just can private practice, independent practices still win against the DSO? Because I think a lot of people are losing faith and confidence and feeling like if I'm not a part of a DSO, I can't win. So I wanna just tickle our brains today. Think about it in a different way and let's have a fun rift because you remember this is the best place. Dental A Team is the place where we are obsessed about helping you have your best life. We call it the Yes Success Model, where we focus on you and your vision, earnings and profitability, and then systems scale and structure for you. So that way you've got the systems, the structure, and the scalability for long term. I'm obsessed with helping teams and doctors align. I'm obsessed with dentistry. My last name's Dent for crying out loud. I love this. So let's do a good rift. Let's talk about. Can we really still win? It feels like DSOs, they got so much money over there and like they don't have to worry about margins. But I don't think that that's necessarily true. And I do believe last year, now I believe I know, DSOs had their first down year last year. So like I said, there is no I don't have a dog in the fight. My dog in the fight is which I don't even know where that phrase comes from. So if someone wants to pen pal me and tell me about this, I mean I could sure I could look it up. But like if you know, send me an email, Hello@TheDentalATeam.com. I'm Kiera. It's fun. It's fun to have a good pen pal over there. But I I think my dog in the fight is what's gonna be the best for dentistry long term. That is my dog in the fight. I wanna make sure that we as a population, money talks. I don't blame you. Getting a good multiple for your practice, like, why not? You're in you're in the golden era of dental practices. Or so they make it put on paper. Make sure it really is the golden era for you if you choose to go that route. and a lot of people have been very happy. A lot of people have also had the shorts burned off them. Tell me why that's a thing too. Like, How does someone get their shorts burned off them? I also want to know that phrase. I'm here for the phrases today, too, on the rift. And also, I'd love to know your opinion on this. So I'm gonna rift on my side and then shoot me an email. I'd love to hear. Hello@TheDentalATeam.com so okay, the rift today. DSOs are all like, okay, let me just go back, backing it up. Ultimately, I hope that we preserve the sanctity of great dentistry for patients forever. That's my hope. I don't care if it's DSOs, I don't care if it's private practice, I don't care what it is. But I do not dental practices are businesses. I also work in dentistry as a clinician. And I'm not okay with us squeezing margins and compromising care to hit margins. To me, that's just unfair. It's unfair for patients. I look at a lot of things in our healthcare system and I don't love it. And so I think that dentistry has been kind of the wild, wild west, and we've stayed out of a lot of it. And so I just hope that while we make Decisions financially and personally, I hope that we think about our long-term consequence. And that comes for a lot of new ones. I think that there's more senior doctors who have been in dentistry for a long time. Like they're not as worried about the multiples. I think a lot of our newer doctors who have a lot of debt on this, there's an easy cash payout. I would just say and a caution and an ask is let's just remember that we've had pioneers ahead of us who have paved the way. Let's continue to be those pioneers that are able to preserve and sanctify dentistry, whether that's through DSOs or private practices. So that's Kiera's dog in the fight. And I hope that you agree. And maybe you have a different opinion. So like let's have a good conversation. This is what we talk about in our monthly masterminds. So come chat, hang out. I'd love to have you there. okay. So can we win? Here are some thoughts of how private practices can still win against DSOs. So like DSOs, they do have more locations. They got bigger budgets, they've got bigger teams. But I do feel like there's still things in private practice that DSOs can't replicate. So I think the biggest threat is if you feel like you can't compete and you give up because there are ways and you can still win. you don't have to outspend a DSO. You don't need to out execute them. I do think that there's still competitive advantages. And so really leaning into whatever is going to be best for you. And again, remember, my my long term opinion is let's just make sure that we protect and sanctify dentistry for the health of our patients. And yes, I want you to be a profitable business owner. I think you can have both. I don't think it has to be one or the other. DSO or non-DSO. Now, private equity. I do have my opinions about private equity and I don't believe that they're always there for ethical reasons. and so I just say like make sure that the decisions you're making with your practice and your patient is going to help long term, the greater good. So those are my two cents on it. So anyway, just we work with a lot of practices. So I think that there's still ways that you can still compete against large corps, corporations. So number one. I think is independent practices, private practices, they have a connection that I think a lot of DSOs lack. DSOs tend to have a burn and churn model. So patients actually knowing their doctor, trusting their team, feeling remembered, feeling valued, and making sure that it's not a constant churn. Now I will have a call out. There are a lot of private practices that I know that are also on a burn and churn. They're churning associates left and right. They're turning team members. I get that it's hard right now, but I will say that if that's your practice, you are not competing against those corporate organizations that have. run of the mill doctors. So I think that private practices can have a consistent provider relationship. I absolutely hate going to the dentist. I actually have transferred away. Like I can get multiple doctors across the road, guys. I don't know if you know what I do for a living, but I work with a lot of dentists. So if I want sporadic care, aka I go into one practice, but I see different providers all the time. And I know we have the whole phrase of they're provide their patients at the practice, which is not wrong. But I'd say the more consistent you can have of team of providers that's going to help you win. I'm not saying to keep team members just because of longevity. I am here to say though, consistent provider relationships I do think will outshine. Just like I hate getting a new hair person, I hate getting a new nail person. People don't like to change that. So I think that that's going to be a zone for you. So personalized experience and long-term patient loyalty. Those are going to be key places that I think private practices can win. Now, DSL is listening, guess what? This is your edge as well. Like this is how you can have an edge. And ultimately we're all here for it. But these things need to stay and maintain. so I do believe that patients who see the same doctor, the same team for years are much less likely to leave based on price or convenience. Like they're going to stick with you. Why why? They don't want to change that up. I don't want to go show someone else my mouth and have that awkward moment where I just Don't like I don't have this, so just know that it's a space where like you gotta you gotta be connected, you've gotta help them feel seen and known. And I do believe that that's that is a zone. Daos are gonna scale systems, but like being able to scale genuine relationships, I do think is a harder thing. So in private practices, watch yourself. Look to see do you have those genuine connections? Are we scaling genuine relationships? Is that something that we're doing in those personalized pieces? You can stand out and still scale and still be profitable. So I just think maybe like if you're looking at this, looking at your team, wanting to have a reflection, what's a way that we can connect and create more personalized patient experiences? How can we keep like, if we want to keep a patient for life, what's gonna make them want to choose us consistently? If there's a competitor of price, if there's a competitor of convenience, how do we make sure those patients stay loyal to us? another thing I think that private practice can win on is believe it or not, in private practice, you can typically move faster. So I know a lot of people when they interview coming to Dental A Team coming from large organizations, like, I love that there's not as much red tape, Kiera. I love that you can move quicker, that we can make decisions faster, we're more nimble. And I think that that is a a huge selling point for private practices to be able to out outpace. You can have faster decision making, you can have faster implementation, you have less red tape, you've got greater flexibility. That can also create chaos. We gotta make sure that we don't we don't flex too much. But If patients are wanting something, we can usually make those calls. We can have a more personalized experience. We can have like great water bottles. We can have different things that make patients have it. So, like you can change scheduling systems, you can change patient experience protocols, you can have like different pieces that your patients can recognize and see. Where in large organizations, a lot of times it does take like months to implement these items. So I do think agility does create opportunity. And so for you to just look at this and think like, all right, what ways can we be more flexible, more adaptable? Like you don't have to have permission to improve things. So if you don't have to have permission to improve, what are we waiting to improve? What things could be improved upon? How can we make a better patient experience? How can we make a better team experience? there's a doctor that I was talking to the other day, and he said, Kiera, I have like a lifestyle practice and I want to attract team members for that lifestyle practice. And I just thought, like, how crazy cool is that? Because I think like that is the agility, that's the flexibility of a private practice that also makes it a great working place. For team members. Like it's not just about patients, it's also about team and attracting those. So that's another zone where I think private practice can still outpace a DSO for sure. and then I do think that systems can win more than size in a lot of ways. So a lot of times we think bigger is better. And I know a lot of times I'm intimidated as a smaller business, if you will, smaller, medium sized business compared to large organizations. You're like, they just have all of it, but it's not true. So I've seen in a lot of our practices very profitable independent practices, really strong leadership, great patient experiences, clear accountability. And you actually can have scheduling protocols and case acceptance systems and leadership systems and financial systems. Smaller scale practices can actually have great, incredible systems in place. So I do believe that a well-run single location or maybe two or three locations oftentimes will outperform on profitability than larger organizations because of. execution is a lot stronger. So it's having those systems that are clear, having those and I know a lot of people are like, but what systems? And our team kind of boiled it down. There's about like 10 to 15 core systems that every practice that they'll implement and execute on, they're going to thrive. And it's scheduling case acceptance, leadership, like our billing protocols, things like that, morning huddles, very basic, non-sexy, but having those systems is going to be much grander than size. And the larger you get, yes, they try to implement these, but I do still feel like those systems can create those predictable experiences. They can have a more customized experience. And then if we don't like it, we can pivot that system, we can refine that system, we can make it better. So I just think it's like, how do you have the best run practice? Not necessarily the biggest practice. And so I would look at your practice and what are the systems? What are the gaps? Where do we slip? Where do patients maybe fill that? Where does our team fill that? Let's fix that. Let's organize that. Let's let's have that. That way we're able to. To be able to outpace. And so there's lots of ways. I think having those genuine connections, making sure that we've got systems that grow with us, having strong leadership teams, having a great, like, I don't know, community feel. There's just something different. Think about it. I have a fee-for-service chiropractor versus corporate chiropractor. I've gone to both. And one is like run of the mill. I come in, I'm out, checked in, checked out, it's cheaper. But they don't know me. They don't have an experience. They don't spend time with me. I do think time that doesn't mean like, 10 minutes, it means genuine connection time. Those things you still can win. And believe it or not, a lot of our practices are sitting at 20, 30, 40% profit margins. You can still be very profitable and not need to be in a DSO. They have a ton, they can scale a lot, but I also think there's an autonomy piece, there's a creativity piece, there's a branding piece of you being able to brand your location as you, to be able to give an experience that is very custom to you, to your audience. So I do believe that there's still a way that private practice can. win against DSLs. I think there's space for both. I don't think one's right or wrong. But again, like I said, my dog in the fight is whatever we choose to do, whatever pieces we're doing, let's just remember to keep the sanctity of dentistry pure. Let's make sure that we're doing what's in the best interest of our patients. Let's make sure we're not cutting corners on dentistry. We're not trying to fit people in to hit production goals just to hit production goals. We're not compromising the the products that we use to be able to to hit the right profit margins. I believe that being the best for our patients will always, always follow profitability. So with that, right now, let's remember, like relationships, you get it. You guys also have speed and flexibility and agility. And you also do have systems that you can put into place that will like beat size. They have other things. They've got billing, they've got it, but think about it, you're gonna sell to them anyway. You might as well get those things in place and you might as well try on your own. So You think they're gonna take over all your problems? You can do this on your own and you don't have to sell your practice. I had a dentist who was wanting to sell his practice. He's like, Kiera, I'm gonna sell to a DSO. And I said, Great, like I'm here for it. Let's rally. And then he's like, I said, let's just think through. Like when they buy you, what are they going to do? He's like, they're gonna expand the practice and they're gonna put systems into place. And I said, Well, do you wanna do that? And he was like, Yeah, like I'm gonna sell to them and they're gonna make all the money on the things that I could just do. And I was like, I'm really proud of you because you're exactly right. So if they're gonna do that anyway, why not do it yourself and reap the rewards? We're here to help you. I helped that practice. We went took them from eight eight ops to 15 ops. They're doing amazing. They're on track to be a five million dollar practice this year. All those things the DSO would have done. We were able to put the systems in place, build the leadership team, expand the practice. They're crushing it. They're doing incredible over there. So again, flexibility, agility, great patient care. And where they go, because people are like, well, a DSO is the only person who's gonna buy me. That's not true. Stop limiting yourself. There's lots of people. There's lots of options. Do you know how many people come out of school wanting to buy? They come out and they do want to buy. So don't limit yourself, you guys. People are like, they can't afford homes. That's not true. There's a lot of people willing to there might now need to be two partners instead of just one partner. It's okay. It's gonna look a little different, but that doesn't mean that that's the only option. So I really do think that private practice dentistry can absolutely thrive. I do believe that the practice is winning today, DSO or private practice. Aren't the ones with multiple the most locations? They're the ones with the strongest leadership, strongest systems, best connection, best patient care. Those are the ones that are winning. So look at it. Do we have the best leadership? Do we have the best systems? Do we have the best patient connection? Do we have the best experience? And if so, fantastic. Make sure that all your patients know it. Make sure that people are talking about it. You guys can do this. I love helping private practices. Whatever your dream is, whatever your goals, if it's DSO or not, I don't care. I just want you to have your best life. And like, hey, before you sell, maybe let's chat. Let's talk about it. Let's see if there's a way that we could actually take that load and that stress and that annoyance off and you reap the rewards before you go and sell. there's so many ways that we can do this. So reach out. Hello@TheDentalATeam.com. And as always, thanks for listening. I'll catch you next time on the Dental A Team podcast.
Listen and subscribe to Money Making Conversations on iHeartRadio, Apple Podcasts, Spotify, www.moneymakingconversations.com/subscribe/ or wherever you listen to podcasts. New Money Making Conversations episodes drop daily. I want to alert you, so you don’t miss out on expert analysis and insider perspectives from my guests who provide tips that can help you uplift the community, improve your financial planning, motivation, or advice on how to be a successful entrepreneur. Keep winning! Two-time Emmy and Three-time NAACP Image Award-winning, television Executive Producer Rushion McDonald, interviewed Jocelyn Pearson. Purpose of the Interview The interview on Money Making Conversations Masterclass with Rushion McDonald and Jocelyn Pearson aimed to: Share Jocelyn’s journey of graduating debt-free by securing $126,350 in scholarships. Educate families on how to avoid student loan debt through her proven system, The Scholarship System. Dispel myths about scholarships and provide actionable steps for parents and students. Key Takeaways Scholarship System Approach Jocelyn developed a six-step process to simplify scholarship applications and avoid overwhelm. Focus on breaking the process into small, manageable steps rather than a vague “go get money” directive. Common Myths Debunked Too early or too late to apply: Start by junior year; it’s never too late—even college seniors can apply. Only perfect students or low-income families qualify: Many scholarships don’t require high GPA or athletic ability. All good scholarships are gone: Smaller, local scholarships ($500–$5,000) add up over time. It takes too much time: With a system and reusable materials, effort decreases each year. Avoiding Scholarship Scams Beware of “easy,” “enter to win,” or sweepstakes-style scholarships—they often sell personal data. Real scholarships require effort and personalization. Role of Parents Parents should help with planning and identifying legitimate scholarships but not complete applications for students. Committees can detect when parents write essays. AI in Scholarship Applications Jocelyn warns against copy-pasting AI-generated essays. Her platform introduced TESS, an AI assistant for ethical guidance and support. Financial Aid Basics Submit FAFSA even if you think you won’t qualify; some colleges and states require it. Combine all sources—government aid, institutional aid, and private scholarships. For Current College Students Check with financial aid offices, academic departments, and organizations for scholarships available after freshman year. Entrepreneurial Journey Jocelyn turned her passion into a business by starting with a book, building an email list, and launching webinars. She emphasizes persistence and ignoring naysayers. Notable Quotes “I had to accumulate my way to getting college paid for—the mere mortals’ way to going to college without tons of debt.” “Most families want scholarships, but they get stuck in the overwhelm.” “There’s no big red easy button—but with clear steps, it feels less daunting.” “We’re saying no to the broken system… It takes, on average, 21 years to pay off student loans.” “With great power comes great responsibility—AI can help, but only if used ethically.” #SHMS #STRAW #BESTSteve Harvey Morning Show Online: http://www.steveharveyfm.com/See omnystudio.com/listener for privacy information.
Plus: trade technology provider Altana acquires AI platform to tackle customs complexity. And Coinbase stock jumps after Clarity Act clears major hurdle. Julie Chang hosts. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.
© 2020-2026 Ivy Podcast Discovery LLC
