Podcasts about layers

If anal play sounds like something you'd have to grit your teeth through, I want to change your mind about one thing first: a painful first try isn't your body's final answer. It's usually a sign something was rushed, unprepared, or out of your control, not proof this isn't for you. In this episode, I'm walking through my five-step, comfort-first framework for exploring anal: getting a real yes before anything starts, setting yourself up for success, warming your whole body up first, moving in layers instead of straight to penetration, and learning to actually read your body's feedback. I'm also busting a few myths along the way, including the one that penetration is the goal at all. Pleasure, not performance. Let's get into it. ABOUT EMILY: Emily Morse is a Doctor of Human Sexuality, author and host of the #1 rated Sex with Emily podcast. Known as a renowned sexologist, Dr. Emily has helped millions of people around the world navigate their sex lives. Her candid and often funny conversations challenge cultural taboos, misinformation and awkward sex talks to create a future where people can deeply connect and embrace pleasure-filled lives. Because, life is too short for bad sex. CONNECT: Instagram: ⁠https://www.instagram.com/sexwithemily/⁠ X: ⁠https://twitter.com/sexwithemily⁠ Facebook: ⁠https://www.facebook.com/sexwithemily⁠ TikTok: ⁠https://www.tiktok.com/@sexwithemily⁠ Threads: ⁠https://www.threads.net/@sexwithemily⁠ WANT MORE? Visit the Website: ⁠https://sexwithemily.com/⁠ which includes FREE guides. Free Downloadable Guides: ⁠https://sexwithemily.com/guides/⁠ Text With Me: ⁠https://sexwithemily.com/text⁠ Receive Sex Tips On The Regular: ⁠https://sexwithemily.com/subscribe⁠ Interested in 1:1 Coaching with Emily? Go to ⁠http://sexwithemily.com/coaching⁠ to apply! This episode is sponsored by: Bellesa: EVERYONE who signs up wins a FREE Rose suction toy with their order! ⁠https://www.bboutique.co/vibe/emilymorse-podcast⁠ ============================= Chapters: 00:00 Cold Open / Welcome to Sex with Emily 01:24 Why Your First Time May Have Hurt 02:08 The Five Step Framework (Meet Dr. Emily) 02:56 Anal Isn't All or Nothing 03:44 Your Nervous System Is the Real Sex Organ 04:07 The Two Rings of Muscle, Explained 04:49 Step 1: Start With a Real Yes 05:28 The Receiving Partner Is Always in Charge 06:00 Step 2: Set Yourself Up for Success (Hygiene) 06:41 Lube, Toys, and the Pre-Talk 07:17 Step 3: Warm Up Your Whole Body 08:22 Step 4: Move in Layers, Not Pressure 09:59 Step 5: Listen to Your Body and Talk About It 10:51 Busting the Biggest Anal Myths 11:11 Recap: The Five Steps to Try Anal Comfortably 11:58 Outro & Where to Find More Learn more about your ad choices. Visit megaphone.fm/adchoices

An AI agent can finish a task and still violate the rules that matter most. In supply chain, that gap can affect cost limits, approved suppliers, compliance requirements, safety protocols, and escalation paths. In this episode of Supply Chain Now, Scott W. Luton speaks with Vin Vashishta, CEO and AI strategist at V-Squared, about intent contracts, audit trails, workflow reorchestration, tokenomics, semantic layers, and evidence-based AI strategy. Vin explains how to evaluate AI by the value it creates, budget for recurring usage costs, work with imperfect information, and require consultants to connect every recommendation to evidence, risk, mitigation, and business-specific ROI. Jump into the conversation: (00:00) Introduction (05:38) Intent contracts and their role in AI agent management (08:38) The need for AI audit trails in supply chain (11:10) CIO considerations for managing AI demand at scale (14:21) Gaps in workflow reorchestration and value quantification (17:00) Lessons from the semantic layer meme on imperfect data (19:36) The outcomes an AI strategy should deliver for a specific business (25:03) AI training programs generating the strongest market response (26:49) Ways to follow Vin and learn more Additional Links & Resources: Connect with Vin: https://www.linkedin.com/in/vineetvashishta/ Learn more about V-Squared: https://vsquaredai.com/ Vin's LinkedIn Post about Managing AI Agents: https://bit.ly/Managing-AI-Agents The cost of intelligence: How CIOs can manage AI demand at scale: https://mck.co/3TUD3mz Vin's LinkedIn Post about the Cost of Intelligence: https://bit.ly/Vin-on-CIO-Managing-AI Vin's LinkedIn Post on The Semantic Layers: https://bit.ly/Vin-on-Semantic-Layers Vin's LinkedIn Post on AI Strategy: https://bit.ly/Vin-on-AI-Strategy-2026 Vin's Training & Certification Classes: https://datascience.vin/ Learn more about our hosts: https://supplychainnow.com/about Learn more about Supply Chain Now: https://supplychainnow.com Watch and listen to more Supply Chain Now episodes here: https://supplychainnow.com/program/supply-chain-now Subscribe to Supply Chain Now on your favorite platform: https://supplychainnow.com/join Work with us! Download Supply Chain Now's NEW Media Kit: https://supplychainnow.com/media-kit/ WEBINAR- SAP AI Inside the Supply Chain: From Silo to Orchestration: https://bit.ly/4bvpz6K WEBINAR- Operational AI in the Supply Chain: How context empowers agents and humans to operate side by side: https://bit.ly/4x7Vd2Z WEBINAR- You Can't Manage What You Can't See: Using Visibility, KPIs, and AI to Optimize Logistics Operations: https://bit.ly/4ql6iem Gartner Announces 2026 Rankings of the Global Supply Chain Top 25: https://www.gartner.com/en/newsroom/press-releases/2026-06-17-gartner-announces-2026-rankings-of-the-global-supply-chain-top-25 This episode was hosted by Scott Luton and produced by Trisha Cordes, Joshua Miranda, and Amanda Luton. For additional information, please visit our dedicated show page at: https://supplychainnow.com/ai-agents-semantic-layers-cio-leadership-1629 The content in this episode, including all audio, videos, visuals, and graphics, is the property of Supply Chain Now and is protected by copyright law. Unauthorized use, reproduction, distribution, modification, or re-uploading of this content in any form is strictly prohibited without explicit written permission from Supply Chain Now.For licensing inquiries or permissions, please contact us at production@supplychainnow.com© 2026 Supply Chain Now. All rights reserved. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

William's Etsy Store - https://www.etsy.com/shop/PermacultureStoreWilliam's Permaculture Design Course - https://patreon.com/ThePermacultureConsultant?utm_medium=unknown&utm_source=join_link&utm_campaign=creatorshare_fan&utm_content=copyLinkWilliam's Channel - www.youtube.com/@ThePermacultureConsultantWilliam's Linktree - https://linktr.ee/ThePermacultureConsultant?utm_source=linktree_profile_share<sid=13182d07-8cfe-4e2f-9b52-aa564df0fcf6Eric Seider's Youtube Channel - http://www.youtube.com/@EricSeiderEric Seider's Tshirts - https://www.ericseider.com/pimpgearPerma Pastures Farm - https://permapasturesfarm.com/Sovereign Health Summit with Barbara O'Neill, October 27-31, 2026 - https://www.sovereignhealthsummit.com/?ref=permaPromo Code - TPC - 5% OffJonathan Otto's Website - https://myredlight.com/Promo Code - bb25 and bb30The Wellness Blanket -https://thewellnessblanket.com/?sca_ref=10936149.IjFZC1tt28OPromo Code - Perma - 10% OffAzure Standard - https://www.azurestandard.com/?a_aid=dd1f60ff5dPromo Code - FOODFORHEALTH1515% Off for New Customers Minimum Order $100Nesa's Hemp - https://www.nesashemp.com/#permapasturesfarmPromo Code - perma - 10% OffBon Charge Blue Light Blocking Glasses - https://boncharge.com/?rfsn=8947983.d7b6efPromo Code: Perma - 15% OffSoil Savior Products - https://www.soilsaviors.org/order?aff=654693f413fad4692e058e9eb0779d3667638550392d22d979d6d2d4daf720b3Cell Saviors - https://www.cellsaviors.org/fulvicPromo Code: detox - Get 10% OffWAVwatch - $100 Off - https://buy.wavwatch.com/?ref=billy100Promo Code: BILLY100Micronic Silver - https://www.micronicsilver.com/?ref=PERMAPASTURESFARMPromo Code - perma 10% offEMF Rocks - https://emfrocks.com/PERMAPASTURESFARMPromo Code - Perma Pastures Farm - 5% OffAir Water Healing Triad Air Filter - https://airwaterhealing.com/Promo Code: perma - Get 10% OffLiving Soil Foundation GiveSendGo - https://givesendgo.com/GE2E8?utm_source=sharelink&utm_medium=copy_link&utm_campaign=GE2E8If you would prefer to send a check:Living Soil FoundationPO Box 2098Mars Hill, NC 28754Richardson Nutritional Center https://rncstore.com/permaPromo Code: perma - Get 10% OffRedmond Products - 15% Off - https://glnk.io/oq72y/permapasturesfarmPromo Code: permaGet $50 Off EMP Shield: https://www.empshield.com Promo Code: permaAbove Phone - https://abovephone.com/perma/Promo Code - PERMA $50 OffHarvest Right Freeze Dryer: https://affiliates.harvestright.com/1247.htmlOnline Pig Processing: https://sowtheland.com/online-workshops-1Patreon: https://www.patreon.com/user

Send us a text

Airtightness is one of the most important—and most measurable—parts of building performance. In Part 2 of the Back to Basics series, Steve, Jake, and Pete break down how air control layers work, why continuity matters, and how blower door testing exposes the difference between theory and reality.The crew starts with several principles that apply across multiple control layers: staying in one plane makes continuity easier, simpler details are more likely to succeed, and combining control functions in a single material can reduce opportunities for failure. They also revisit one of their favorite ideas: eliminate “confused spaces.” Attics, crawlspaces, and other areas should clearly belong either inside or outside the building enclosure.Then the conversation gets specific to airtightness. Blower doors do not lie. ACH50 provides the quantitative result, but finding the actual leaks—their size, location, and number—is just as important. The boys discuss why one blower door test is rarely enough, how wind, stack effect, and fans drive air leakage, and why two mediocre air control layers do not equal one good one.For renovations, the sequence is simple: blower door test first, air seal second, insulate third.Pete's Resources:EPA – Moisture Control Guidance for Building Design, Construction, and MaintenanceENERGY STAR Thermal Bypass ChecklistA House Needs to Breathe...Or Does It? An Introduction to Building Science by Allison Bailes

⁠

Self-care podcast exploring The Hidden Layers Below Emotional Eating Triggers, Why One Trigger Can Create a Whole Eating Spiral & Steps to Addressing Emotional Eating Triggers. TOPICS:: ** The Hidden Layers Below Emotional Eating Triggers (06:33). ** Why One Trigger Can Create a Whole Eating Spiral (19:19). ** Steps to Addressing Emotional Eating Triggers (35:04).   NOTES:: Show notes: amberapproved.ca/podcast/686 Leave me a review at amberapproved.ca/review Email me at info@amberapproved.ca   Take the NEW Free Hormone Imbalance Quiz here: https://amberapproved.ca/hormone-imbalance-quiz    Subscribe to newsletter: https://amber-romaniuk.mykajabi.com/newsletter-sign-up    SHOW LINKS: Click below to schedule a 30 minute Complimentary Body Freedom Consultation https://amberapproved.ca/body-freedom-consultation/  Take my free Emotional Eating Quiz here: http://amberapproved.ca/emotional-eating-quiz Listen to Episode 668 about what it's like to work with me here: https://amberapproved.ca/podcast/668  Follow me on Instagram www.instagram.com/amberromaniuk Youtube Channel: https://www.youtube.com/@amberromaniuk/    MY PARTNERS: Designs For Health Blood Sugar Bundle! One of the hardest parts of overcoming my emotional eating was the INSANE SUGAR and carb cravings. It was the intense sugar and carb cravings. That's why I created my Blood Sugar Bundle with Designs for Health to support your body while you work on emotional eating and breaking binge patterns. It includes chromium for blood sugar balance, L-glutamine to help curb cravings fast, a high-quality probiotic for gut health, and a clean Pure Paleo protein powder to keep you full and stable. I only recommend Designs for Health to my clients because they are third-party tested, family-owned, and use the highest quality ingredients. Quality matters when it comes to truly supporting your body and getting results. Get 30% off The Blood Sugar Bundle in USA and Canada automatically applied at checkout below! Canada Blood Sugar Bundle here for 30% off!  USA Blood Sugar Bundle here for 30% off! You can also get 30% off any Designs for Health supplements anytime, it's my gift to you. Canada: www.designsforhealth.ca  (code AMBER30) USA: www.designsforhealth.com (code AMBER88)   MY PARTNERS: HERBAL FACE FOOD Stubborn eczema, red spots, aging spots, or acne on your face, chest, arms, or back from hormones or hard water damage? I have something SO amazing for you. Our mineral-heavy water started impacting my skin the moment we moved, and I had never experienced eczema or skin issues in my life. Hard water can strip natural oils, disrupt the microbiome, and weaken the skin barrier — leaving you inflamed, reactive, and stuck in flare-ups. That's exactly what happened to my neck, and within almost two weeks of using Herbal Face Food, my eczema is almost gone. This stuff is legit — and I only share my favorite things that actually work. Their formulas are made with ultra-potent, organic, whole-plant botanicals rich in antioxidants, polyphenols, and phytonutrients that calm inflammation, rebuild the skin barrier, and fight visible signs of aging instead of masking symptoms. The Cure is their most targeted antioxidant treatment designed to visibly improve stubborn concerns like eczema, melasma, rosacea, scarring, sun damage, and deeper signs of aging. The Cream is a deeply hydrating, ultra-potent botanical moisturizer that firms, smooths, strengthens the skin barrier, and helps reduce fine lines and wrinkles. The Soap gently cleanses without stripping, using powerful plant concentrates to protect, nourish, and support healthy, youthful-looking skin on both the face and body. If your skin has been struggling from hormones, environmental stress, hard water damage, or premature aging, I'm sharing exactly what I'm using and why it's working. Shop through my link in the show notes or visit https://herbalfacefood.com/?ref=AMBER88 and use code AMBER88 at checkout for 30% off your entire order.

Psychologist Russ Poldrack is a mind reader of sorts, but not in the traditional sense of the word. Instead, he makes movies of blood flow in the brain using functional MRI to draw insights into the function of specific regions of the brain to decipher how they control behavior. In one example, Poldrack has shown how people with obsessive-compulsive disorder have enhanced “salience networks” that turn on when surprising things happen. This finding could have “very profound” outcomes in diagnosis and treatment, Poldrack tells host Russ Altman in this wide-ranging exploration of the implications and the ethics of functional neuroimaging on Stanford Engineering's The Future of Everything podcast. Have a question for Russ? Send it our way in writing or via voice memo, and it might be featured on an upcoming episode. Please introduce yourself, let us know where you're listening from, and share your question. You can send questions to thefutureofeverything@stanford.edu. Episode Reference Links: Stanford Profile: Russ Poldrack Poldrack Lab website: http://poldracklab.org The MyConnectome project: https://www.myconnectome.org/ Stanford Center for Open and Reproducible Science: https://datascience.stanford.edu/cores Connect With Us: Episode Transcripts >>> The Future of Everything Website Connect with Russ >>> Threads / Bluesky / Mastodon Connect with School of Engineering >>> Twitter/X / Instagram / LinkedIn / Facebook Chapters: (00:00:00) Introduction Russ Altman introduces guest Russ Poldrack, a professor of psychology and of psychiatry and behavioral science at Stanford University. (00:03:15) Path into Neuroimaging How Poldrack moved from cognitive psychology into functional MRI research. (00:04:18) Functional MRI How MRI can be used to track brain activity. (00:06:59) Mapping Brain Activity How tasks and behaviors are linked to patterns of brain activation. (00:08:02) Resting Functional MRI How resting scans revealed large-scale brain networks. (00:09:00) Brain Networks and Behavior How the brain controls stopping, switching, and habits. (00:10:34) Layers of Brain Organization Why neuroimaging studies the brain across multiple scales. (00:13:04) Individual Brain Differences How shared brain networks vary across individuals. (00:13:34) Scanning Himself Why Poldrack repeatedly scanned his own brain. (00:15:47) Mapping One Brain Deeply How repeated scans revealed reliable individual brain patterns. (00:18:59) Neuroimaging and Mental Health How brain network differences may relate to depression and OCD. (00:21:42) Reproducibility in Brain Imaging How larger and deeper data sets improve reliability. (00:24:51) Analytic Variability Why the same data can produce different conclusions. (00:26:39) Multiverse Analysis How testing many analyses helps identify stable findings. (00:27:55) Open Science How data sharing and standards improve transparency. (00:30:29) Privacy and Brain Data How researchers protect participants when sharing neuroimaging data. (00:32:17) Future In a Minute Rapid-fire Q&A: neuroimaging, reproducibility, and AI in science. (00:33:22) Conclusion   Connect With Us:Episode Transcripts >>> The Future of Everything WebsiteConnect with Russ >>> Threads / Bluesky / MastodonConnect with School of Engineering >>>Twitter/X / Instagram / LinkedIn / Facebook Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Send us Fan Mail

In this episode of Make Your Message a Movement, host Tiffany Neuman explores the concept of building an authority brand and the invisible ceiling that founders encounter when their external brand no longer matches their internal identity, or vice versa. Rather than treating branding as a superficial collection of logos and content, Tiffany explains that a truly powerful authority brand requires deep alignment across four essential layers: magnetic messaging, visual expression of authority, a conversion-based platform, and energetic congruence.Key TakeawaysThe Four Layers of an Authority Brand:Magnetic Messaging: Clear positioning and a compelling point of view that addresses audience psychographics, preventing you from getting lost in a sea of generic, AI-generated voices.Visual Expression of Authority: Intentional, modern, and distinct design choices that close the gap between your true expertise and how the market perceives your credibility.Conversion-Based Platform: A strategically built website (not just a digital brochure) that confirms credibility, reduces friction, and guides high-level clients, event organizers, and partners to take the next step.Energetic Congruence: The ability to fully inhabit the brand position you have built—standing in your authority, quoting your rates without apology, and expanding your capacity to receive new opportunities without shrinking back.Branding for Where You're Going: When you reverse-engineer a brand for the next 3 to 10 years of your business, your perception catches up and reality accelerates.The Invisible Ceiling: Launching an upgraded brand can sometimes trigger your ego or stretch your capacity. The real expansion often begins after the website launches as the market responds to your elevated positioning and you learn to step into that room.Mentioned ResourcesConnect with Tiffany on LinkedIn: https://www.linkedin.com/in/tiffanyneuman/Schedule a call with Tiffany: https://tidycal.com/tiffany1/bosfdiscoverycallRate, Review, and Follow on Your Favorite Platform! If you loved this episode, leave us a review. And always make sure you're following the podcast so you never miss an episode. Follow now!

Special Episode: My World-Famous Minimal Packing List   Dr. Travelbest Adventures Minimal packing list for a 90-day trip around the world. Note: laundry every 7 days, with a goal of smart layers and tiny toiletries. The core •   *Carry-on backpack (35–50L) + packable daypack/tote Packing cubes, nylon (4, one for toiletries) •   Money belt or two if you can for different currencies. •   Small lock + ten zip pouches clear, for meds, tech, misc. toiletries Clothing Tops •   3–4 tops (mix of short + long sleeve; quick-dry) •   1 nicer top (dinner/church/ photo ops) •   Scot-T vest with 10 pockets (extra-sized so that you can overlay) 5 Steps to Solo Travel  |  Part C Bottoms •          1 pair of travel pants •          1 pair of shorts and 1 skirt Layers •          1 light warm layer •          1 packable rain jacket (hooded) •          1 thin "modesty" layer: light scarf or overshirt (also sun/AC/ temple coverage) Dress option •          1 packable permanent press nylon dress Underwear •          7-10 underwear •          2 bras (or 1 bra + 1 bralette/sports bra) •          5-7 pairs of socks (include 2 warmer and no-show socks) Swim •          1 swimsuit, cap, goggles + quick-dry shorts Shoes (keep it to 2) •          1 primary walking shoe (supportive sneakers) •          1 lighter flats second shoe (sandals/flip-flops for showers/ beach) Toiletries (tiny + refillable) •          Toothbrush, small toothpaste, floss •          Deodorant stick •          Travel-size shampoo/soap OR solid bar •          Minimal skincare + sunscreen (you'll restock locally) •          Razor •          Tweezer •          Small brush/comb + hair ties/clips •          Mini first aid: a few bandages, blister care, tiny antiseptic wipes •          Sanitary female items and toilet paper as needed.   Health + meds (don't guess here) •          Prescriptions in original bottles + photo of prescriptions •          Pain/fever meds, antihistamine, anti-diarrheal, motion sickness (as needed) •          A few electrolyte packets •          Masks (a couple) + hand sanitizer   Tech (only what you'll actually use) •          Phone + charging cable with wall connection •          Universal adapter (a second one for South Africa travel) •          Power bank •          Earbuds/headphones (noise-canceling if you have them) •   Optional: lightweight laptop/tablet, only if it's essential for your work. Consider NOT bringing it for weight and security reasons. I did not. •   A few items saved: passport scan, insurance, itinerary, key addresses, and photos of loved ones. •   E- eSIM code: MARY2856 with Airalo if you don't have an eSim yet on your phone for data) You.will receive a discount, and I will receive a smaller fee.   Travel documents + money •   Passport + 1 backup ID, Global Entry Card •   2 credit/debit cards (stored separately), other IDs for discounts like Triple A (AAA), •   Some cash in USD + local starter cash when possible •   Travel and health insurance info and cards •   Pen (immigration forms and misc.)   Safety + comfort (high value, low space) •   Sleep kit: earplugs (mentioned above) and ask •   Reusable water bottle •   Small microfiber towel   A few "don't regret it" items •   Compression socks (long flights) •   Light tote bag for groceries/laundry •   Small "confidence items" (lip color, earrings, or a scarf)— makes outfits feel new on day 47 •   Washer sheets for laundry 118 Appendices •   Fanny pack or crossbody bag with adjustable strap and many zipper pockets   Minimal "rules" that keep you light •   No duplicates or "just in case" items (buy locally if needed). •   Wear the bulkiest: main shoes + jacket on travel days. •   Choose a color palette so everything matches. (ex. Navy blue, black, brown)   https://www.amazon.com/s?k=5+steps+to+solo+travel+dr+travelbest&crid=1ZVA4W0JU46M8&     Connect with Dr. Travelbest 5 Steps to Solo Travel website Dr. Mary Travelbest X Dr. Mary Travelbest Facebook Page Dr. Mary Travelbest Facebook Group Dr. Mary Travelbest Instagram Dr. Mary Travelbest Podcast Dr. Travelbest on TikTok Dr.Travelbest on YouTube In the news  

In this episode, Josh Bray discusses how endurance training shaped his coaching perspective, the role of games and hidden intention in athletic development, and how roughhousing and free play build adaptable movers. He also explores bucketing athletes as elastic, muscular, or hybrid, using force-plate data without overcomplicating the process, balancing readiness with meaningful training stress, and employing open-ended prescriptions that let athletes self-regulate volume while preserving intent, quality, and engagement. Today's episode is brought to you by Hammer Strength.

Water is the number one enemy of building durability—and the first control layer every builder, architect, and designer needs to understand. In Part 1 of a four-part Back to Basics series, Steve, Jake, and Pete break down the principles of water management that every high-performance building depends on.The conversation starts where most rain begins: the roof. From roof design and drainage to flashing strategies and capillary action, the crew explores how liquid water moves through buildings and why good design is the first line of defense. They also discuss how patterns of water damage in both historic and modern buildings reveal timeless lessons about moisture management.One of the episode's most interesting discussions comes when building science theory meets real-world practice. Steve approaches water management from an architect's perspective, while Jake explains the practical realities builders face every day. The result is a thoughtful conversation about balancing ideal building science with constructability in the field.This is Part 1 of a four-part series covering the four fundamental building control layers: Water, Air, Vapor, and Thermal.Pete's Resources:Joe Lstiburek – Moisture Control for BuildingsEPA – Moisture Control Guidance for Building Design, Construction, and MaintenanceWater in Buildings: An Architect's Guide to Moisture and Mold

For more of my latest content, subscribe to my YouTube channel, Dark Asia with Megan and join our awesome community. Your support means everything, and I can't wait to share more Asian cases with you! On Other Platforms: • TikTok: https://www.tiktok.com/@darkasiawithmegan • Instagram: https://www.instagram.com/darkasiawithmegan • Facebook: https://www.facebook.com/darkasiameganlee Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Lighting has evolved from a functional necessity into one of the most expressive tools in the interior designer's palette. With intelligent fixtures, light can now shift in colour, temperature and intensity throughout the day, warming mornings, energising afternoons, and creating cinematic evenings, all tuned to how we actually live. It can transform art, sculpture and materials within the home, add layers and drama, quietly shaping mood, emotion and experience. How are interior designers and lighting specialists working together to use light and shade in new and exciting ways? How could we see these ideas in action at WOW!house 2026 and what is the likely impact of such creative, design sensitive collaborations on the future of luxury residential design? Host Jeff Hayward is joined by WOW!house 2026 room designer, Samantha Bartlett, Luke Thomas from John Cullen Lighting, James Ratcliffe from Homeplay and Shaun Gumbs from Lutron in front of a live audience of designers to find out. This episode was recorded at the Design Club in the Design Centre Chelsea Harbour. Our thanks to our partners, Lutron, and to the Design Centre for their support. The interior design business is a Wildwood Plus production.      Chapters (00:00:00) - Wonders of the Interior Design Business(00:01:59) - Lighting Design for Luxury Homes(00:03:22) - Interior Lighting(00:08:58) - Lighting Design and Interior Design(00:13:24) - What are Layers in Lighting?(00:14:46) - What is the difference between LEDs and halogen light Bulbs?(00:18:26) - The Making of a Kitchen(00:19:12) - How Do You Design a Lighting Project?(00:19:54) - The bird's beak(00:21:49) - Living With Quetra(00:27:17) - Intelligent Lighting(00:32:49) - Wireless Lighting Installation(00:38:00) - Lighting(00:43:15) - Does Lighting Affect Your Projects?(00:44:09) - Lighting Your Home(00:45:39) - The Interior Design Business

rWotD Episode 3381: Hush (Asobi Seksu album) Welcome to random Wiki of the Day, your journey through Wikipedia's vast and varied content, one random article at a time.The random article for Thursday, 6 August 2026, is Hush (Asobi Seksu album).Hush is the third studio album by American shoegaze band Asobi Seksu. It was released on February 17, 2009 by Polyvinyl Record Co., marking the band's first album for the label. Hush was recorded in the summer of 2008 and was produced by Chris Zane, who had also worked on Asobi Seksu's previous album Citrus (2006). The album demonstrated a shift from the more shoegaze-inspired work of prior releases to a mellower, quieter sound.Hush produced four singles: "Me & Mary", released on November 17, 2008; "Familiar Light", released on February 16, 2009; "Transparence", released on August 21, 2009; and "Layers", released on December 7, 2009.This recording reflects the Wikipedia text as of 01:00 UTC on Thursday, 6 August 2026.For the full current version of the article, see Hush (Asobi Seksu album) on Wikipedia.This podcast uses content from Wikipedia under the Creative Commons Attribution-ShareAlike License.Visit our archives at wikioftheday.com and subscribe to stay updated on new episodes.Follow us on Bluesky at @wikioftheday.com.Also check out Curmudgeon's Corner, a current events podcast.Until next time, I'm standard Joanna.

Corey Goes Live: We got an impromptu pre-show from Corey Feldman on his Instagram Live. What a great way to start the show.Jump Back Black Cat: A new Corey Feldman song has dropped! I hope you kept the pact to enjoy it with the boys.Corey's Twitter: Corey is in hype mode on his Twitter with his new music release and his upcoming Birthday Bash.COREY FELDMAN!, SHOW STOPPER!, LET'S JUST TALK!, DON CHEADLE!, BOOGIE NIGHTS!, JIM AND THEM IS POP CULTURE!, COREY FELDMAN SHOW!, REAL ONES!, PO BOX!, ALL DAY GOONS!, SPECIAL FRIDAY NIGHT!, PRE-SHOW!, COREY FELDMAN INSTAGRAM LIVE!, PROMOTION!, MONSTER TRUCKS!, PARANOIA!, FELDDOGSUMMER2!, HACKED IN REAL TIME!, ISSUES!, SKIPPED!, FULL SIGNAL!, LIVES!, KISSING MY 22!, TOUR!, WHISTLEBLOWER!, MISSION FROM BAAL!, TROLLS!, CONSPIRED OPPOSITION!, PAID!, SELL TICKETS!, CONSPIRACY!, ARTIST!, MEAN!, DARK VIBE!, WAITER!, MONSTER TRUCKS!, ASSASSINATION!, THE ODYSSEY!, PEOPLE FROM THE SEA!, JUMP BACK BLACK CAT (GIMME THAT ROCK & ROLL)!, THE PACT!, NEW MUSIC!, ROCK AND ROLL!, ROCKING CAT!, RED HOT CHILI PEPPERS!, PAULA ABDUL!, SKA PUNK VERSION!, A CAPELLA!, VOCALS ISOLATED!, BETTER MUSIC!, ETHOS!, NOISE!, SCREAMING!, LAYERS!, AWFUL!, GUITAR SOLO!, FREEBIRD!, BIRTHDAY BASH!, SKA PUNK!, STUDIOS!, PROMOTION!, PHIL THE ARTIST!, SHREDDING!, GOOGLE ASSHAT!, NAME NAMES!, HDM!, ANNA!, HDM'S BIRTHDAY!You can find the videos from this episode at our Discord RIGHT HERE!

Agne Malisauskiene joins Payman as the clinical lead at the Baltic Academy of Aesthetic Dentistry and one of the world's best-known minimally invasive composite teachers, having trained thousands of dentists from her purpose-built facility in Vilnius.She talks candidly about swapping punishing NHS-style hours in Lithuania for a much gentler introduction to UK private dentistry, and how a year at a US military boarding school as a teenager shaped the discipline and leadership instincts she still relies on today.This is also a conversation about ambition and what it costs. From the competitive streak that pushed her toward the hardest degree to get into, to the guilt and logistics of teaching abroad most weekends while raising two young boys, Agne is refreshingly honest about the trade-offs behind a career built on repetition, pattern recognition and constant travel.Expect sharp reflections on gender and ambition in dental education, why composite bonding took off early in Lithuania, and a clinical mistake that changed how she treats every tooth since.In This Episode00:00:50 – Introduction00:02:40 – Building the Vilnius teaching facility00:11:57 – Moving to the UK00:18:27 – Boarding school in America00:30:57 – Family background and choosing dentistry00:39:41 – Ambition, competitiveness and sacrifice00:46:47 – Family life and growing up in Soviet-era Lithuania00:53:12 – A typical teaching week and travel logistics01:02:32 – Joining Biomimetic01:07:33 – Gender, ambition and teaching01:15:53 – Filming through the loupe camera01:22:47 – Blackbox thinking01:26:32 – Favourite resource: Layers 201:30:27 – Fantasy dinner partyAbout Agne MalisauskieneAgne Malisauskiene is an internationally recognised minimally invasive direct composite expert and the clinical lead at the Baltic Academy of Aesthetic Dentistry in Vilnius, where she trains dentists from around the world. She is a member of Biomimetic and Enlighten's distributor in Lithuania, and splits her time between her own patients and a busy international lecturing schedule.

What if the most valuable business intelligence in your company is sitting on the shop floor, earning $25 an hour? In this episode, Sean sits down with Dustin Snyder, a human systems consultant who helps founders and CEOs understand what's actually happening on the ground inside their organizations. Dustin explains how leaders often design broken systems without realizing it, and why the frontline workers they rarely talk to hold insights that no spreadsheet can capture. You'll also hear why delivering that truth to founders requires as much emotional intelligence as it does hard data, especially when the leader's own decisions are part of the problem. If you've ever assumed you had a clear picture of your business just because the numbers looked right, this conversation will challenge that assumption entirely.──────────────────────────────────────────────────────

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

Welcome back to today's Friday Review where I'll be breaking down the best of the week!     I'll be sharing specifics on these topics:     Red Light Toothbrush (product review) Healing Happens in Layers (tip of the week) Alcohol & Brain Health (research) Lion's Breath Breathing (research)   For all the details tune in to today's Cabral Concept 3829 – Enjoy the show and let me know what you thought!   - - - For Everything Mentioned In Today's Show: StephenCabral.com/3829 - - - Get a FREE Copy of Dr. Cabral's Book: The Rain Barrel Effect - - - Join the Community & Get Your Questions Answered: CabralSupportGroup.com - - - Dr. Cabral's Most Popular At-Home Lab Tests: > Complete Minerals & Metals Test (Test for mineral imbalances & heavy metal toxicity) - - - > Complete Candida, Metabolic & Vitamins Test (Test for 75 biomarkers including yeast & bacterial gut overgrowth, as well as vitamin levels) - - - > Complete Stress, Mood & Metabolism Test (Discover your complete thyroid, adrenal, hormone, vitamin D & insulin levels) - - - > Complete Food Sensitivity Test (Find out your hidden food sensitivities) - - - > Complete Omega-3 & Inflammation Test (Discover your levels of inflammation related to your omega-6 to omega-3 levels) - - - Get Your Question Answered On An Upcoming HouseCall: StephenCabral.com/askcabral - - - Would You Take 30 Seconds To Rate & Review The Cabral Concept? The best way to help me spread our mission of true natural health is to pass on the good word, and I read and appreciate every review!  

Battler's Game to prove he understand's Beatrice's truth is almost complete. Surely we'll get some answers now! Wait, why is there a wedding? Wait, why is Ange back? Wait, who the hell is Featherine? RYUKISHIIIIIIII!!!!!!

Healing and Wellness: Building Your Body's Foundations In this episode, Cher explores the crucial principles behind healing, emphasizing that health is a process rooted in foundational support and stewardship. She uses a house analogy to illustrate how understanding layers and layers of the body's systems—and addressing root causes—can lead to sustainable wellness. Cher encourages a holistic view, highlighting that symptoms are alerts and that healing occurs gradually through proper layers, data, and understanding.   Key topics covered: The importance of starting with the foundation: level ground and balance before building How symptoms signal deeper root causes, not just surface issues The interconnectedness of body systems and why a holistic approach is vital The role of testing and data in understanding your body's needs Why quick fixes and extreme diets often miss the root cause The process of healing as layered, requiring patience and proper groundwork The difference between conventional and holistic perspectives on health The significance of personalizing health strategies for individual needs Trusting the body's innate ability to heal when supported properly God's design of the body as a miracle that can heal itself with proper care Practical steps: paying attention to signs, reviewing lifestyle, and collaborating with health professionals Timestamps: 00:00 - Using the house analogy to understand foundational health 02:00 - Roots of imbalance: lifestyle factors and emotions influence health 04:00 - The gradual nature of healing and recognizing small progress 06:00 - Layers of the body: understanding the importance of foundational layers like insulation and electrical wiring 08:00 - The significance of testing and data for individual health insights 10:00 - Why quick fixes fail without foundational support 12:00 - The body's natural healing processes and God's design 14:00 - The importance of integrating conventional and holistic care 16:00 - Recognizing that symptoms are warnings, not enemies 18:00 - Building health in stages, like house construction 20:00 - God's purpose for our bodies and the role of stewardship    Resources & Links: Ephesians 2:10 Style and Stewardship YouTube Channel Start the Application Process for personalized health support   Connect with Cher: Instagram Facebook  

I want to talk about what took Leonardo da Vinci 16 years and at least 30 layers of paint to create, and what that has to do with the life you're building right now. The Mona Lisa is not famous despite its layers of mistakes and revisions. It is famous because of them. In this episode, I break down why the struggles, failures, and seasons where nothing seems to work are not obstacles to your legacy. They are the actual material your legacy is made from. If your goal is to be legendary, you cannot afford to keep treating your hardest chapters like something to bury. Key Takeaways Greatness takes time. Sacrificing legacy for immediacy is one of the most costly trades you can make. Your mistakes and setbacks are not liabilities. They are layers that create a texture nobody else can replicate. The biggest smiles are built on the most tears. What you have been through gives you a perspective and depth that others cannot copy. Your current difficult season is not your final form. It is the foundation being built beneath what the world will eventually see. Do not bury your failures. Own them, because nobody can take your lived experience from you. Action Steps Identify one past failure or painful experience and write down one specific insight or strength it gave you that you could not have gained any other way. Audit where you are rushing something important for the sake of completion rather than letting it become great. Commit to the longer timeline. The next time something does not work out, say out loud: 'This is a layer, not a loss.' Interrupt the habit of treating setbacks as reasons to quit. Notable Quote I love the mistakes because nobody can take them from me. I love overcoming the failures because nobody can take it from you.

Send us Fan MailThree things every investor needs to hear. First: a banker flags the growing SPV-on-SPV problem in today's VC market — 24-hour close emails, unverifiable shares, unvested stock — and explains why reading the actual paperwork has never mattered more. Second: on the PE side, capital scarcity means investors are winning more board seats, tighter terms, and preferential return structures. Third — and most shareable — a serial entrepreneur delivers a direct warning: AI is a tsunami, the window to get ahead of it is five years or less, and the operators who integrate it first will eat everyone else's lunch across every industry. His point: this is arguably the world's biggest arbitrage opportunity, and most people are still asleep.About Family Office ClubThe world's largest investor club in the family office space. 19 years. 300+ events. 16 million members. $1B+ in community transactions.

A £2 million business and a £2 billion business shouldn't be running the same AI Search strategy. In this executive briefing, Charlie Marchant (CEO of Exposure Ninja) breaks down the Pyramid of AI Search, a layered framework for understanding exactly where your brand sits and what to build next to show up in ChatGPT and other AI platforms.Why there's no single "correct" AI Search strategy, and how your starting point depends on your business sizeThe foundation layer: technical SEO, site health, and getting schema right across 100+ typesThe middle layer: which on-site pages (About, comparison, product) tend to get referenced in AI answers, and how to shape their messagingThe tip of the pyramid: third-party sites like Trustpilot, G2, Reddit, and trade publications, and how to reverse engineer which ones AI is already citing about youGetting cited by AI platforms isn't about ticking one box. It's about building a consistent, accurate brand story that holds up from your own website through to what customers and third parties say about you elsewhere.Future-proof your brand with our AI Search Audit

Originally aired February 25, 2025 "We all walk around in our little cultural bubble, and sometimes, without realizing it, that bubble shapes how we show up in the playroom." – Dr. Carmen Cubillo In this thought-provoking episode, Lisa sits down with Dr. Carmen Cubillo, a clinical psychologist, Registered Play Therapist, and cultural advocate based in Australia, for an essential conversation about culture—our own, our clients', and how both shape the therapeutic relationship. Together, Lisa and Carmen dive into: ✨ The therapist's cultural lens: How our personal cultural background influences the way we show up in the playroom and why it's crucial to reflect on it. ✨ Layers of cultural curiosity: Beyond heritage, how systemic and societal influences shape our clients' experiences and why we must stay open and adaptable. ✨ Cultural safety in therapy: What it means to be a culturally safe therapist and how unconscious biases can create barriers to connection. ✨ The importance of rupture and repair: Why acknowledging cultural missteps—rather than avoiding them—deepens trust and strengthens relationships. ✨ Connecting to the land: How understanding the cultural history of where you live can ground your practice and help you become more attuned to the experiences of the children and families you work with. This episode is an invitation to lean into cultural awareness, embrace the discomfort of learning, and grow into more attuned and connected therapists. With Carmen's wisdom, vulnerability, and deep experience working with Aboriginal and Torres Strait Islander communities, this conversation will challenge, inspire, and expand your perspective.

Most buyers no longer trust marketing claims at face value. They investigate, compare, and verify before they believe. In this solo episode of StrategyCast, Lori Jones explains the four layers of proof that help brands build buyer trust in an increasingly skeptical marketplace.As AI makes content easier to create, unsupported claims are losing value. Buyers now examine leadership behavior, customer experiences, thought leadership, consistency, and visible expertise before they speak with a sales team. The brands that stand out will not simply make stronger claims. They will make their credibility easier to observe.And don't forget! You can crush your marketing strategy with just a few minutes a week by signing up for the StrategyCast Newsletter. You'll receive weekly bursts of marketing tips, clips, resources, and a whole lot more. Visit https://strategycast.com/ for more details.==Let's Break It Down==00:00 Introduction and 600th episode milestone03:31 Understanding the trust shift08:44 Evolving role of proof in marketing10:00 Limitations of borrowed proof15:37 Future of Marketing Trends17:13 Developing a trust architecture framework20:01 Closing thoughts and contact info==Where You Can Find Us==Website: https://strategycast.com/Instagram: https://www.instagram.com/strategy_cast/Facebook: https://www.facebook.com/strategycast==Leave a Review==Hey there, StrategyCast fans!If you've found our tips and tricks on marketing strategies helpful in growing your business, we'd be thrilled if you could take a moment to leave us a review on Apple Podcasts. Your feedback not only supports us but also helps others discover how they can elevate their business game!

Our awards shortlist book club rolls on with a selection of books we’re much more receptive to: “A Drop of Corruption” by Robert Jackson Bennett, “The Raven Scholar” by Antonia Hodgson, and The Everlasting by Alix E. Harrow. Jason Snell with Aleen Simms, Erika Ensign, Scott McNulty, Heather Berberet and Paul Weimer.

Our awards shortlist book club rolls on with a selection of books we’re much more receptive to: “A Drop of Corruption” by Robert Jackson Bennett, “The Raven Scholar” by Antonia Hodgson, and The Everlasting by Alix E. Harrow. Jason Snell with Aleen Simms, Erika Ensign, Scott McNulty, Heather Berberet and Paul Weimer.

We've become a culture that believes every situation demands a hero and a villain. We don't know how to sit in tension anymore, but what if multiple things are true? Layers reveal different perspectives. Let's converse.

Panel with Dr Sergio Sanchez and Inventor Mike Acerra on how far AI goes to render truth and become a truth machine.Discover the truth about AI and its potential to be a truth machine. In this video, we delve into the capabilities and limitations of artificial intelligence in determining what is true and what is not. CHAPTERS00:00 Welcome to Cyber Crime Junkies + Guest Introductions02:00 Anthropic's Secret AI Mythos Finds Decade-Old Vulnerabilities04:30 How AI Is Finding Security Holes at Scale in Small Businesses06:30 New Books: Moving Target, The God Prompt, and Future Visions09:00 What American Education Is Getting Wrong11:30 Why Forgetting History Destroys Countries14:00 Nixon, the Deep State, and Political Accountability17:00 Property Taxes, Home Ownership, and the Banking Lie20:00 Tax Season Stress and Why Big Refunds Are a Bad Sign23:00 Flat Tax, the IRS, and Why the System Is Built to Confuse You26:00 The Knights Templar Invented Modern Banking29:00 AI as a Truth Machine Against Political Propaganda32:00 How Kennedy Beat Nixon on TV but Lost on Radio37:00 Hack or Hype: Employee Security Risks After Termination40:00 Hack or Hype: The Kaseya Supply Chain Ransomware Attack43:00 Hack or Hype: Does Cyber Insurance Actually Pay Out49:00 What SMBs Get Wrong When Applying for Cyber Insurance52:00 Windows Defender, Malware, and Layers of Endpoint Security Questions? Text our Studio direct. We read these and when helpful we give a special shout out for those to contact us.True crime enters our homes and businesses daily. Learn from actual people who fight it daily and show you how in a thriller story. The Moving Target Trilogy. Book 3 to be released September 22nd, 2026. Start with any of them. Be a Moving Target.Special Author pricing (30% off) The Moving Target Trilogy. Book 3 to be released September 22nd, 2026. Start with any of them. Be a Moving Target.Special Author pricing (30% off) Growth without Interruption. Get peace of mind. Stay Competitive-Get NetGain. Contact NetGain today at 844-777-6278 or reach out at DMauro@NetGainIT.com or find more at www.NETGAINIT.com   Support the showNew Exclusive Offers for our Listeners! New non-fiction Book Series is out! Moving Target: The Art of Online Camouflage drops April 14.Moving Target: The Obedient Machine drops April 21.Book 3 -- Ghost and the Machine -- out soon!

Hour 2 of The Ken Carman Show with Anthony Lima

Carina talks to fellow Riley Blake fabric designer Jill Finely. Learn how Jill works in fabric, retreats, notions, and so much more! This is a cut from an hour long interview Carina did with Jill for the University of Arts & Design's DES 610 course.Learn more about Jill at https://www.jillilystudio.com/Learn more about this course, degrees, and our new Continuing Education Program at the University of Arts & Design at www.uad.educationJoin a Design Bootcamp at www.designsuitecourses.com/designbootcamp Get my free gift to you here: https://www.designsuitecourses.com/intentional

If you have ever improved, then relapsed, or felt like you are chasing the same symptoms in circles, this episode will help explain why. Dr. Holly Carling and Alicia walk through what it actually means to ask why, not just once, but layer by layer, until you reach the core of what is driving chronic illness. Using the image of an onion, Dr. Holly explains that symptoms are the outermost layer, and that real restoration usually lives at the fourth or fifth layer down. She walks through a case example to illustrate how each layer builds on the last, and why treating only the outer layers leads to improvement followed by relapse. In this episode: Why chronic illness is multifactorial, and why stopping at the first or second layer of why so often leads to managing illness rather than resolving it How to recognize system dysfunction and deeper drivers as distinct layers requiring different support Why digestion, sleep, blood sugar stability, and nervous system regulation must be addressed before deeper drivers can safely be resolved How acupuncture regulates systems rather than symptoms, and why that makes it central to this layered approach   For full show notes, resources and links head to: https://vitalhealthcda.com/podcasts/ The Vital Health for You Podcast is for everyone. Get to know us more by connecting with us at our website or on our Facebook page.   *Disclaimer: The statements made in this episode about specific products have not been evaluated by the U.S. Food & Drug Administration and are not intended to diagnose, treat, cure or prevent disease. All information provided is for informational purposes only and is not intended as a substitute for advice from your physician or other healthcare professional.   

Enjoy this message from our NextGen Pastor Austin Young!

Game Exercise: Close your eyes and follow along with an entire Chess game using the audio below. On each move, try to conceptualize the position clearly and understand how it has changed. Try to follow the game until the end to stretch the amount of moves you can see ahead. To learn more about Don't Move Until You See It and get the free 5-day Conceptualizing Chess Series, head over to https://dontmoveuntilyousee.it/conceptualization PGN for today's exercise: d4 Nf6 2. c4 Nc6 3. d5 Ne5 4. b3 e6 5. Bb2 Bb4+ 6. Nd2 Ne4 7. Bc1 Qf6 0-1

One of the most complicated elements of the Magic rules is something known as "layers." I sit down with the man in charge of the rules, Jess Dunks, to help walk through an introduction to layers.

Episode SummaryIn this episode of the Two Piers Podcast, Erica D'Eramo welcomes back author, creative coach, and facilitator GG Renee Hill for a conversation about writing, creativity, identity, and the stories we carry.GG's latest book, Story Work: Field Notes on Self Discovery and Reclaiming Your Narrative, explores how inherited beliefs and limiting stories shape the way we see ourselves, make decisions, and move through the world. Together, Erica and GG discuss how writing can help us uncover those stories, examine whether they still serve us, and reconnect with our own voice and values.In this episode, they explore:How writing can support healing, self-understanding, and personal clarityThe stories we inherit from family, culture, and earlier experiencesHow limiting beliefs become embedded in our behaviorThe three sections of Story Work: Roots and Origins, Truth and Lies, and Voice and VisionCreativity as something much broader than traditional artistic expressionHow perfectionism and the inner critic can interfere with journaling and creative workThe value of practicing imperfection and allowing work to remain messy or unfinishedWhy the process of creating can matter more than the final resultGG also shares her own path from corporate America into writing, coaching, and facilitation. She reflects on how therapy and journaling helped her discover “the story beneath the story” and eventually led her toward work centered on creative expression, self-discovery, and community.One of the central takeaways from the episode is that writing does not need to be polished, impressive, or intended for anyone else. It can simply be a place to witness yourself more honestly.As Erica puts it: don't deprive the world of your imperfect mess.About GG Renee HillGG Renee Hill is an author, creative coach, and facilitator whose work explores writing as a pathway to wellness, personal clarity, creativity, and collective growth.She is the author of:Story Work: Field Notes on Self Discovery and Reclaiming Your NarrativeThe Self-Care Check-InA Year of Self-ReflectionGG also facilitates writing workshops, creative writing cohorts, monthly writing prompts, and free public writing sessions designed to help people reconnect with their voices in a supportive, low-pressure environment.Connect with GG Renee HillVisit All the Many Layers: https://www.allthemanylayers.com/Subscribe to GG's weekly newsletter and Substack, Writing the Layers: https://thelayers.substack.com/Follow GG on Instagram: https://www.instagram.com/ggreneewritesConnect with GG on LinkedIn: https://www.linkedin.com/in/gina-gg-renee-hill-28686762/Through All the Many Layers, you can also learn more about GG's monthly writing prompts, First Friday free-writing sessions, workshops, creative writing cohorts, and other resources for writers and anyone interested in reflective creative practice.More from Two PiersLearn more about Two Piers Consulting: https://www.twopiersconsulting.comBrowse podcast episodes, summaries, and transcripts: https://www.twopiersconsulting.com/podcastAffiliate disclosure: Some book links in this post are Bookshop.org affiliate links. If you make a purchase through one of these links, Two Piers may receive a small commission at no additional cost to you. These commissions help support the work we do at Two Piers.

Doug Schmell is back....Shakedown lightning to open...Top notch Dirty Dead throughout set one. Layers of an insane Deal gem...Help Slip Frank...A never ending DEW jam, Garcia's resilient and Brilliant...Layers

Drew Weatherhead is a Canadian bestselling author, Brazilian Jiu-Jitsu black belt, and content creator known for blending philosophical inquiry with storytelling. He has published two non-fiction books—Consciousness Reality & Purpose and Layers of Truth—that explore deep questions about human existence, truth, and perception from subjective viewpoints. His debut fiction work, Fractures in Love, launches the epic Fractures fantasy series, following a young enslaved girl who discovers hidden powers in a richly imagined world; with his new book Fractures in Peace set to debut July 17th. Cornerstone Forum 26'https://shaunnewmanpodcast.substack.com/Drew's Book: https://www.amazon.ca/dp/B0H56ZK19P?dplnkId=6bce6aac-3430-4e12-8801-f12c3510475c Silver Gold Bull Links:Website: https://silvergoldbull.ca/Email: SNP@silvergoldbull.comText Grahame: (587) 441-9100Bow Valley Credit UnionBitcoin: www.bowvalleycu.com/en/personal/investing-wealth/bitcoin-gatewayEmail: welcome@BowValleycu.com Get your voice heard: Text Shaun 587-217-8500

If you've ever poured months into building a semantic layer only to watch it become shelfware the moment the business pivoted, Jacob Matson has some thoughts. And a metaphor. Your data is a jungle—and a semantic layer is a highway. Great if you need to get somewhere fast and reliably (monthly active users: highway, please). But the interesting business questions? The slicing, the dicing, the nuanced dimensions that actually differentiate your company from its competitors? There's no highway for that. There never will be. Jacob, a developer advocate at MotherDuck with deep roots in accounting and ERP systems, joined Michael, Moe, and Julie to talk through what comes after the semantic layer—or at least alongside it. The conversation covered why the most important parts of any business are precisely the parts that resist being modeled in someone else's framework, why AI is actually pretty good at writing SQL but not so great at remembering what it figured out yesterday, and whether the real job to be done here is less about modeling and more about search. Oh, and the uncomfortable truth that at episode 300, we still don't have a great answer for metric drift. But we've got some really good questions. For complete show notes, including links to items mentioned in this episode and a transcript of the show, visit the show page.

-#RuffinosRants -LSU Playoff Picture | Toughest SEC Schedule? -LSU Recruiting Update | What To Know -#AskBlake Powered By @247Sports | @Geaux247

When a child struggles with swallowing, chewing, or food transitions, our first instinct is often to look directly at traditional feeding strategies. But what if the missing piece of the puzzle isn't the food itself, but the foundational resting posture and function of the orofacial muscles?In this solo episode, Hallie Bulkin demystifies myofunctional therapy (Myo) and explores its critical, undeniable overlap with pediatric feeding therapy. She breaks down how addressing underlying myofunctional dysfunction can drastically accelerate your clinical progress, protect airway safety, and create long-term, sustainable outcomes for the children on your caseload.Hallie addresses common misconceptions surrounding Myo, discusses structural considerations like tongue-ties, and explains why a whole-system approach—looking at tongue posture, breathing, and body alignment—is non-negotiable. If you're ready to stop looking at oral motor function in a vacuum and want practical steps to seamlessly weave myofunctional thinking into your next feeding evaluation, this episode is exactly what you need.Key Topics & TakeawaysDefining the Scope of Myo: Understanding what myofunctional therapy actually is and how it targets the resting postures and functions of the oral and facial muscles.The Perfect Partners: Why feeding therapy and Myo should never be treated as entirely separate disciplines, but rather as deeply interconnected systems that support one another.The Trifecta of Function: Exploring how tongue resting posture, nasal breathing, and physical body posture directly dictate a child's success with chewing and safe swallowing mechanics.Debunking Common Misconceptions: Shedding light on the myths surrounding myofunctional therapy and highlighting the evidence-based research that supports its clinical efficacy.Integrating the Assessment: Practical, realistic steps to incorporate orofacial muscle function and structural considerations (like tongue-ties) into your standard feeding evaluations without blowing your timeline.Soundbites"Feeding and Myo are partners, not separate disciplines. When you treat them as a connected system, your outcomes transform.""Addressing myofunctional dysfunction speeds up feeding progress. We cannot build functional feeding skills on top of poor oral resting postures.""Myo literacy makes you a better clinician in any specialty. It completely shifts the lens through which you analyze a child's struggles."Timestamps00:02:29 | Defining Myofunctional Therapy00:03:32 | The Root Cause vs. Symptom Lens00:07:09 | Breaking Through Feeding Plateaus00:11:56 | Where Feeding and Myo Overlap00:14:41 | Airway Management & Nasal Breathing00:18:12 | Debunking the "Just Exercises" Myth00:23:54 | How to Run a Myo Assessment00:30:12 | The 5-Step Integration Framework00:33:33 | The Connected Child SystemLinks & ResourcesClinical Tool: Streamline your assessments and screen for muscle dysfunction F.A.S.T. MYO SCREENING PACKET: Need a simple & science-backed way to screen your patients for potential orofacial myofunctional disorders?WORTH A LISTEN: CONTINUE YOUR JOURNEYThe 4 Layers of Feeding: How to Finally Know Where to StartWhen You Screen a Child and Think 'Now What?STAY CONNECTED

The early modern period was a rocky time for religion across Europe, fundamental pinnings of everyday life were being questioned and changing, as the reformation shifted authority away from the church. Thoughts on death and the afterlife were turned upside down, and in villages across England, restless spirits were making a comeback. To confront these supernatural intruders, communities turned to their local clergy to carry out the practice of ghost-laying, an exorcism ritual that sought to bind and banish the dead. At least, that's what Victorian authors would have liked to believe. Part religion, part folklore, and part fear, the subject of ghost-laying is fairly well complicated, and defining just how much of it is actually true, even more so. SOURCES Walsh, Brendan C. (2023) ‘He Could Raise and Lay Ghosts at His Will': Victorian Folklorists and the Creation of Early Modern Clerical Ghost-Laying. Folklore, 134:3, 281-303, DOI: 10.1080/0015587X.2023.2187157 Bottrell, William (1870) Traditions & Hearthside Stories of West Cornwall. Beare & Son, Penzance, UK. Hawker, Robert Steven (1870) Footprints of Former Men in Far Cornwall. James C. Commin, Exeter, UK. Andrews, William (1898) The Church Treasury of History, Custom, Folk-Lore, etc. London, UK. Bond, Thomas (1823) Topographical and Historical Sketches of the Boroughs of East and West Looe, in the County of Cornwall. J. Nichols & Son, London, UK. Hunt, Robert (1865) Popular Romances of the West of England, Or, The Drolls, Traditions and Superstitions of Old Cornwall. John Camden Hotten, London, UK. Courtney, Margaret Anne (1973) Cornish Feasts & Folklore. EP Publishing Ltd. UK. ------ For almost anything, head over to the podcasts hub at ⁠⁠⁠⁠⁠⁠⁠darkhistories.com ⁠⁠⁠⁠⁠⁠⁠ Support the show by visiting our Patreon for bonus episodes and Early Access: ⁠⁠⁠⁠⁠⁠⁠https://www.patreon.com/darkhistories⁠⁠⁠⁠⁠⁠⁠ The Dark Histories books are available to buy here: ⁠⁠⁠⁠⁠⁠⁠http://author.to/darkhistories⁠⁠⁠⁠⁠⁠⁠ Dark Histories merch is available here: ⁠⁠⁠⁠⁠⁠⁠https://bit.ly/3GChjk9⁠⁠⁠⁠⁠⁠⁠ Connect with us on Facebook: ⁠⁠⁠⁠⁠⁠⁠http://facebook.com/darkhistoriespodcast⁠⁠⁠⁠⁠⁠⁠ Or find us on Twitter: ⁠⁠⁠⁠⁠⁠⁠http://twitter.com/darkhistories⁠⁠⁠⁠⁠⁠⁠ & Instagram: ⁠⁠⁠⁠⁠⁠⁠https://www.instagram.com/dark_histories/⁠⁠⁠⁠⁠⁠⁠ Or you can contact us directly via email at ⁠⁠⁠⁠⁠⁠⁠contact@darkhistories.com⁠⁠⁠⁠⁠⁠⁠ or join our Discord community: ⁠⁠⁠⁠⁠⁠⁠https://discord.gg/cmGcBFf⁠⁠⁠⁠⁠⁠⁠ The Dark Histories Butterfly was drawn by Courtney, who you can find on Instagram @bewildereye Music was recorded by me © Ben Cutmore 2017 Other Outro music was Paul Whiteman & his orchestra with Mildred Bailey - All of me (1931). It's out of copyright now, but if you're interested, that was that. Learn more about your ad choices. Visit megaphone.fm/adchoices

How is trauma stored in the body?Josh Trent welcomes double board-certified physician, Dr. Aimie Apigian, to the Wellness + Wisdom Podcast, episode 820, to reveal the connection between neuroscience, functional medicine, and attachment theory, how trauma becomes a physiological pattern stored in the body, and why healing trauma can become easy when we understand how it forms.

What's going on, Household! Are you itching for more Summer House Reunion discussion? Well, you're in luck. In case you missed it, here's a snippet of what we discussed on Pop Extra last week. Listen to the full episode by subscribing to Viall Files +, via Supporting Cast. Sign up today at ViallFiles.com.