Podcasts about ai machine learning

פרק מספר 492 של רברס עם פלטפורמה, שהוקלט ב-20 בפברואר 2025. רן מארח ב-Remote את עמית מהבנק הדיגיטלי One-Zero כדי לדבר על פרויקט של בוט חדש - Ella - שמפותח בחברה.

Send us a textIn this episode of The Digital Executive, Shashank Kapadia, a leader in machine learning, shares insights into how AI-driven personalization is transforming industries. Drawing from his experience with companies like Walmart and Randstad, he discusses the challenges of deploying large-scale AI solutions, the myths of real-time AI, and the critical role of ethical and transparent machine learning systems. His expertise in NLP and domain-specific AI applications further highlights the evolving landscape of AI-driven decision-making.Kapadia also explores key emerging trends, including the democratization of foundational models, the rise of edge learning, and the importance of ethical AI in business strategy. His forward-thinking approach offers valuable takeaways for leaders navigating AI adoption. Tune in to gain expert insights into the future of machine learning and its impact on enterprises worldwide.Disclaimer:Shashank Kapadia's comments and opinions are provided in their personal capacity and not as a representative of Walmart.  They do not reflect the views of Walmart and are not endorsed by Walmart.

We've all been there. That unsolicited, lengthy WhatsApp message from a "consultant" offering a laundry list of services or training programs. You scroll, you sigh, and you likely ignore or block the sender. It's a common occurrence, and it's a perfect example of how not to do outreach. Let's dissect a recent outreach attempt I received, and why it failed miserably: The Offending Message: "Good Day, trust your day is going great? My name is XYZ, an IT Consultant at TECHXYZ AFRICA. I am inviting you and your loved ones to join our internship training program, designed to equip you with the most global in-demand tech skills like Data Science, Web Development, Cybersecurity, AI & Machine Learning, Digital Marketing, Product Design (UI/UX), Video Production, Networking (CCNA), Mobile App Development, Cloud Computing (AWS), Music Production, and more! Our training program comes with; ✅ international certifications ✅ International Job Opportunities ✅ Internship Opportunities ✅ Free CV revamp ✅ Job interview preparation Let me know which course you're interested in. Call or WhatsApp 090XYZ to kick-start your journey today! We have both online and Physical classes Please Don't keep this to yourself, share it with friends & family!" Why This Approach Fails: * Too Long: In the mobile-first world, lengthy messages are a death sentence. People simply won't scroll through endless paragraphs. * Unsolicited: This message came out of the blue. I had no prior interaction or expressed interest in any of these tech areas. Unsolicited messages are often perceived as spam and lead to blocking. * Trying to Be Everything to Everyone: Listing every possible tech skill dilutes the message. It lacks focus and fails to resonate with any specific need. * Confusing Call to Actions (CTAs): The message throws multiple CTAs at the recipient: choose a course, share with friends, inquire about benefits. This confusion leads to inaction. * Lack of Personalization: The generic greeting and impersonal approach make it clear this is a mass-sent message. The Right Way to Do WhatsApp Outreach: Instead of broadcasting a novel, adopt a more strategic, conversational approach: * Start with a Problem-Based Opener: Instead of a generic greeting, open with a question that addresses a specific need. For example, instead of the long message, a good first message would be something like, "Good Day (name), I got your number from XYZ business directory. Are you or any of your loved ones interested in acquiring tech skills?". * Qualify and Engage: If they respond, engage in a conversation to understand their specific interests and needs. Ask questions about their goals and challenges. * Focus on One Area: Instead of listing every service, focus on the one that aligns with their needs. * Build Rapport: Prioritize building trust and rapport before pitching your services. * Offer Value: Highlight the benefits that matter most to them, such as certifications, job opportunities, or skill development. * Use a Clear and Singular CTA: Guide the recipient to take one specific action, such as scheduling a call or visiting a website. The Takeaway: In today's digital age, it's tempting to rely on mass messaging. However, effective outreach requires a personalized, strategic approach. Focus on building relationships, understanding needs, and providing value. Stop trying to do what's easy, and begin doing what is effective. Remember, building trust and rapport leads to higher conversion rates, repeat business, and valuable referrals. It's worth the effort. #WhatsAppOutreach #SalesTips #DigitalMarketing #BusinessStrategy #LeadGeneration

Technica specializes in AI, machine learning, and cybersecurity, shares Steve Hatch, Head of Human Resources and Communications. Learn from Steve's interviewing tips including being confident in your skills, doing your homework, and showing interest and enthusiasm. And it's another shout-out for the STAR method on your resume and when interviewing, from Steve. Want to learn how to do that well? Listen and learn!3:46 Technica's Innovations Lab has engineers, software developers, and scientists looking at potential issues or challenges we'll be facing in the future.10:55 Interviewing tips. Understand the company. Critical factor in assessing someone for a position – if they put in an effort to learn and research the company before the interview.14:32 Use the STAR method – it's relevant for interviews and resumes. Situation, Task, Action and Result.Find complete show notes at: https://clearedjobs.net/technica-ai-machine-learning-and-cyber-podcast/_ This show is brought to you by ClearedJobs.Net. Have feedback or questions for us? Email us at rriggins@clearedjobs.net. Sign up for our cleared job seeker newsletter. Create a cleared job seeker profile on ClearedJobs.Net. Engage with us on LinkedIn, Facebook, Instagram, X, or YouTube. _

In this episode, recorded during the IPPE show in Atlanta in January, Juan DeVillena, senior vice president of quality assurance and food safety at Wayne-Sanderson Farms, walks through Wayne-Sanderson's digitalization journey and how the company is working to implement artificial intelligence (AI) and machine learning today. DeVillena also discusses how the chicken processor rolled out its training and helped its employees work through the process. He also hints at how AI and machine learning could help the company's food safety and quality efforts down the road.

Have you considered living outside the US? It feels like wherever we are, that's where we have to stay but your dream life could be someplace else and you just have to get to it.Well in this chat we have Sonaya Williams here to teach us what it really takes to move and live outside the US. Sonaya has the experience and is here to share her wisdom for not just singles but for whole families who want to relocate!As you've heard me say, it's free to find out! So jump into this chat to learn from Sonaya.Thanks for being here today! Find Sonaya on Substack at https://www.yourexpatlife.com/ and head over to Instagram https://instagram.com/nicolewaltersso we can hang out!Watch this chat (and others!) on YouTube at https://nicolewalters.com/youtubeGet the rest of the links and resources from this episode at https://nicolewalters.com/episode451Episode Sponsors: Get 20% off your first order with code DREAM20 at nanit.com.Use code NICOLE at checkout for 15% off your entire order at www.vionicshoes.com when you log into your account. 1 time use only.Download the Earnin app and type in Nicole Walters under PODCAST when you sign - it'll really help the show.Download the CFO's Guide to AI Machine Learning at NetSuite.com/NICOLE.To listen to MissUnderstood: The ADHD in Women Channel, just search for MisUnderstood in you podcast app.See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

Venture capital funding into Irish technology SMEs treaded water in the third quarter of this year, finishing at €192.5m, just 1% ahead of the same period last year, according to the Irish Venture Capital Association Venture Pulse survey, published today in association with William Fry. This compares to a global trend where VC funding in the third quarter fell by 15% year over year. VC funding in Ireland for the nine months to the end of September fell by 18% to €945.3m from €1.1bn in 2023. Globally, VC funding fell by 3.7% over the same period. International VC funding into Ireland fell by almost €200m or 24% for the year to the end of September. International funding for the third quarter of €85m amounted to 44% of the total for the period. Gerry Maguire, chairperson of the Irish Venture Capital Association, commented: "A positive feature of the third quarter is that the number of deals rose by 55% to 59. While there were no rounds in the €30m plus range, this was compensated for by an increase in deals under €10m, including seed funding. These underlying trends reflect a buoyant ecosystem in Ireland for early-stage companies, many of which are involved in cutting-edge technologies such as Artificial Intelligence (AI), Cyber Security, Quantum Computing, MedTech and Envirotech. "However, the potential for an isolationist-leaning incoming administration in the White House to implement policies that could be negative for Ireland emphasises the need for us to take these Indigenous companies to the next level and grow our own tech champions rather than rely on less committed overseas investment." Sarah-Jane Larkin, director general of the Irish Venture Capital Association, said that while there were no deals over €30m in this quarter, there was an encouraging bounce in smaller rounds. Deals in the €3-€5m range rose by 52% to €29m. €1m-€3m rounds grew by 16% to €26m, and deals under €1m rose by nearly a third to €9m. Seed funding, or first rounds raised by SMEs, more than doubled to €33.5m. Commenting on the opportunities for an incoming Government to enhance Ireland's competitiveness in an increasingly uncertain global market, Ms Larkin welcomed the small business provisions in a number of party manifestos, which emphasised the need for scaling up funding and, in particular, the backing of pension funds in this regard. She said that while the support for early-stage firms was a positive feature of the Irish economy, a warning signal was that the average deal size in Ireland has fallen since 2018 compared to comparative European countries. "Drip feeding of small rounds is not efficient. As a result, the proportion of Irish firms going out of business is higher than the European average." The largest deals in the third quarter were medtech company Neurent Medical (€18.2m); renewable energy firm Circal (€15m); EV and solar installation provider ePower (€15m); medical device company Luminate (€13.9m) and medtech firm, Loci Orthopaedics (€13.8m). Sarah-Jane Larkin added: "Access to public capital for VCs through the likes of EI (Enterprise Ireland) and ISIF (Ireland Strategic Investment Fund) in Ireland and EIF (European Investment Fund in Europe, has never been better. However, there is now a severe shortage of matching private capital in Ireland from investors such as pension funds, family offices and corporates. Government policy can have a significant impact in fixing this, as is happening in the UK, and in several EU countries such as Denmark and France." Life sciences were the leading sector in the nine months to the end of September raising 42% of the total (€392.8m), followed by Envirotech (13%), Regtech (10%), Fintech and Software (both 9%). In the third quarter, AI & Machine Learning was the third highest sector raising 11%.

Join us for an insightful episode with Carla Wade, Chief Revenue Officer at Lotlinx, to discuss the current challenges facing car dealerships in our "State of the Union" segment. With dealers scrambling to offload bloated inventories, we delve into the evolving landscape of automotive sales where traditional advertising methods are becoming obsolete. As consumers spend 13-15 hours making vehicle decisions over a span of 1-3 months, inflation and rising interest rates have made buyers more cautious, often driving them over 50 miles to find the right deal. Carla highlights the pressing need for dealers to enhance their online presence, emphasizing the importance of VIN-level marketing and effective dealership website management. In this episode, Carla breaks down the critical differences between AI and machine learning, explaining how each can be leveraged to make smarter inventory management decisions. She discusses how Lotlinx stands at the forefront of this technological evolution, guiding dealerships to choose the right tools that not only streamline operations but also provide valuable insights into market dynamics. Discover how improved vehicle descriptions, competitive pricing, and high-quality photos can attract buyers in today's market. Carla shares insights on using AI and machine learning to optimize pricing strategies and enhance inventory visibility, allowing dealerships to make informed decisions without the overwhelming manual effort. Learn how technology can help pinpoint low-funnel buyers and ensure that vehicles are seen by the right customers at the right time. Carla emphasizes the importance of understanding specific dealership needs to effectively harness data, leading to better decision-making and increased profitability. With practical metrics and a simple checklist, she breaks down how dealers can maximize efficiency and profitability while reducing the need for discounts. Tune in to learn how to navigate the complexities of today's automotive market and harness the power of Lotlinx's expertise to reclaim your time and boost your bottom line! Dealer Talk with Jen Suzuki Podcast |Jennifer@edealersolution.com | 800-625-1590 | edealersolutions.com  

Accelerate your post-production career: https://mixinglight.comFull episode notes and additional links: https://mixinglight.com/color-grading-tutorials/colorist-jason-bowdach-pixeltools-color-timer-podcast/Buy the Color Timer Shirt, now for sale: https://vincenttaylorcolor.myshopify.com/===Today, I will be speaking to Jason Bowdach. Jason is a fellow contributor here on Mixing Light. He is a colorist and online editor with an impressive CV that includes major networks such as Fox and Disney. Jason creates a suite of colorist tools through his company, PixelTools. Jason shares what inspires him to create software offerings for the industry and his approach to a new project. He also has an interesting view on AI / Machine Learning and its likely impact on our industry.This is putting the shoe on the other foot at last. Jason has his very cool podcast, Color & Coffee, on which I was a guest, so now it's time for Jason to answer some of my questions! Outside of Pixel Tools and his own grading work, Jason also teaches about color and post-workflow, so he keeps fairly busy, to say the least.- - -Editor: Rich RoddmanExecutive Producer: https://mixinglight.comPodcast Home: https://colortimerpodcast.mixinglight.com (00:00) - - Introduction (02:29) - - Starting the Timer: What is Pixel Tools? (05:27) - - How LUTs inspired PixelTools' first product (08:08) - - How do you reverse engineering a LUT? (11:36) - - How Jason got into DCTLs (14:53) - - Exploring filmic contrast (18:20) - - What is Jason's take on AI? (20:23) - - Closing remarks

Neil Leyland is a Chief Contact Center Strategist at InterVision, and he is an exceptional PMP and six Sigma black belt certified senior-level program leader and a proven problem-solver. A highly motivated achiever with a career history in sales and operations management for international multi-unit retail operations.  Possessing excellent interpersonal, presentation, written and verbal communication skills, which are used to solve problems, consult on technology projects and develop long-term collaborative relationships.  Questions · Could you share with our listeners a little bit about your journey, how you got from where you were to where you are today. · Could you tell our listeners a little bit about InterVision, what they do and what your role at InterVision is? · When you say holistic approach using AI and machine learning, can you explain to us what that means? Is it that robots are going to replace human beings? Or are you looking more from a support side, just tell us how it is that you actually see it working? · Could you share with our listeners, what's the one online resource, tool, website or application that you absolutely can't live without in your business? · Could you also share with our listeners maybe one or two books that you've read that have had a great impact on you, it could be a book that you read a very long time ago, or even one that you've read recently that has impacted you either professionally or personally. · Can you also share with our listeners what's the one thing that's going on in your life right now that you are really excited about, either something you're working on to develop yourself or your people. · Where can listeners find you online? · Now, before we wrap our episodes up, Neil, we always like to ask our guests, do you have a quote or a saying that during times of adversity or challenge, you will tend to revert to this quote if for any reason you get derailed or you get off track, the quote kind of helps to get you back on track. Do you have one of those? Highlights Neil's Journey Me: Neil, could you share with our listeners a little bit about your journey, how you got from where you were to where you are today.   Neil stated that you can probably tell, he's not from the US, although he's based there now. Finishing up from university, he went straight into working for enterprise, Rent a Car, did probably 20 or so years with them. Worked from the main counter at a rental office all the way through to leading an area. Then he moved countries during that time and started working more on the technology side and being far more strategic as opposed to tactical.   Then settled in St. Louis, got married, had kids, and then moved through several different companies, picking up good and bad practices along the way, and he's ended up at just coming up to 12 months now with InterVision.   About InterVision Me: Now, could you tell our listeners a little bit about InterVision, what they do and what your role at InterVision is?   Neil shared InterVision is an AWS premier tier partner. They specialize in transforming contact centers to their flagship product, Connective CX, powered by Amazon Connect. They also integrate AI to deliver seamless omnichannel engagements. They address common pain points for call centers, like reducing call volume and the costs associated, they do this through improve engaging efficiency and also enhance customer satisfaction. They can be found at www.intervision.com.  His role within InterVision, he would say is a Contact Center Evangelist. So, he works with clients and look at problems that they face, and then help them find what is either the best operational or technological solution to best satisfy customers' needs or solve problems that the companies have been satisfying those needs, so tying together his history working in retail, as well as time and technology, and sort of blending the two, to give what people consider to be a best in class solution for them.   Understanding the Approach in AI and Machine Learning in the Contact Centre Me: So, you are in the contact center space, and your strategy is to ensure that you have a holistic approach using AI and machine learning, two very popular words that are being used very frequently in the CX space. When you say holistic approach using AI and machine learning, can you explain to us what that means? Is it that robots are going to replace human beings? Or are you looking more from a support side, just tell us how it is that you actually see it working?   Neil stated that he thinks it's actually good to approach it from a journey perspective, if you will. So, if you think about somebody that has a transaction, whether that be online or in person, and then they need some level of support. So, they come through to a contact center, and at that point, contact centers have really embraced AI and machine learning to help customers come through and get a better level of experience. So, whether that be at the starting point when they answer the phone, you can have chatbots either on the website or on the IVR that are able to answer and interact with customers and provide them with quick hit answers and potentially resolve problems for them quickly and efficiently. Now that's one use of AI.  People say, well, is that going to replace people?  He doesn't think it does, because it solves the simple problems AI and ML doesn't have the ability to solve.  So, when people do get to an agent or somebody on the phone, or whether a chat or send an email and get a reply, the agents are able to spend that little bit more time to solve a problem, so it elevates the customer experience even though it's not necessarily AI based.  When you think about that side though, you get AI does weave its way in there and provide agents with the ability to serve customers or call us better. So, you get crazy things like, there are AI tools out there now that will listen to the conversation, will understand the context of it, understand the ask and serve up knowledge or information to the agent real time, so they can better solve the problem.  So, it will literally know this customer sounds like they have a question about x, here's the most common answers to x, is this the right thing to say and serve that up to the agent. So, the agents might not have any real experience of the problem, but they've got a proven history of other people being able to solve that question, or a very similar question, quickly and efficiently, and they can copy it. And then that really helps agents appear to be more efficient, more friendly, and for everybody that's listening, and everyone's been put on hold. No one likes to be put on hold, or “I don't know the answer to that, let me transfer you”, that can go away, which is really, really profound and gives a perception of quality well and above the norm.  And then the other side of it that's kind of cool, is you can have sentiment monitoring. So, if somebody's listening to this call, the AI or ML in the background will be monitoring it, and they can flag calls to supervisors or to other people to say, “Hey, Neil's really happy with this call. Neil's unhappy with this call. We might need some help. Somebody may need to join this call because Neil's struggling with it.”  So, it basically not only gives the ability to empower people and have them answer questions well, but it gives them monitoring so that people can actually get involved and engaged and help customers that have got problems and prevent issues, if that makes sense.   Me: Yes, it absolutely does. I attended a conference, I think it was the first and second of May, hosted by a company called CX Outsourcers Mindshare. They brought together, I believe, close to 80 persons from all over the world, from all different continents, that were in the contact center space. And my role at the event was, I sat on a podcast panel with a podcaster from Brazil and one from South Africa talking about the influence of podcasting on customer experience and the impact that it will have in the contact center space.  One of the things that I found fascinating at the conference, and this was predominantly I believe in, I know in the Caribbean for sure, and definitely in Africa, and you can let me know what your feedback is based on your exposure and experience that hiring, in terms of recruitment was a big issue that they were facing in the contact centers and trying to integrate AI and more importantly, as it relates to recruitment, ensuring that as they go forward and AI is more integrated into the whole process of solving customers problems, having AI do the more simplistic activities and tasks, and then having the agents do more complex tasks. Is that a trend that you've seen happening? Or is there anything else that you'd like to add to that conversation?    Neil shared that it definitely is. It's causing an upskilling, or an appearance of upskilling of people that answer the phone or answer the chats. Because when he started in the contact center space, companies would train an agent for 4, 6, 8 weeks, maybe even more to make sure that they have the skill set and the knowledge to be able to answer not just 80%. Neil shared that Yanique is absolutely correct that AI and machine learning is having a profound impact on the agents and upskilling, because the ability for machines to take away the simpler tasks means that agents can do things that are more interesting and rewarding for one. So, that makes the job more fun, that's an important component.   The other side is, years ago, as companies brought people on board, they'd spend weeks and weeks training them, and nowadays that's just not necessary, because most companies have invested, or are looking to invest in a single pain so all of the information is shared to them, and when that's augmented by machine learning to provide sensible text or answers or knowledge that's appropriate, agents appear to be more knowledgeable with less training, faster and that gives a great different for companies that are embracing it. It's a great differentiate. It really helps the agents feel valued, enjoy the job, and therefore more likely to be retained and that skill level is retained.  So, generally, companies that companies that retain the skill gets better over time, and it also the other piece is, he thinks it helps companies attract people, because the job is more fun and more rewarding. So, the benefits not only in the people that work there, it's about getting the better talent in the front door as well.   App, Website or Tool that Neil Absolutely Can't Live Without in His Business When asked about online resources that he can't live without in his business, Neil shared that for him personally, and this is going to sound a little bit old school, he absolutely loves using YouTube, and he will go visit YouTube looking for how to build a presentation, looking for information, ways to do things, learnings, classes, he finds a great value with day to day, he's looking at YouTube and watching videos on lots and lots of topics continually, because he thinks it's a quick and easy way to learn how to do something new or refine what he's doing based on somebody else's best practices, whether that be consultants that have classes on how to do PowerPoint presentations, or even people that do public speaking regularly and share tips and trades on how they do it. So, he uses YouTube a lot, and slowly but surely, he thinks that's starting to be replaced a little bit by TikTok, because he likes 60 second bites as opposed to 20-minute videos.   Books that Have Had the Biggest Impact on Neil  When asked about books that have had an impact, Neil shared that books that he really enjoyed and got a lot out of is a book called Good to Great: Why Some Companies Make the Leap…And Others Don't by Jim Collins, it's an older book now, but it's been around for a good few years that definitely influenced his working life, because the ideology is all in the title. So, how can you be better, and how can you differentiate yourself or the company you work for, and make a difference, and then elevate to go from being a good company to a great company or a good employee to a great employee. So, that's one of the books that definitely influenced his career. And he really enjoyed the fact that it had case studies in there that you were able to look at, read and understand, and then it gives you that a little bit more of a practical application when there's case studies that you can look at.   What Neil is Really Excited About Now! When asked about something that he's excited about, Neil shared that he's been with InterVision approximately 12 months, and in that time, they've had a tremendous amount of growth.  They are working a significant amount with Amazon, on Amazon Connect in the contact center space, and watching how that's changing the contact center space is really, really incredible. And with that, they have releases on a weekly or biweekly basis, and you see new technologies and new items come out, and it's actually an interesting challenge making sure that his team is not only at the cutting edge of technology, but what's new and modern today is, for want of a better description, a month old in a month's time.  And making sure that his team are kept current and up to date with all of these technology changes, specifically around AI and ML, that's really an interesting challenge, because the solutions of a year ago aren't solutions for today, and he finds that both interesting challenge from a business perspective, but it's also rewarding because you get the opportunity to have people do training classes and learnings to make sure that they're at the top of their skill game to be able to deliver the best in class products that they like to offer.   Where can listeners find Neil online? LinkedIn – Neil Leyland Website – www.intervision.com   Quote or Saying that During Times of Adversity Neil Uses Me: Now, before we wrap our episodes up, Neil, we always like to ask our guests, do you have a quote or a saying that during times of adversity or challenge, you will tend to revert to this quote if for any reason you get derailed or you get off track, the quote kind of helps to get you back on track. Do you have one of those?    When asked about a quote or saying that he tends to revert to, Neil shared that it's not quite that, but there's a phrase that he often thinks about in challenges when he's working with his colleagues, or they're looking at a project and how to move forward, and it's directly related to customer service and it's, “The tolerance of poor behaviour is worse than the behaviour itself.”   Me: That's such a powerful statement.   Neil shared that he loves it because it's applicable everywhere. In your personal life, you can choose not to go to the gym, or you can go to the gym. In work, you can watch people do things and managers do things or accept things and that they shouldn't and as soon as a behaviour becomes ingrained, it's far more challenging to remove it.  Me: Yeah, agreed. Thank you so much for sharing. So, we just want to extend our deepest gratitude to you, Neil, for hopping on our podcast today, sharing about InterVision, about your journey, as well as what you're doing at InterVision, the impact of AI and machine learning, the opportunity that workers have in the centers as agents to upskill their competencies and behaviours so that they can better serve customers and solve problems quicker. It was really a rewarding and engaging conversation, and I want to just extend our deepest gratitude to you. So, thank you so much.    Please connect with us on Twitter @navigatingcx and also join our Private Facebook Community – Navigating the Customer Experience and listen to our FB Lives weekly with a new guest   Links •    Good to Great: Why Some Companies Make the Leap…And Others Don't by Jim Collins   The ABC's of a Fantastic Customer Experience Grab the Freebie on Our Website – TOP 10 Online Business Resources for Small Business Owners  Do you want to pivot your online customer experience and build loyalty - get a copy of “The ABC's of a Fantastic Customer Experience.” The ABC's of a Fantastic Customer Experience provides 26 easy to follow steps and techniques that helps your business to achieve success and build brand loyalty. This Guide to Limitless, Happy and Loyal Customers will help you to strengthen your service delivery, enhance your knowledge and appreciation of the customer experience and provide tips and practical strategies that you can start implementing immediately! This book will develop your customer service skills and sharpen your attention to detail when serving others. Master your customer experience and develop those knock your socks off techniques that will lead to lifetime customers. Your customers will only want to work with your business and it will be your brand differentiator. It will lead to recruiters to seek you out by providing practical examples on how to deliver a winning customer service experience!

Idelfonso Nogueira is a brazilian scientist from Salvador Bahia, currently an Associate Professor at the Department of Chemical Engineering at NTNU. Nogueira is a Ph.D. in Chemical and Biological Engineering and has more than 10 years of experience in the field of machine learning. Nogueria  has conducted research on the scents and flavors of AI and his fields of interest and research include, among others, merging AI, systems optimization and automation. See omnystudio.com/listener for privacy information.

Part 2 of 4. My guest for this week's episode is Noam Solomon, CEO and co-founder at Immunai, a pioneering biotech company that is comprehensively mapping and reprogramming the immune system with single-cell biology and AI to power new therapeutic discoveries, accelerate drug development, and improve patient outcomes.

In episode 95 of Cybersecurity Where You Are, Sean Atkinson is joined by Randy Rose, VP of Security Operations & Intelligence at the Center for Internet Security® (CIS®).Together, they discuss AI augmentation in terms of how cyber defenders are using generative artificial intelligence to enhance their capabilities.Here are some highlights from our episode:01:16. How artificial intelligence has changed the landscape for cybersecurity defenders03:49. How AI is starting to augment threat detection10:12. What security researchers are exploring around AI and cyber defense20:54. Key challenges and limitations for AI-based cyber defense30:54. Future trends and innovations for cybersecurity defenders' use of AIResourcesEpisode 56: Cybersecurity Risks and Rewards of LLMsEpisode 59: Probing the Modern Role of the PentestSEC595: Applied Data Science and AI/Machine Learning for Cybersecurity Professionalsfr0gger / Awesome-GPT-AgentsThe LLM Misinformation Problem I Was Not ExpectingSeparating FUD from Practical for Post-Quantum CryptographyIf you have some feedback or an idea for an upcoming episode of Cybersecurity Where You Are, let us know by emailing podcast@cisecurity.org.

AI is machine learning. It's great for repetitive tasks. It collects knowledge and synthesizes it into useful compilations. AI machine learning can be predictable. “I” learning is what we do when we interact with people. It's as unpredictable as an amateur golf swing. Interacting with humans requires empathy, psychology, self-awareness, and patience. It is time-consuming and can have different results on different days. Interacting with people is how we do business. People do business with people, and although AI can help, it cannot replace human-to-human interaction. In the world of generating leads, social networks, and business in the post-pandemic economy, it's not just connections that matter. It's who you are connecting with and who they have in their networks that matter. It's your job to stay connected and keep you, your business, and your content and concepts top of mind with them. Active connections create relationships, businesses, and referrals.

In this episode of the Eye on AI podcast, we delve into the power of AI in business and data science and data management with Shub Bhowmick, CEO and co-founder of Tredence.   Explore how Tredence is revolutionizing industries by integrating AI and data science to drive business outcomes. Shub shares his journey from management consulting to leading a team of 2,500 data scientists and engineers. Learn how Tredence is supercharging Fortune 200 clients across retail, consumer packaged goods (CPG), financial services, healthcare, and more, with custom AI solutions.   We dive deep into the challenges and opportunities in leveraging AI, particularly in the CPG sector, and how Tredence addresses data silos, enhances data quality, and implements real-time data processing. Shub also discusses the implementation of generative AI, detailing the company's journey from foundational models to enterprise adoption and responsible AI practices.   Discover Tredence's Gen AI as a service platform, designed to help enterprises navigate the complexities of AI integration, optimize infrastructure, and unlock business value through innovative AI solutions. Shub highlights real-world applications of AI, such as demand forecasting and predictive analytics, demonstrating how these technologies can significantly reduce costs and improve efficiency.   Gain insights into Tredence's growth strategy, including their expansion into new verticals, service lines, and geographies. Shub emphasizes the importance of a multidisciplinary approach, combining domain expertise with cutting-edge technology to deliver customized, impactful AI solutions.   Don't miss out on this insightful conversation. Like, subscribe, and hit the notification bell for more expert discussions on the technologies driving the AI revolution     This episode of Eye on AI  is sponsored by BetterHelp. If you're thinking of starting therapy, give BetterHelp a try. It's entirely online. Designed to be convenient, flexible, and suited to your schedule. Just fill out a brief questionnaire to get matched with a licensed therapist, and switch therapists any time for no additional charge. Visit https://www.betterhelp.com/eyeonai today to get 10% off your first month. This episode is sponsored by Shopify.  Shopify is a commerce platform that allows anyone to set up an online store and sell their products. Whether you're selling online, on social media, or in person, Shopify has you covered on every base. With Shopify you can sell physical and digital products. You can sell services, memberships, ticketed events, rentals and even classes and lessons. Sign up for a $1 per month trial period at http://shopify.com/eyeonai Stay Updated: Craig Smith Twitter: https://twitter.com/craigss Eye on A.I. Twitter: https://twitter.com/EyeOn_AI (00:00) Preview and Introduction (02:42) Introducing Shub Bhowmick, CEO of Tredence (03:29) What Tredence Does and Their Clients (04:32) Why Tredence Focuses on Consumer Packaged Goods (06:18) AI Transformation Example in Retail (09:24) Building a Solid Data Foundation (10:42) Improving Reporting with Real-Time Data (11:44) Starting the AI and Data Science Journey (13:32) How Tredence Uses Generative AI (14:10) Stages of Implementing Generative AI (19:09) Applying AI Across Different Industries (23:43) Overview of Gen AI as a Service Platform (25:37) Features and Benefits of the Platform (28:42) How Gen AI is Integrated into Businesses (33:02) Tredence's Unique Approach to AI Solutions (36:08) Importance of Domain-Specific AI Solutions (40:42) Tredence's Growth and Expansion Plans (47:57) Size and Reach of Tredence (50:01) Competing in the AI Market

מאיר פנסטר ואילן ריס, עורכי פטנטים בקבוצת ארליך מארחים את רועי מלצר ראש מחלקת טכנולוגיות מחשוב בקבוצת ארליך, איך כותבים פטנט באמצעות AI וזה מזה Machine Learning?

В этом выпуске опытный разработчик Сергей Вяткин делится своим опытом поиска работы в IT в США после 60 лет. Он рассказывает о своем пути от первых шагов в программировании в СССР до современной работы с облачными технологиями. Сергей делится ценными советами по поиску работы, прохождению собеседований и адаптации резюме для старших кандидатов. Мы обсудили вопросы возрастной дискриминации в IT, особенности работы с legacy-кодом, важность постоянного обучения новым технологиям. Также затронули темы работы на контрактной основе, подработки в сфере AI и Machine Learning, и как оставаться востребованным специалистом в любом возрасте. Смотрите видео: https://youtu.be/cSWMzqT-TcE Сергей Вяткин (Sergey Vyatkin), Senior Software Engineer at Moody's Corporation (ex TuneIn, General Electric, Bank of America) LinkedIn: https://www.linkedin.com/in/sergey-vyatkin-1072456/ Статьи от Сергея по поиску работы на Хабре: https://habr.com/ru/users/Sergunka/publications/articles/ Рассказ Сергея о том, как он искал работу в 2005 году, с чего и началось в русскоязычном зарубежье обсуждения поиска работы и приемы ведения дискуссии на интервью: https://sergey-vyatkin.livejournal.com/1031.html Статистика поиска работы от Сергея (отклики, собеседования): https://docs.google.com/spreadsheets/d/17dP-JtpWQN7OuS9I3olChz49bXHK3iZFz7KD8v7tGV8/edit?usp=sharing Ссылки на сайты по ИИ обучению: https://www.dataannotation.tech/ https://app.outlier.ai/ *** Записывайтесь на карьерную консультацию (резюме, LinkedIn, карьерная стратегия, поиск работы в США): https://annanaumova.com Онлайн курс "Идеальное резюме и поиск работы в США": https://go.mbastrategy.com/resumecoursemain Гайд "Идеальное американское резюме": https://go.mbastrategy.com/usresume Гайд "Как оформить профиль в LinkedIn, чтобы рекрутеры не смогли пройти мимо" (предзаказ): https://link.coursecreator360.com/widget/form/ObfVCQ2clIWTdNcQBAkf Мой Telegram-канал: https://t.me/prodcastUSA Мой Instagram: https://www.instagram.com/prodcast.us/

Louis Lehot, Partner and Business Lawyer at Foley & Lardner LLP, hosted a live-stream webinar for startup founders and investors in 2023, a year marked as the era of generative AI. New software, including ChatGPT, DALL-E, Midjourney, and Bard, collectively created one of those rare "This is going to change everything!" moments in Silicon Valley and beyond. For startup founders and investors, the question was: What does Artificial Intelligence and Machine Learning mean, and how do we navigate the opportunities ahead?See the full video discussion here.Discover more about Louis Lehot and explore additional professional insights on his website: https://louislehot.comExplore Related Content: This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit louislehotattorney.substack.com

Join us for an insightful episode as we delve into the world of service and sales department challenges and solutions with Maria Maleki, Regional Sales Manager, a seasoned expert at CallRevu. In this interview, Maria sheds light on the critical aspects of service department communication, sales training, and development, particularly focusing on the Service BDC (Business Development Center). Maria highlights common weaknesses that hinder effective communication in service departments, emphasizing the importance of building strong relationships with customers. She points out that newer and younger personnel often struggle when faced with complex customer inquiries related to service work, pricing, and timeframes, which can lead to missed opportunities and customer dissatisfaction.  Throughout the conversation, Maria provides valuable tips to address these challenges, emphasizing the significance of active listening, rapport-building, and effective questioning techniques. By recapping conversations, expanding on responses, and demonstrating understanding, service advisors can enhance the customer experience, increase appointment conversions, and foster continued engagement. Maria also shares key tactics to position service advisors as trusted problem solvers, focusing on value rather than price alone. By showcasing the process of diagnostics in a transparent and non-intimidating manner, advisors can instill confidence in customers and emphasize the importance of choosing their services. Furthermore, Maria discusses the use of AI technology to analyze customer interactions, flagging keywords related to common issues such as pricing concerns or competitor mentions. By leveraging AI insights to identify and address recurring problems, service departments can proactively coach and train staff for better outcomes, ultimately improving customer satisfaction and retention. Tune in to discover practical strategies and techniques to enhance service department communication, build customer relationships, and drive business success in the competitive automotive service industry. Contact Maria: mariamaleki@callrevu.com Dealer Talk with Jen Suzuki Podcast |Jennifer@edealersolution.com | 800-625-1590 | edealersolutions.com  

The University of Oxford is made up of over 30 colleges and halls spread across the city centre and beyond. These colleges are at the heart of Oxford's reputation as one of the best universities in the world and make it a very special place to study and live. With the oldest colleges being over 700 years old, it is a rare occurrence for a new college to be established. In 2019, Reuben Colleges was established as the newest college at Oxford University. It is a graduate college dedicated to fostering interdisciplinary exchange within an inclusive, innovative and impact-oriented community. My guest in this episode is the first president of Reuben College. Professor Lord Lionel Tarassenko CBE FREng FMedSci is the first president of Reuben College, and Theme Lead for the AI & Machine Learning research cluster. He is a world-leading expert in the application of signal processing and machine learning to healthcare, with a strong track record in translation to clinical medicine. Professor Tarassenko's work has had a major impact on the identification of deterioration in acute care and on the management of chronic disease. The system which he designed for patient monitoring in critical care was the first machine learning system to gain FDA approval (in 2008). Prior to that, Professor Tarassenko had been closely involved in the development of some of the jet engine monitoring software at the core of the Rolls-Royce TotalCare® package. This won him the Rolls-Royce Chairman's Award for Technical Innovation in 2001 and the Sir Henry Royce High Value Patent Award in 2008. Professor Tarassenko was elected to a Fellowship of the Royal Academy of Engineering in 2000, and to a Fellowship of the Academy of Medical Sciences in 2013. He has won many awards for his work, including the British Computer Society (BCS) Medal, the Silver Medal of the Royal Academy of Engineering, and the Institute of Engineering & Technology IT Award, among others. He was made a Commander of the British Empire (CBE) for services to engineering in the 2012 New Year's Honours List. In May of 2024, Professor Tarassenko was appointed to the House of Lords as a non-party-political peer and will join the House of Lords as a cross-bencher (Baron Tarassenko of Headington).

Theresa Young, CSAA Insurance Group; Michael Angelina, St. Joseph's College; and Steve Clarke, Verisk, said issues also arise from lack of regulation of third-party vendors, public adjusters, and service providers, which can lead to higher claim payouts than were anticipated in pricing.

April 9: First, we join Justin Coran, Chief Analytics Officer at Renown Health, and Steven Ramirez, Chief Information Security & Technology Officer at Renown Health, at the HIMSS conference. The duo shares their ambitious initiatives for the year, focusing on transitioning to a multi-cloud environment with AWS and Azure, and the importance of building a sustainable infrastructure that ensures security and prevents data leakage. They emphasize the critical role of analytics maturity in healthcare systems, highlighting the integration of artificial intelligence and machine learning to revolutionize clinical and administrative operations. Governance and data literacy emerge as pivotal themes, underscoring the need to prepare the workforce for an AI-driven future. Then, we dive in with Perry Welch, Chief Sales Officer at Airwavz Solutions, a company at the forefront of solving in-building connectivity challenges within healthcare facilities. As hospitals and healthcare institutions evolve into sprawling complexes, the issue of reliable cellular service becomes increasingly critical, not just for convenience but for essential medical operations. Welch delves into how Airwavz addresses this problem by creating custom-designed, carrier-agnostic networks that ensure seamless communication within these complex environments.Categories: Cloud, AI / Machine Learning

In a new episode of the pharmaphorum podcast recorded live on site at WIRED Health in London in March, web editor Nicole Raleigh spoke with Elise de Reus, co-founder of Cradle, the generative AI platform that helps scientists to design and program proteins.

Explore the journey from idea to triumph in the tech realm with Kumar Ujjwal, Founder and CEO at DwellFi, on this episode of GSD Presents: Silicon Valley Tech & AI. Kumar, a visionary leader and serial entrepreneur, boasts a remarkable track record with successes like DwellFi and Revni, specializing in SaaS, AI/Machine Learning, Blockchain, and Fintech. With a $500 million exit from Punchh and accolades including recognition as one of the Emerging 50 fund managers in 2022 by Signature Block, Kumar's strategic acumen and fusion of technical expertise and business savvy promise an insightful exploration into the recipe for entrepreneurial triumph.

Step into the future of branding with Steve Red and Tessa Burg. They unravel how your company can forge deeper connections with your audience through authentic storytelling. Discover the secrets to transforming data points into compelling stories that do more than sell—they create authentic brand experiences. Creativity and analytics come together as we discuss how to harness AI and ML not just to streamline operations but to infuse marketing with insights that drive growth and resonate on a human level. Learn how to make data your ally in the creative process, and shape campaigns that speak directly to the hearts and minds of your ideal customers. You'll hear ideas on how to combine AI with the human touch to tell stories that are as data-driven as they are emotionally charged. Join us to redefine what branding means in the age of technology, and discover a new era of customer engagement. Leader Generation is hosted by Tessa Burg and brought to you by Mod Op.    About Steve Red: Steve is the Co-Chief Creative Officer at Mod Op. The founder of Red Tettemer O'Connell + Partners, recently acquired by Mod Op, Steve touches virtually everything that goes out of the agency door from a creative and strategic perspective. He believes people are smart and advertising should be too, and that the only real martini is made with gin. Someday … long after his teenagers aren't teenagers anymore, he'll settle down and start painting again, which is what his dad says he should have done all along. Steve can be reached at Steve.Red@ModOp.com. About Tessa Burg: Tessa is the Chief Technology Officer at Mod Op and Host of the Leader Generation podcast. She has led both technology and marketing teams for 15+ years. Tessa initiated and now leads Mod Op's AI/ML Pilot Team, AI Council and Innovation Pipeline. She started her career in IT and development before following her love for data and strategy into digital marketing. Tessa has held roles on both the consulting and client sides of the business for domestic and international brands, including American Greetings, Amazon, Nestlé, Anlene, Moen and many more. Tessa can be reached on LinkedIn or at Tessa.Burg@ModOp.com.

Welcome to a special series where I interview industry leaders LIVE during the NADA Show! Meet Aleks Keric, General Manager of the Murgado Automotive Group in Chicago! We get into the world of used car strategies and risk management in the automotive industry. In this insightful conversation, Aleks sheds light on how managing risk can be a powerful strategy, often overlooked by operators who fail to see its potential impact. We explore the challenges of taking in a high volume of trade-ins on a busy day, with a slammed service department, and the risks involved in owning inventory for retail. Alex shares his unique approach to risk management, likening it to monitoring the stock market, and emphasizes the importance of focusing on at-risk cars with low demand to optimize sales. As a dedicated user of LotLinx, Aleks reveals how AI and machine learning tools have revolutionized his decision-making process, particularly in identifying cars that require more marketing activity. By leveraging LotLinx's VVO tool and collaborating closely with his marketing team, Alex streamlines his operations and accelerates the sales cycle, ultimately saving costs and boosting efficiency. Discover the innovative strategies Aleks has implemented to adapt his business in 2024, including the implementation of a strict 60-day hard turn policy. Learn how he maximizes vehicle descriptions with compelling language, incorporating ChatGPT and emojis to enhance customer engagement and drive sales. Tune in to this episode to gain valuable insights from Alex Carrick on transforming risk management into a strategic advantage and optimizing used car operations in the ever-evolving automotive landscape. Dealer Talk with Jen Suzuki Podcast |Jennifer@edealersolution.com | 800-625-1590 | edealersolutions.com

In the last 12 months, artificial intelligence (AI) has taken the world by storm, and you've likely felt the effects at your organization! From the rapid rise of ChatGPT and other AI tools, it can be hard to keep up – and hard to know when and how to dive in. BDI's Chief Intelligence Officer, Mike Rogers, and Chief Creative Officer, James Read, are eager to help. In this episode, you'll learn…A helpful history of AI and how it worksPossible applications of AI in fundraisingHow AI can make your work more efficient and effectiveListen now to be encouraged and inspired to harness the power of AI at your nonprofit organization!Episode Highlights:Meet Mike and James and hear their experiences in the nonprofit industryA brief history of artificial intelligenceHow does AI technology show up in our lives every day?Why is AI at the forefront of our conversations today?Major milestones of OpenAI and ChatGPTTypes of AI: Machine Learning vs. Deep Learning or Generative AIHow BDI has adopted AI tools into our workflowPLUS… a game: “Is it real or is it AI?”Connect with us!Connect with Katrina Williams: LinkedInConnect with Mike Rogers: LinkedInConnect with James Read: LinkedInConnect with BDI: Website | LinkedIn | Facebook | InstagramResources:Article: How AI is Revolutionizing Nonprofit Fundraising & MarketingArticle: 5 Steps to Integrate AI Into Your OrganizationArticle: The Absolute Beginner's Guide to AI-Powered Productivity

Bret and Nirmal are joined by Michael Irwin, DevRel at Docker, to talk about all the products and features Docker shipped in 2023, and what's coming in early 2024. Michael has been on this show many times as a Docker Captain and now as a Docker employee, and it's always great to dig into the details of the products with someone who's been using them for so many years as an end-user and now staff at Docker. Docker did some big things in 2023, but they also shipped some smaller features that we will help you catch up on in this episode.The live recording of the complete show from December 28, 2023 is on YouTube (Ep. #247)Creators & Guests Cristi Cotovan - Editor Beth Fisher - Producer Bret Fisher - Host Nirmal Mehta - Host Michael Irwin

Michael Hoffman, or “Hoff” as most know him, is the co-founder and CEO at IQXR, a company solving the hardest problems facing global-scale, enterprise XR deployments, and doing so with an open source approach.Previously Michael spent nearly a decade working with the Microsoft Hololens team. He was a Principal Engineering Lead at Microsoft for a couple of years, left to be the founding partner of Object theTheory, where he and his team worked with enterprises to leverage AR and VR technologies, often in combination with IoT and AI/Machine Learning. And then he went back to Microsoft for a couple of years to lead the development of the Mixed Reality Toolkit (MRTK) project.Earlier in his career, Michael worked in software engineering roles at Google, Nike, and several startups.In this conversation, Hoff describes how 3D visualization, with AR and VR technologies, changes our comprehension of digital information, contributes to the value of having your hands free to interact with the world, and enables better efficiency and better insights.Within the enterprise setting, Hoff notes it's relatively easy to get to a pilot and prove value, but it's really difficult to deliver that value at scale.We go on to talk about making AR/VR solutions viable within an enterprise setting at scale, including challenges around visual and audio haptics, working both online and offline, and other key bits of plumbing, as well as the misconceptions that many enterprises have about the technology.We also discuss:- the rationale and corporate strategy for building open source solutions, - the role of AI in accelerating software and content development,- the art of the AI prompt, and- how Apple Vision Pro accelerates the market.Hoff wraps up by discussing neurodivergence and his own growing awareness and acceptance of the challenges and benefits of neurodivergence for both his children and himself.You can find all of the show notes at thearshow.com. Please consider contributing to my Patreon at https://www.patreon.com/theARshow.Links From The Episode- Press Release: [Microsoft Talent Joins The Mesmerise Group to Drive Growth of Immersive Technology Solutions for the Enterprise](https://www.prnewswire.com/news-releases/microsoft-talent-joins-the-mesmerise-group-to-drive-growth-of-immersive-technology-solutions-for-the-enterprise-301856247.html)- Article: [What is ikigai and how can it change my life?](https://www.betterup.com/blog/what-is-ikigai) by Elizabeth Perry for BetterUp- Book: [Ikigai: The Japanese Secret to a Long and Happy Life](https://amzn.to/48pf15s) by Héctor García and Francesc Miralles - Book: [The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers](https://amzn.to/2ZpiQ8m) by Ben Horowitz- Book: [Ready Player One](https://amzn.to/2X9Eu2t) by Ernest Cline

Jonathan Cassar, the Chief Technology Officer and Head of Information Security at the Malta Information Technology Agency (MITA), provides insights into global cybersecurity trends and their implications for the public sector. He discusses strategies to bolster cybersecurity in government and critical infrastructure, approaches to tackling the cybersecurity talent shortage, the integration of human expertise with technology to combat threats effectively, and best practices for automating security operations in government. Olivia Neal [host] | LinkedIn   Alvaro Vitta | LinkedIn Jonathan Cassar | LinkedIn MITA Microsoft Public Sector Center of Expertise for more information and transcripts of all episodes  Discover and follow other Microsoft podcasts at aka.ms/microsoft/podcasts   

Are you ready to accelerate your business's growth through the right level of technology? If so, today's guest Don Finley, Founder of FINdustries, and I discuss the power of AI and software development. We discuss the keys to the early adoption of AI. 1) Workflow Analysis 2) Identify Access Turn Around Time 3) Assign Ownership within your Team FINdustries is a relationship-focused web and software development agency. Their expertise is in the acceleration of revenue growth for organizations through tailoring the right development to successfully bring the vision of our clients into reality, specializing in harnessing the power of blockchain and AI/Machine Learning technologies. They have successfully completed dozens of projects for a variety of industries, and have refined their workflows and delivery processes to easily handle rapidly changing requirements. Ready for your software to make waves? To learn more or hire Don and the FINdustries team visit: https://fin.dustries.com/ Thank you for listening to another episode of the Perky Collar Radio Show. Warmest Regards, David M. Frankel Perky Collar Inventor, Perky, LLC Founder, Perky Collar Radio Show Host, Commercial Real Estate Broker & Business Broker www.PerkyLLC.com, www.BBOTC.net Feel free to join my Entrepreneur Group on Facebook www.Facebook.com/Groups/CharlotteEntrepreneurThinkTank Feel free to learn more about The Fenx and join fellow successful Entrepreneurs https://entrepreneurs-maclackey.thrivecart.com/the-fenx-monthly/?ref=cettsupport Feel free to connect with me on Linkedin www.Linkedin.com/in/DavidMFrankel Ready to write a book and share your story with the world? Let me help you get it done every step of the way. Go to https://perky.bookpublishingagency.com/ --- Support this podcast: https://podcasters.spotify.com/pod/show/perkycollaradioshow/support

Guest: Chloe Messdaghi, Head of Threat Research, Protect AIOn ITSPmagazine | https://www.itspmagazine.com/itspmagazine-podcast-radio-hosts/chloe-messdaghiOn Twitter | https://twitter.com/ChloeMessdaghiOn LinkedIn | https://www.linkedin.com/in/chloemessdaghiOn Instagram | https://www.instagram.com/chloemessdaghi/Website | https://www.securebychloe.com__________________________SponsorsAre you interested in sponsoring an ITSPmagazine Channel?

Guest: Chloe Messdaghi, Head of Threat Research, Protect AIOn ITSPmagazine | https://www.itspmagazine.com/itspmagazine-podcast-radio-hosts/chloe-messdaghiOn Twitter | https://twitter.com/ChloeMessdaghiOn LinkedIn | https://www.linkedin.com/in/chloemessdaghiOn Instagram | https://www.instagram.com/chloemessdaghi/Website | https://www.securebychloe.com__________________________SponsorsAre you interested in sponsoring an ITSPmagazine Channel?

Description: In this thought-provoking episode, we have the privilege of hosting Shakeel Aslam Mohammed, a visionary entrepreneur and the founder of Avant Safety and Pathegy. With a keen focus on health safety data analytics and productivity, Shakeel's venture, Avant Safety, is reshaping the way industries view occupational safety. Our discussion delves into how his company's innovative approach, particularly through Shift Analytics, is minimizing disruption and elevating health and safety standards. Shakeel's diverse entrepreneurial journey doesn't end there; he's also the brain behind Pathegy, a career and business coaching enterprise helping individuals and businesses stride confidently on the path of success. The conversation unfolds into the realm of artificial intelligence and machine learning, where Shakeel shares his insights on the challenges, the rapid pace of development, and the future concerns these technologies harbor. His reflections on how AI and machine learning can be harnessed to further enhance workplace safety and productivity are nothing short of enlightening. We also get personal with Shakeel, exploring his roots, family, and the core values that drive his ventures. From strategy deliberations to personal motivations and fears, particularly around AI, Shakeel provides a candid look into the mind of a forward-thinking entrepreneur. Our discussion also touches on the theme of 'Equipping for Success', which is not just a business mantra, but a societal call to action for Shakeel. His dedication towards enhancing the wellbeing of humanity shines through as he articulates his goals and the broader impact he envisions. Join us in this riveting episode as we navigate the intersections of technology, business innovation, and the human-centric approach that underscores Shakeel's mission. This is a conversation loaded with nuggets of wisdom, and a clarion call to equip ourselves for the success and challenges of the modern world. Disclaimer: Not advice. Educational purposes only. Not an endorsement for or against. Results not vetted. Views of the guests do not represent those of the host or show. Do your due diligence. Click here to join PodMatch (the "AirBNB" of Podcasting): https://www.joinpodmatch.com/drchrisloomdphd We couldn't do it without the support of our listeners. To help support the show: CashApp- https://cash.app/$drchrisloomdphd Venmo- https://account.venmo.com/u/Chris-Loo-4 Buy Me a Coffee- https://www.buymeacoffee.com/chrisJx Thank you to our sponsor, CityVest: https://bit.ly/37AOgkp Click here to schedule a 1-on-1 private coaching call: https://www.drchrisloomdphd.com/book-online Click here to purchase my books on Amazon: https://amzn.to/2PaQn4p Follow our YouTube channel: https://www.youtube.com/chL1357 Follow us on Twitter: https://www.twitter.com/drchrisloomdphd Follow us on Instagram: https://www.instagram.com/thereal_drchrisloo Follow us on TikTok: https://www.tiktok.com/@drchrisloomddphd Follow the podcast on Spotify: https://podcasters.spotify.com/pod/show/christopher-loo Thank you to our advertisers on Spotify. Financial Freedom for Physicians, Copyright 2023

The latest Nassau Re/Imagine Podcast episode explores the crucial steps in integrating AI and machine learning into businesses: from pilot programs to seizing opportunities and fostering experimentation. Join us in this insightful discussion recorded during Hartford Innovation Week, where we unpack the profound impact of AI, ML, and Quantum Computing on our society, businesses, and personal lives. Dive into an engaging panel discussion hosted by Nassau Financial Group, featuring experts Francesco Ricigliano from AdvanceCT, Sanjeevnayak Nayak, and Mike DiDonato from the University of Connecticut. Their expertise provides a solid groundwork for the intriguing conversations that ensued. If you're passionate about these transformative technologies and keen to be part of future discussions, drop your interest in the comments below. Let's continue this dialogue and drive innovation forward! #AI #MachineLearning #Innovation #BusinessStrategy  

Dr. Amin Madani is the director of the Surgical AI Research Academy (SARA) at the University Health Network (UHN). He is an endocrine and acute care surgeon at UHN and assistant professor of surgery at the University of Toronto. Dr. Madani talked to us and showed us some of the work he is doing on AI in surgery and in particular on computer vision. He really breaks down for us the terms AI, machine learning, and data science, and highlighted some of the promise and challenges for AI in surgery. Email Dr. Madani: amin.madani@uhn.ca Watch full unedited interview on YouTube: https://youtu.be/aawdQy90v2Q Links: 1. https://temertysimcentre.com/surgical-artificial-intelligence-research-academy-sara/ 2. Video of Go No Go: https://www.youtube.com/watch?v=MmcW8JK1Qv4 3. Surgical Data Science: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9135051/ 4. Artificial Intelligence for Intraoperative Guidance: Using Semantic Segmentation to Identify Surgical Anatomy During Laparoscopic Cholecystectomy. https://pubmed.ncbi.nlm.nih.gov/33196488/

We are honored to be joined today by Guru Banavar. Guru Banavar, the founding CTO of Viome, has been instrumental in its success by integrating AI with biological insights to offer in-depth reports on gene expression. Before Viome, he served as the global VP and Chief Science Officer at IBM, and was the founding VP of the Watson AI Research team. With awards like the Leadership in Technology Management and a National Innovation Award from India's President, his 35+ US patents and contributions have been highlighted by top media outlets including the New York Times and BBC. You won't want to miss this episode! Topics: 1. Understanding Healthspan and Lifespan - Distinction between healthspan and lifespan - Importance of healthspan 2. The Role of Gene Expression - Influence on health and lifespan - Process of DNA to RNA transcription - Consideration of microbial gene expression 3. Host-Microbe Interactions - Dynamics and significance 4. Viome's mRNA Analysis - Testing process - Data interpretation and diet recommendation - Oxalate sensitivity analysis explanation - AI's role in personalizing treatments - Machine learning for glycemic response 5. Viome Test Offerings - Available tests and their unique insights - Differences and recommendations: oral vs. gut microbiome tests. Thanks for tuning in! Order a Viome Test NOW! Get Chloe's Book Today! "⁠⁠75 Gut-Healing Strategies & Biohacks⁠⁠" If you liked this episode, please leave a rating and review or share it to your stories over on Instagram. If you tag @synthesisofwellness, Chloe would love to personally thank you for listening! Follow Chloe on Instagram ⁠⁠⁠⁠⁠⁠⁠⁠@synthesisofwellness⁠⁠⁠⁠⁠⁠⁠⁠⁠ Follow Chloe on TikTok @chloe_c_porter Visit ⁠⁠⁠⁠⁠⁠⁠⁠⁠synthesisofwellness.com⁠⁠⁠⁠⁠⁠⁠⁠⁠ to purchase products, subscribe to our mailing list, and more! Or visit ⁠⁠⁠⁠⁠⁠⁠⁠⁠linktr.ee/synthesisofwellness⁠⁠⁠⁠⁠⁠⁠⁠⁠ to see all of Chloe's links, schedule a BioPhotonic Scanner consult with Chloe, or support the show! Thanks again for tuning in! --- Support this podcast: https://podcasters.spotify.com/pod/show/chloe-porter6/support

EdTech Speaks welcomes another esteemed guest, Steven Emanuel, a legal educator and visionary who has spent nearly five decades transforming how law students approach learning. Join us as he shares his insights into revolutionizing education through AI and machine learning. As a legal scholar and author, Steven has been a prominent figure in the field of legal education for over 48 years. Steven's journey began in 1974, when he started writing and publishing books for law students, focusing on simplifying and enhancing the learning process. Today, as the principal author of the acclaimed Emanuel Law Outlines series, he's dedicated to helping law students effectively master complex subjects. He has challenged traditional "Socratic method" and case-oriented approaches, developing more efficient ways for students to grasp the substance of law-school subjects. EmanuelAYCE.com provides law students access to a vast library of study aids through a subscription model, fostering a more personalized and effective learning process.Steven says, "Education is about to be transformed by the use of AI/Machine Learning methods (like Chat-GPT), which will give each student a truly customized learning experience. Anyone who wants to "teach" others should be actively exploring how these ML-based methods can create an entirely new way of delivering educational content, because these will (I predict) largely supplant traditional methods of instruction."For more insights and to explore the world of legal education transformation, visit www.EmanuelAYCE.com and discover the array of resources designed to empower law students on their academic journey.Learn more and connect with Steven here:www.emanuelAYCE.comGet your complimentary 2-month subscription: email semanuel@westnet.com, mention "Free Trial of EmanuelAYCE" Get bonus content on Patreon Hosted on Acast. See acast.com/privacy for more information.

FreightWaves' Mike Baudendistel interviews Manish Ghosh, VP of Strategy and Consumer Industries at Blue Yonder. Follow The Stockout Podcast Other FreightWaves Shows Learn more about your ad choices. Visit megaphone.fm/adchoices

FlashAttention was first published by Tri Dao in May 2022 and it had a deep impact in the large language models space. Most open models you've heard of (RedPajama, MPT, LLaMA, Falcon, etc) all leverage it for faster inference. Tri came on the podcast to chat about FlashAttention, the newly released FlashAttention-2, the research process at Hazy Lab, and more. This is the first episode of our “Papers Explained” series, which will cover some of the foundational research in this space. Our Discord also hosts a weekly Paper Club, which you can signup for here. How does FlashAttention work?The paper is titled “FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness”. There are a couple keywords to call out:* “Memory Efficient”: standard attention memory usage is quadratic with sequence length (i.e. O(N^2)). FlashAttention is sub-quadratic at O(N). * “Exact”: the opposite of “exact” in this case is “sparse”, as in “sparse networks” (see our episode with Jonathan Frankle for more). This means that you're not giving up any precision.* The “IO” in “IO-Awareness” stands for “Input/Output” and hints at a write/read related bottleneck. Before we dive in, look at this simple GPU architecture diagram:The GPU has access to three memory stores at runtime:* SRAM: this is on-chip memory co-located with the actual execution core. It's limited in size (~20MB on an A100 card) but extremely fast (19TB/s total bandwidth)* HBM: this is off-chip but on-card memory, meaning it's in the GPU but not co-located with the core itself. An A100 has 40GB of HBM, but only a 1.5TB/s bandwidth. * DRAM: this is your traditional CPU RAM. You can have TBs of this, but you can only get ~12.8GB/s bandwidth, which is way too slow.Now that you know what HBM is, look at how the standard Attention algorithm is implemented:As you can see, all 3 steps include a “write X to HBM” step and a “read from HBM” step. The core idea behind FlashAttention boils down to this: instead of storing each intermediate result, why don't we use kernel fusion and run every operation in a single kernel in order to avoid memory read/write overhead? (We also talked about kernel fusion in our episode with George Hotz and how PyTorch / tinygrad take different approaches here)The result is much faster, but much harder to read:As you can see, FlashAttention is a very meaningful speed improvement on traditional Attention, and it's easy to understand why it's becoming the standard for most models.This should be enough of a primer before you dive into our episode! We talked about FlashAttention-2, how Hazy Research Group works, and some of the research being done in Transformer alternatives.Show Notes:* FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness (arXiv)* FlashAttention-2* Together AI* From Deep Learning to Long Learning* The Hardware Lottery by Sara Hooker* Hazy Research* Is Attention All You Need?* Nvidia CUTLASS 3* SRAM scaling slows* Transformer alternatives:* S4* Hyena* Recurrent Neural Networks (RNNs)Timestamps:* Tri's background [00:00:00]* FlashAttention's deep dive [00:02:18]* How the Hazy Research group collaborates across theory, systems, and applications [00:17:21]* Evaluating models beyond raw performance [00:25:00]* FlashAttention-2 [00:27:00]* CUDA and The Hardware Lottery [00:30:00]* Researching in a fast-changing market [00:35:00]* Promising transformer alternatives like state space models and RNNs [00:37:30]* The spectrum of openness in AI models [00:43:00]* Practical impact of models like LLAMA2 despite restrictions [00:47:12]* Incentives for releasing open training datasets [00:49:43]* Lightning Round [00:53:22]Transcript:Alessio: Hey everyone, welcome to the Latent Space podcast. This is Alessio, Partner and CTO-in-Residence at Decibel Partners. Today we have no Swyx, because he's in Singapore, so it's a one-on-one discussion with Tri Dao. Welcome! [00:00:24]Tri: Hi everyone. I'm Tri Dao, excited to be here. [00:00:27]Alessio: Tri just completed his PhD at Stanford a month ago. You might not remember his name, but he's one of the main authors in the FlashAttention paper, which is one of the seminal work in the Transformers era. He's got a lot of interest from efficient transformer training and inference, long range sequence model, a lot of interesting stuff. And now you're going to be an assistant professor in CS at Princeton next year. [00:00:51]Tri: Yeah, that's right. [00:00:52]Alessio: Yeah. And in the meantime, just to get, you know, a low pressure thing, you're Chief Scientist at Together as well, which is the company behind RedPajama. [00:01:01]Tri: Yeah. So I just joined this week actually, and it's been really exciting. [00:01:04]Alessio: So what's something that is not on the internet that people should know about you? [00:01:09]Tri: Let's see. When I started college, I was going to be an economist, so I was fully on board. I was going to major in economics, but the first week I was at Stanford undergrad, I took a few math classes and I immediately decided that I was going to be a math major. And that kind of changed the course of my career. So now I'm doing math, computer science, AI research. [00:01:32]Alessio: I had a similar thing. I started with physics and then I took like a programming course and I was like, I got to do computer science. I don't want to do physics. So FlashAttention is definitely, everybody's using this. Everybody loves it. You just released FlashAttention 2 last week. [00:01:48]Tri: Yeah. Early this week on Monday. Yeah. [00:01:53]Alessio: You know, AI time. Things move fast. So maybe let's run through some of the FlashAttention highlights, some of the innovation there, and then we can dive into FlashAttention 2. So the core improvement in FlashAttention is that traditional attention is a quadratic sequence length. And to the two, FlashAttention is linear, which obviously helps with scaling some of these models. [00:02:18]Tri: There are two factors there. So of course the goal has been to make attention go faster or more memory efficient. And ever since attention became popular in 2017 with the Transformer paper, lots and lots of folks have been working on this. And a lot of approaches has been focusing on approximating attention. The goal is you want to scale to longer sequences. There are tons of applications where you want to do that. But scaling to longer sequences is difficult because attention scales quadratically in sequence length on both runtime and memory, as you mentioned. So instead of trying to approximate attention, we were trying to figure out, can we do the same computation and maybe be more memory efficient? So in the end, we ended up being the memory is linear in sequence length. In terms of computation, it's still quadratic, but we managed to make it much more hardware friendly. And as a result, we do get wall clock speed up on the order of 2 to 4x, which really helps because that just means that you'll be able to train with 2 to 4x longer sequence length for the same cost without doing any approximations. As a result, lots of folks have been using this. The thing is available in a lot of libraries that do language model training or fine tuning. [00:03:32]Alessio: And the approximation thing is important because this is an exact thing versus a sparse. So maybe explain a little bit the difference there. [00:03:40]Tri: For sure. So in addition, essentially you compute pairwise similarity between every single element in a sequence against each other. So there's been other approaches where instead of doing all that pairwise computation, you only compute similarity for some pairs of elements in the sequence. So you don't do quadratic number of comparison. And this can be seen as some form of sparsity. Essentially you're ignoring some of the elements. When you write down the matrix, you essentially say, OK, I'm going to pretend there's zero. So that has some benefits in terms of runtime and memory. But the trade-off is that it tends to do worse in terms of quality because you're essentially approximating or ignoring some elements. And I personally have worked on this as well for a few years. But when we talk to practitioners who actually train models, especially at large scale, they say, tend not to use these approximate attention methods. Because it turns out, this was surprising to me at the time, was that these approximation methods, even though they perform fewer computation, they tend to not be faster in walk-on time. So this was pretty surprising because back then, I think my background was more on the theoretical side. So I was thinking of, oh, how many flops or floating point operations are you performing? And hopefully that correlates well with walk-on time. But I realized that I was missing a bunch of ideas from the system side where flops or floating point operations don't necessarily correlate with runtime. There are other factors like memory reading and writing, parallelism, and so on. So I learned a ton from just talking to systems people because they kind of figured this stuff out a while ago. So that was really eye-opening. And then we ended up focusing a lot more on memory reading and writing because that turned out to be the majority of the time when you're doing attention is reading and writing memory. [00:05:34]Alessio: Yeah, the I.O. awareness is probably one of the biggest innovations here. And the idea behind it is, like you mentioned, the FLOPS growth of the cards have been going up, but the memory bandwidth, not as much. So I think maybe that was one of the assumptions that the original attention paper had. So talk a bit about how that came to be as an idea. It's one of those things that like in insight, it's like, obviously, why are we like rewriting to like HBM every time, you know, and like once you change it, it's clear. But what was that discovery process? [00:06:08]Tri: Yeah, in hindsight, a lot of the ideas have already been there in the literature. And I would say is it was somehow at the intersection of both machine learning and systems. And you kind of needed ideas from both sides. So on one hand, on the system side, so lots of systems folks have known that, oh, you know, kernel fusion is great. Kernel fusion just means that instead of performing, you know, loading the same element, instead of performing an operation, write it down, load it back up and perform the second operation, you just load it once, perform two operations and then write it down again. So that saves you kind of memory read and write in the middle there. So kernel fusion has been a classic. There's been other techniques from the system side, like tiling, where you perform things in the form of computations in block, again, so that you can load it into a really fast memory. Think of it as a cache. And this is, again, classical computer science ideas, right? You want to use the cache. So the system folks have been thinking about these ideas for a long time, and they apply to attention as well. But there were certain things in attention that made it difficult to do a complete kernel fusion. One of which is there is this softmax operation in the middle, which requires you to essentially sum across the row of the attention matrix. So it makes it difficult to kind of break it, because there's this dependency. So it makes it difficult to break things into a block. So on the system side, people have been thinking about these ideas, but it's been difficult to kind of do kernel fusion for the entire operation. On the machine learning side, people have been thinking more algorithmically. They say, okay, either we can approximate attention, or there's this trick called the online softmax trick, which says that because of softmax, the way it's written mathematically, you can actually break it up into smaller pieces, do some rescaling, and still get the right answer. So this online softmax trick has been around for a while. I think there was a paper from NVIDIA folks back in 2018 about this. And then there was a paper from Google. So Marcus, Rob, and Stats wrote a paper late 2021 on using this online softmax trick to break attention up into smaller pieces. So a lot of the ideas were already there. But it turns out, you kind of need to combine ideas from both sides. So you need to understand that, hey, we want to do kernel fusion to reduce memory written writes. But we also need this online softmax trick to be able to break the softmax into smaller pieces so that a lot of the systems tricks kind of carry through. We saw that, and it was kind of a natural idea that we ended up using ideas from both sides, and it ended up working pretty well. Yeah. [00:08:57]Alessio: Are there any downsides to kernel fusion? If I think about databases and the reasons why we have atomic operations, you know, it's like, you have observability and fallback in between them. How does that work with attention? Is there anything that we lose by fusing the operations? [00:09:13]Tri: Yeah, I think mostly on the practical side is that you lose a little bit of flexibility in the sense that, hey, now you have, for example, faster attention, it's just a subroutine that you would call to do attention. But as a researcher, let's say you don't want that exact thing, right? You don't want just attention, let's say you want some modification to attention. You want to do, hey, I'm going to multiply the query and key, but then I'm going to do this extra thing before I carry on. So kernel fusion just means that, okay, we have a subroutine that does the entire thing. But if you want to experiment with things, you won't be able to use that fused kernel. And the answer is, can we have a compiler that then automatically does a lot of this kernel fusion? Lots of compiler folks are thinking about this, either with a new language or you can embed it in PyTorch. PyTorch folks have been working on this as well. So if you write just your code in PyTorch and they can capture the graph, can they generate code that will fuse everything together? That's still ongoing, and it works for some cases. But for attention, because of this kind of softmax rewriting stuff, it's been a little bit more difficult. So maybe in a year or two, we'll have compilers that are able to do a lot of these optimizations for you. And you don't have to, for example, spend a couple months writing CUDA to get this stuff to work. Awesome. [00:10:41]Alessio: And just to make it clear for listeners, when we say we're not writing it to memory, we are storing it, but just in a faster memory. So instead of the HBM, we're putting it in the SRAM. Yeah. [00:10:53]Tri: Yeah. [00:10:54]Alessio: Maybe explain just a little bit the difference there. [00:10:56]Tri: Yeah, for sure. This is kind of a caricature of how you think about accelerators or GPUs in particular, is that they have a large pool of memory, usually called HBM, or high bandwidth memory. So this is what you think of as GPU memory. So if you're using A100 and you list the GPU memory, it's like 40 gigs or 80 gigs. So that's the HBM. And then when you perform any operation, you need to move data from the HBM to the compute unit. So the actual hardware unit that does the computation. And next to these compute units, there are on-chip memory or SRAM, which are much, much smaller than HBM, but much faster. So the analogy there is if you're familiar with, say, CPU and RAM and so on. So you have a large pool of RAM, and then you have the CPU performing the computation. But next to the CPU, you have L1 cache and L2 cache, which are much smaller than DRAM, but much faster. So you can think of SRAM as the small, fast cache that stays close to the compute unit. Physically, it's closer. There is some kind of asymmetry here. So HBM is much larger, and SRAM is much smaller, but much faster. One way of thinking about it is, how can we design algorithms that take advantage of this asymmetric memory hierarchy? And of course, lots of folks have been thinking about this. These ideas are pretty old. I think back in the 1980s, the primary concerns were sorting. How can we sort numbers as efficiently as possible? And the motivating example was banks were trying to sort their transactions, and that needs to happen overnight so that the next day they can be ready. And so the same idea applies, which is that they have slow memory, which was hard disk, and they have fast memory, which was DRAM. And people had to design sorting algorithms that take advantage of this asymmetry. And it turns out, these same ideas can apply today, which is different kinds of memory. [00:13:00]Alessio: In your paper, you have the pyramid of memory. Just to give people an idea, when he says smaller, it's like HBM is like 40 gig, and then SRAM is like 20 megabytes. So it's not a little smaller, it's much smaller. But the throughput on card is like 1.5 terabytes a second for HBM and like 19 terabytes a second for SRAM, which is a lot larger. How do you think that evolves? So TSMC said they hit the scaling limits for SRAM, they just cannot grow that much more. HBM keeps growing, HBM3 is going to be 2x faster than HBM2, I think the latest NVIDIA thing has HBM3. How do you think about the future of FlashAttention? Do you think HBM is going to get fast enough when maybe it's not as useful to use the SRAM? [00:13:49]Tri: That's right. I think it comes down to physics. When you design hardware, literally SRAM stays very close to compute units. And so you don't have that much area to essentially put the transistors. And you can't shrink these things too much. So just physics, in terms of area, you don't have that much area for the SRAM. HBM is off-chip, so there is some kind of bus that essentially transfers data from HBM to the compute unit. So you have more area to essentially put these memory units. And so yeah, I think in the future SRAM probably won't get that much larger, because you don't have that much area. HBM will get larger and faster. And so I think it becomes more important to design algorithms that take advantage of this memory asymmetry. It's the same thing in CPU, where the cache is really small, the DRAM is growing larger and larger. DRAM could get to, I don't know, two terabytes, six terabytes, or something, whereas the cache stays at, I don't know, 15 megabytes or something like that. I think maybe the algorithm design becomes more and more important. There's still ways to take advantage of this, I think. So in the future, I think flash attention right now is being used. I don't know if in the next couple of years, some new architecture will come in and whatnot, but attention seems to be still important. For the next couple of years, I still expect some of these ideas to be useful. Not necessarily the exact code that's out there, but I think these ideas have kind of stood the test of time. New ideas like IO awareness from back in the 1980s, ideas like kernel fusions, tiling. These are classical ideas that have stood the test of time. So I think in the future, these ideas will become more and more important as we scale models to be larger, as we have more kinds of devices, where performance and efficiency become much, much more important. [00:15:40]Alessio: Yeah, and we had Jonathan Frankle on the podcast, and if you go to issattentionallyouneed.com, he has an outstanding bet, and he does believe that attention will be the state of the art architecture still in a few years. Did you think flash attention would be this popular? I'm always curious on the research side, you publish a paper, and obviously you know it's great work, but sometimes it just kind of falls flat in the industry. Could you see everybody just starting to use this, or was that a surprise to you? [00:16:11]Tri: Certainly, I didn't anticipate the level of popularity. Of course, we were extremely happy to have people using this stuff and giving us feedback and so on, and help us improve things. I think when we were writing the paper, I remember sending an email to one of my advisors, and like, hey, I'm excited about this paper, but I think the most important thing will be the artifact, which is the code. So I knew that the code will be valuable. So we kind of focus a lot on the code and make sure that the code is usable and as fast as can be. Of course, the idea, the paper presents the ideas and explain it and have experiments that validate the idea, but I knew that the artifact or the code was also pretty important. And that turned out to be the right focus, which is, you know, we put out the paper, we release the code and continue working on the code. So it's a team effort with my co-authors as well. [00:17:07]Alessio: We mentioned Hazy Research a bunch of times on the podcast before. I would love for you to spend five minutes just talking about how does the group work? How do people get together? How do you bounce ideas off of each other? Yeah. [00:17:21]Tri: So Hazy Research is a research group at Stanford led by one of my advisors, Chris Re. I love the people there. It was one of the best experiences I had. They've made my PhD so much more enjoyable. And I think there are a couple of ways that the group has been working pretty well. So one is, I think there's a diverse pool of people who either, you know, some of them focus on algorithms and theory, some of them focus on building systems, some of them focus on applications. And as a result, there is this flow of idea. So as an example, some of us were working on like more algorithms and theory, and then we can talk to the folks building systems and say, hey, let's try it out and let's put it in the systems and see how it is. And there you will get feedback from systems folks. They will say, hey, we implemented this, or we tried this and this is where it doesn't work, something like that. And once we put it in the systems, the application folks can use the algorithm or new methods or new models. And we again get great feedback from them because the application folks, for example, some of my good friends, they focus on medical imaging or seizure detection. And that is the problem they care about. And if your method doesn't work on the task they care about, they will tell you. Whereas I think a lot of people in machine learning, they're a little bit more flexible. So they will be like, hey, it doesn't work on seizure detection. Let's try some other task, right? But having that direct feedback of like, hey, it doesn't work there, let's figure out why. I think that that feedback allows us to do better work. And I think that kind of process of exchanging ideas, validating it in a real system so that applications folks can try it out and give you feedback. That cycle has been very, very useful. And so that's one, having a diverse group of people. The other one is, and this is something I really appreciate from advice from Chris was try to understand the fundamental, right? And he's happy letting me go off and read some textbooks and playing with things because I think a lot of research ideas come from understanding the old literature and see how it fits with the new landscape. And so if you just new archive papers every day, that's great, but you also need to read textbooks. And that's one advice I got from Chris, which is understand the fundamentals. And I think that allows us to do more impactful work. [00:19:46]Alessio: How do you think about academia versus industry? I feel like AI / Machine Learning has been an area where up until three, four years ago, most of the cutting edge work was being done in academia. And now there's all these big industry research labs. You're obviously going to Princeton, so you're an academia believer. How should people think about where to go? Say I'm doing my master's, I have to decide between doing a PhD and going into OpenAI Anthropic. How should I decide? [00:20:15]Tri: I think they kind of play a complementary role, in my opinion. Of course, I also was considering different paths as well. So I think right now, scaling matters a lot, especially when you talk about language models and AI and so on. Scaling matters a lot. And that means that you need compute resources and you need infrastructure and you need engineers time. And so industry tends to have an advantage when it comes to scaling things. But a lot of the ideas actually came from academia. So let's take Attention, which got popular with the Transformer in 2017. Attention actually has been around for a while. So I think the first mention was in 2014, a paper from Bernadot and others and Yoshua Bengio, which is coming from academia. A lot of ideas did come from academia. And scaling things up, of course, I think OpenAI has been great at scaling things up. That was the bet that they made after, I think, GPT-2. So they saw that scaling these things up to back then was 1.5 billion parameter seemed to give you amazing capabilities. So they really committed to that. They really committed to scaling things. And that turned out to be, it's been a pretty successful bet. I think for academia, we're still trying to figure out exactly what we're doing in this shifting landscape. And so lots of folks have been focusing on, for example, evaluation. So I know the Stanford Center for Foundation Model led by Percy, they have this benchmark called HELM, which is this holistic benchmark. So trying to figure out, okay, characterizing the landscape of different kinds of models, what people should evaluate, what people should measure, and things like that. So evaluation is one role. The other one is understanding. So this has happened historically where there's been some development in the industry and academia can play a role in explaining, understanding. They have the luxury to slow down trying to understand stuff, right? So lots of paper on understanding what's really going on, probing these models, and so on. I think I'm not as familiar with the NLP literature, but my impression is there's a lot of that going on in the NLP conferences, which is understanding what these models are doing, what capabilities they have, and so on. And the third one I could see is that the academia can take more risky bets in the sense that we can work on stuff that is quite different from industry. I think industry, my impression is you have some objective. You're trying to say, hey, for this quarter, we want to scale the model in this particular way. Next quarter, we want the model to have these capabilities. You're trying to get objectives that maybe, I don't know, 70% that will work out because it's important for the company's direction. I think for academia, the way things work is you have many, many researchers or PhD students, and they're kind of pursuing independent directions. And they have a little bit more flexibility on, hey, I'm going to try out this seemingly crazy idea and see, let's say there's a 30% chance of success or something. And however you define success, for academia, a lot of the time, success just means like, hey, we found something interesting. That could eventually go into industry through collaboration and so on. So I do see academia and industry kind of playing complementary roles. And as for someone choosing a career, I think just more and more generally, industry would be probably better in terms of compensation, in terms of probably work-life balance. But my biased perspective is that maybe academia gives you a little bit more freedom to think and understand things. So it probably comes down to personal choice. I end up choosing to be a professor next year at Princeton. But of course, I want to maintain a relationship with industry folks. I think industry folks can provide very valuable feedback to what we're doing in academia so that we understand where the field is moving because some of the directions are very much influenced by what, for example, OpenAI or Google is doing. So we want to understand where the field is moving. What are some promising applications? And try to anticipate, okay, if the field is moving like this, these applications are going to be popular. What problems will be important in two, three years? And then we try to start thinking about those problems so that hopefully in two, three years, we have some of the answers to some of these problems in two, three years. Sometimes it works out, sometimes it doesn't. But as long as we do interesting things in academia, that's the goal. [00:25:03]Alessio: And you mentioned the eval side. So we did a Benchmarks 101 episode. And one of the things we were seeing is sometimes the benchmarks really influence the model development. Because obviously, if you don't score well on the benchmarks, you're not going to get published and you're not going to get funded. How do you think about that? How do you think that's going to change now that a lot of the applications of these models, again, is in more narrow industry use cases? Do you think the goal of the academia eval system is to be very broad and then industry can do their own evals? Or what's the relationship there? [00:25:40]Tri: Yeah, so I think evaluation is important and often a little bit underrated. So it's not as flashy as, oh, we have a new model that can do such and such. But I think evaluation, what you don't measure, you can't make progress on, essentially. So I think industry folks, of course, they have specific use cases that their models need to do well on. And that's what they care about. Not just academia, but other groups as well. People do understand what are some of the emerging use cases. So for example, now one of the most popular use cases is Chatbot. And then I think folks from Berkeley, some of them are from Berkeley, call them MLCs. They set up this kind of Chatbot arena to essentially benchmark different models. So people do understand what are some of the emerging use cases. People do contribute to evaluation and measurement. And as a whole, I think people try to contribute to the field and move the field forward, albeit that maybe slightly different directions. But we're making progress and definitely evaluation and measurement is one of the ways you make progress. So I think going forward, there's still going to be just more models, more evaluation. We'll just have better understanding of what these models are doing and what capabilities they have. [00:26:56]Alessio: I like that your work has been focused on not making benchmarks better, but it's like, let's just make everything faster. So it's very horizontal. So FlashAttention 2, you just released that on Monday. I read in the blog post that a lot of the work was also related to some of the NVIDIA library updates. Yeah, maybe run us through some of those changes and some of the innovations there. Yeah, for sure. [00:27:19]Tri: So FlashAttention 2 is something I've been working on for the past couple of months. So the story is the NVIDIA CUTLASS team, they released a new version of their library, which contains all these primitives to allow you to do matrix multiply or memory loading on GPU efficiently. So it's a great library and I built on that. So they released their version 3 back in January and I got really excited and I wanted to play with that library. So as an excuse, I was just like, okay, I'm going to refactor my code and use this library. So that was kind of the start of the project. By the end, I just ended up working with the code a whole lot more and I realized that, hey, there are these inefficiencies still in Flash Attention. We could change this way or that way and make it, in the end, twice as fast. But of course, building on the library that the NVIDIA folks released. So that was kind of a really fun exercise. I was starting out, it's just an excuse for myself to play with the new library. What ended up was several months of improvement, improving Flash Attention, discovering new ideas. And in the end, we managed to make it 2x faster and now it's pretty close to probably the efficiency of things like matrix multiply, which is probably the most optimized subroutine on the planet. So we're really happy about it. The NVIDIA Cutlass team has been very supportive and hopefully in the future, we're going to collaborate more. [00:28:46]Alessio: And since it's an NVIDIA library, can you only run this on CUDA runtimes? Or could you use this and then run it on an AMD GPU? [00:28:56]Tri: Yeah, so it's an NVIDIA library. So right now, the code we release runs on NVIDIA GPUs, which is what most people are using to train models. Of course, there are emerging other hardware as well. So the AMD folks did implement a version of Flash Attention, I think last year as well, and that's also available. I think there's some implementation on CPU as well. For example, there's this library, ggml, where they implemented the same idea running on Mac and CPU. So I think that kind of broadly, the idea would apply. The current implementation ended up using NVIDIA's library or primitives, but I expect these ideas to be broadly applicable to different hardware. I think the main idea is you have asymmetry in memory hierarchy, which tends to be everywhere in a lot of accelerators. [00:29:46]Alessio: Yeah, it kind of reminds me of Sara Hooker's post, like the hardware lottery. There could be all these things that are much better, like architectures that are better, but they're not better on NVIDIA. So we're never going to know if they're actually improved. How does that play into some of the research that you all do too? [00:30:04]Tri: Yeah, so absolutely. Yeah, I think Sara Hooker, she wrote this piece on hardware lottery, and I think she captured really well of what a lot of people have been thinking about this. And I certainly think about hardware lottery quite a bit, given that I do some of the work that's kind of really low level at the level of, hey, we're optimizing for GPUs or NVIDIA GPUs and optimizing for attention itself. And at the same time, I also work on algorithms and methods and transformer alternatives. And we do see this effect in play, not just hardware lottery, but also kind of software framework lottery. You know, attention has been popular for six years now. And so many kind of engineer hours has been spent on making it as easy and efficient as possible to run transformer, right? And there's libraries to do all kinds of tensor parallel, pipeline parallel, if you use transformer. Let's say someone else developed alternatives, or let's just take recurrent neural nets, like LSTM, GRU. If we want to do that and run that efficiently on current hardware with current software framework, that's quite a bit harder. So in some sense, there is this feedback loop where somehow the model architectures that take advantage of hardware become popular. And the hardware will also kind of evolve to optimize a little bit for that kind of architecture and software framework will also evolve to optimize for that particular architecture. Right now, transformer is the dominant architecture. So yeah, I'm not sure if there is a good way out of this. Of course, there's a lot of development. Things like, I think compilers will play a role because compilers allow you to maybe still be much more efficient across different kinds of hardware because essentially you write the same code and compiler will be able to make it run efficiently different kinds of hardware. So for example, there's this language Mojo, they're compiler experts, right? And their bet is AI models will be running on different kinds of devices. So let's make sure that we have really good compilers with a good language that then the compiler can do a good job optimizing for all kinds of devices. So that's maybe one way that you can get out of this cycle. But yeah, I'm not sure of a good way. In my own research, I have to think about both the algorithm new model and how it maps to hardware. So there are crazy ideas that seem really good, but will be really, really difficult to run efficiently. And so as a result, for example, we can't really scale some of the architectures up simply because they're not hardware friendly. I have to think about both sides when I'm working on new models. [00:32:50]Alessio: Yeah. Have you spent any time looking at some of the new kind of like AI chips companies, so to speak, like the Cerebras of the world? Like one of their innovations is co-locating everything on the chip. So you remove some of this memory bandwidth issue. How do you think about that? [00:33:07]Tri: Yeah, I think that's an interesting bet. I think Tesla also has this Dojo supercomputer where they try to have essentially as fast on-chip memory as possible and removing some of these data transfer back and forth. I think that's a promising direction. The issues I could see, you know, I'm definitely not a hardware expert. One issue is the on-chip memory tends to be really expensive to manufacture, much more expensive per gigabyte compared to off-chip memory. So I talked to, you know, some of my friends at Cerebros and, you know, they have their own stack and compiler and so on, and they can make it work. The other kind of obstacle is, again, with compiler and software framework and so on. For example, if you can run PyTorch on this stuff, lots of people will be using it. But supporting all the operations in PyTorch will take a long time to implement. Of course, people are working on this. So I think, yeah, we kind of need these different bets on the hardware side as well. Hardware has, my understanding is, has a kind of a longer time scale. So you need to design hardware, you need to manufacture it, you know, maybe on the order of three to five years or something like that. So people are taking different bets, but the AI landscape is changing so fast that it's hard to predict, okay, what kind of models will be dominant in, let's say, three or five years. Or thinking back five years ago, would we have known that Transformer would have been the dominant architecture? Maybe, maybe not, right? And so different people will make different bets on the hardware side. [00:34:39]Alessio: Does the pace of the industry and the research also influence the PhD research itself? For example, in your case, you're working on improving attention. It probably took you quite a while to write the paper and everything, but in the meantime, you could have had a new model architecture come out and then it's like nobody cares about attention anymore. How do people balance that? [00:35:02]Tri: Yeah, so I think it's tough. It's definitely tough for PhD students, for researchers. Given that the field is moving really, really fast, I think it comes down to understanding fundamental. Because that's essentially, for example, what the PhD allows you to do. It's been a couple of years understanding the fundamentals. So for example, when I started my PhD, I was working on understanding matrix vector multiply, which has been a concept that's been around for hundreds of years. We were trying to characterize what kind of matrices would have theoretically fast multiplication algorithm. That seems to have nothing to do with AI or anything. But I think that was a time when I developed mathematical maturity and research taste and research skill. The research topic at that point didn't have to be super trendy or anything, as long as I'm developing skills as a researcher, I'm making progress. And eventually, I've gotten quite a bit better in terms of research skills. And that allows, for example, PhD students later in their career to quickly develop solutions to whatever problems they're facing. So I think that's just the natural arc of how you're being trained as a researcher. For a lot of PhD students, I think given the pace is so fast, maybe it's harder to justify spending a lot of time on the fundamental. And it's tough. What is this kind of explore, exploit kind of dilemma? And I don't think there's a universal answer. So I personally spend some time doing this kind of exploration, reading random textbooks or lecture notes. And I spend some time keeping up with the latest architecture or methods and so on. I don't know if there's a right balance. It varies from person to person. But if you only spend 100% on one, either you only do exploration or only do exploitation, I think it probably won't work in the long term. It's probably going to have to be a mix and you have to just experiment and kind of be introspective and say, hey, I tried this kind of mixture of, I don't know, one exploration paper and one exploitation paper. How did that work out for me? Should I, you know, having conversation with, for example, my advisor about like, hey, did that work out? You know, should I shift? I focus more on one or the other. I think quickly adjusting and focusing on the process. I think that's probably the right way. I don't have like a specific recommendation that, hey, you focus, I don't know, 60% on lecture notes and 40% on archive papers or anything like that. [00:37:35]Alessio: Let's talk about some Transformer alternatives. You know, say Jonathan Franco loses his bet and Transformer is not the state of the art architecture. What are some of the candidates to take over? [00:37:49]Tri: Yeah, so this bet is quite fun. So my understanding is this bet between Jonathan Franco and Sasha Rush, right? I've talked to Sasha a bunch and I think he recently gave an excellent tutorial on Transformer alternatives as well. So I would recommend that. So just to quickly recap, I think there's been quite a bit of development more recently about Transformer alternatives. So architectures that are not Transformer, right? And the question is, can they do well on, for example, language modeling, which is kind of the application that a lot of people care about these days. So there are methods based on state space methods that came out in 2021 from Albert Gu and Curran and Chris Re that presumably could do much better in terms of capturing long range information while not scaling quadratically. They scale sub-quadratically in terms of sequence length. So potentially you could have a much more efficient architecture when sequence length gets really long. The other ones have been focusing more on recurrent neural nets, which is, again, an old idea, but adapting to the new landscape. So things like RWKV, I've also personally worked in this space as well. So there's been some promising results. So there's been some results here and there that show that, hey, these alternatives, either RNN or state space methods, can match the performance of Transformer on language modeling. So that's really exciting. And we're starting to understand on the academic research side, we want to understand, do we really need attention? I think that's a valuable kind of intellectual thing to understand. And maybe we do, maybe we don't. If we want to know, we need to spend serious effort on trying the alternatives. And there's been folks pushing on this direction. I think RWKV scale up to, they have a model at 14 billion that seems pretty competitive with Transformer. So that's really exciting. That's kind of an intellectual thing. We want to figure out if attention is necessary. So that's one motivation. The other motivation is Transformer Alternative could have an advantage in practice in some of the use cases. So one use case is really long sequences. The other is really high throughput of generation. So for really long sequences, when you train with Transformer, with flash attention and so on, the computation is still quadratic in the sequence length. So if your sequence length is on the order of, I don't know, 16K, 32K, 100K or something, which some of these models have sequence length 100K, then you do get significantly slower in terms of training, also in terms of inference. So maybe these alternative architectures could scale better in terms of sequence length. I haven't seen actual validation on this. Let's say an RNN model release with context length, I don't know, 100K or something. I haven't really seen that. But the hope could be that as we scale to long sequences, these alternative architectures could be more well-suited. Not just text, but things like high resolution images, audio, video, and so on, which are emerging applications. So that's one, long sequences. Number two is a high throughput generation, where I can imagine scenarios where the application isn't like an interactive chatbot, but let's say a company wants to batch as many requests as possible on their server, or they're doing offline processing, they're generating stuff based on their internal documents, that you need to process in batch. And the issue with Transformer is that during generation, it essentially needs to keep around all the previous history. It's called the KV cache. And that could take a significant amount of memory, so you can't really batch too much because you run out of memory. I am personally bullish on RNNs. I think RNNs, they essentially summarize the past into a state vector that has fixed size, so the size doesn't grow with the history. So that means that you don't need as much memory to keep around all the previous tokens. And as a result, I think you can scale to much higher batch sizes. And as a result, you can make much more efficient use of the GPUs or the accelerator, and you could have much higher generation throughput. Now, this, I don't think, has been validated at scale. So as a researcher, I'm bullish on this stuff because I think in the next couple of years, these are use cases where these alternatives could have an advantage. We'll just kind of have to wait and see to see if these things will happen. I am personally bullish on this stuff. At the same time, I also spend a bunch of time making attention as fast as possible. So maybe hatching and playing both sides. Ultimately, we want to understand, as researchers, we want to understand what works, why do the models have these capabilities? And one way is, let's push attention to be as efficient as possible. On the other hand, let's push other alternatives to be as efficient at scale, as big as possible, and so that we can kind of compare them and understand. Yeah, awesome. [00:43:01]Alessio: And I think as long as all of this work happens and open, it's a net positive for everybody to explore all the paths. Yeah, let's talk about open-source AI. Obviously, together, when Red Pajama came out, which was an open clone of the LLAMA1 pre-training dataset, it was a big thing in the industry. LLAMA2 came out on Tuesday, I forget. And this week, there's been a lot of things going on, which they call open-source, but it's not really open-source. Actually, we wrote a post about it that was on the front page of Hacker News before this podcast, so I was frantically responding. How do you think about what open-source AI really is? In my mind, in open-source software, we have different levels of open. So there's free software, that's like the GPL license. There's open-source, which is Apache, MIT. And then there's kind of restricted open-source, which is the SSPL and some of these other licenses. In AI, you have the open models. So Red Pajama is an open model because you have the pre-training dataset, you have the training runs and everything. And then there's obviously RandomLens that doesn't make it one-to-one if you retrain it. Then you have the open-weights model that's kind of like StableLM, where the weights are open, but the dataset is not open. And then you have LLAMA2, which is the dataset is not open, the weights are restricted. It's kind of like not really open-source, but open enough. I think it's net positive because it's like $3 million of flops donated to the public. [00:44:32]Tri: How do you think about that? [00:44:34]Alessio: And also, as you work together, what is your philosophy with open-source AI? Right, right. [00:44:40]Tri: Yeah, I think that's a great question. And I think about it on maybe more practical terms. So of course, Meta has done an amazing job training LLAMA1, LLAMA2. And for LLAMA2, they make it much less restrictive compared to LLAMA1. Now you can use it for businesses, unless you are a monthly active user or something like that. I think just this change will have a very significant impact in the kind of landscape of open-source AI, where now lots of businesses, lots of companies will be using, I expect will be using things like LLAMA2. They will fine-tune on their own dataset. They will be serving variants or derivatives of LLAMA2. Whereas before, with LLAMA1, it was also a really good model, but your business companies weren't allowed to do that. So I think on a more practical term, it's kind of shifting the balance between a closed-source model like OpenAI and Anthropic and Google, where you're making API calls, right? And maybe you don't understand as much of what the model is doing, how the model is changing, and so on. Versus now, we have a model with open weight that is pretty competitive from what I've seen in terms of benchmarks, pretty competitive with GPT 3.5, right? And if you fine-tune it on your own data, maybe it's more well-suited for your own data. And I do see that's going to shift the balance of it. More and more folks are going to be using, let's say, derivatives of LLAMA2. More and more folks are going to fine-tune and serve their own model instead of calling an API. So that shifting of balance is important because in one way, we don't want just a concentration of decision-making power in the hands of a few companies. So I think that's a really positive development from Meta. Of course, training the model takes a couple of millions of dollars, but engineers have and I'm sure they spend tons of time trying many, many different things. So the actual cost is probably way more than that. And they make the weights available and they allow probably a lot of companies are going to be using this. So I think that's a really positive development. And we've also seen amazing progress on the open source community where they would take these models and they either fine-tune on different kinds of data sets or even make changes to the model. So as an example, I think for LLAMA1, the context lane was limited to 2K. Like a bunch of folks figured out some really simple methods to scale up to like 8K. [00:47:12]Alessio: Like the RoPE. [00:47:13]Tri: Yes. I think the open source community is very creative, right? And lots of people. LLAMA2 will, again, kind of accelerate this where more people will try it out. More people will make tweaks to it and make a contribution and then so on. So overall, I think I see that as still a very positive development for the field. And there's been lots of libraries that will allow you to host or fine-tune these models, like even with quantization and so on. Just a couple of hours after LLAMA2 was released, tons of companies announcing that, hey, it's on our API or hosting and so on and together did the same. So it's a very fast-paced development and just kind of a model with available weights that businesses are allowed to use. I think that alone is already a very positive development. At the same time, yeah, we can do much better in terms of releasing data sets. Data sets tend to be... Somehow people are not incentivized to release data sets. So philosophically, yeah, you want to be as open as possible. But on a practical term, I think it's a little bit harder for companies to release data sets. Legal issues. The data sets released tend to be not as eye-catchy as the model release. So maybe people are less incentivized to do that. We've seen quite a few companies releasing data sets together. Released a red pajama data set. I think Cerebus then worked on that and deduplicate and clean it up and release slim pajama and so on. So we're also seeing positive development on that front, kind of on the pre-training data set. So I do expect that to continue. And then on the fine-tuning data set or instruction tuning data set, I think we now have quite a few open data sets on instruction tuning and fine-tuning. But these companies do pay for human labelers to annotate these instruction tuning data set. And that is expensive. And maybe they will see that as their competitive advantage. And so it's harder to incentivize these companies to release these data sets. So I think on a practical term, we're still going to make a lot of progress on open source AI, on both the model development, on both model hosting, on pre-training data set and fine-tuning data set. Right now, maybe we don't have the perfect open source model since all the data sets are available. Maybe we don't have such a thing yet, but we've seen very fast development on the open source side. I think just maybe this time last year, there weren't as many models that are competitive with, let's say, ChatGPT. [00:49:43]Alessio: Yeah, I think the open data sets have so much more impact than open models. If you think about Elusive and the work that they've done, GPT-J was great, and the Pythia models are great, but the Pyle and the Stack, everybody uses them. So hopefully we get more people to contribute time to work on data sets instead of doing the 100th open model that performs worse than all the other ones, but they want to say they released the model. [00:50:14]Tri: Yeah, maybe the question is, how do we figure out an incentive structure so that companies are willing to release open data sets? And for example, it could be like, I think some of the organizations are now doing this where they are asking volunteers to annotate and so on. And maybe the Wikipedia model of data set, especially for instruction tuning, could be interesting where people actually volunteer their time and instead of editing Wikipedia, add annotation. And somehow they acknowledge and feel incentivized to do so. Hopefully we get to that kind of level of, in terms of data, it would be kind of like Wikipedia. And in terms of model development, it's kind of like Linux where people are contributing patches and improving the model in some way. I don't know exactly how that's going to happen, but based on history, I think there is a way to get there. [00:51:05]Alessio: Yeah, I think the Dolly-15K data set is a good example of a company saying, let's do this smaller thing, just make sure we make it open. We had Mike Conover from Databricks on the podcast, and he was like, people just bought into it and leadership was bought into it. You have companies out there with 200,000, 300,000 employees. It's like, just put some of them to label some data. It's going to be helpful. So I'm curious to see how that evolves. What made you decide to join Together? [00:51:35]Tri: For Together, the focus has been focusing a lot on open source model. And I think that aligns quite well with what I care about, of course. I also know a bunch of people there that I know and trust, and I'm excited to work with them. Philosophically, the way they've been really open with data set and model release, I like that a lot. Personally, for the stuff, for example, the research that I've developed, like we also try to make code available, free to use and modify and so on, contributing to the community. That has given us really valuable feedback from the community and improving our work. So philosophically, I like the way Together has been focusing on open source model. And the nice thing is we're also going to be at the forefront of research and the kind of research areas that I'm really excited about, things like efficient training and inference, aligns quite well with what the company is doing. We'll try our best to make things open and available to everyone. Yeah, but it's going to be fun being at the company, leading a team, doing research on the topic that I really care about, and hopefully we'll make things open to benefit the community. [00:52:45]Alessio: Awesome. Let's jump into the lightning round. Usually, I have two questions. So one is on acceleration, one on exploration, and then a takeaway. So the first one is, what's something that already happened in AI machine learning that you thought would take much longer than it has? [00:53:01]Tri: I think understanding jokes. I didn't expect that to happen, but it turns out scaling model up and training lots of data, the model can now understand jokes. Maybe it's a small thing, but that was amazing to me. [00:53:16]Alessio: What about the exploration side? What are some of the most interesting unsolved questions in the space? [00:53:22]Tri: I would say reasoning in the broad term. We don't really know how these models do. Essentially, they do something that looks like reasoning. We don't know how they're doing it. We have some ideas. And in the future, I think we will need to design architecture that explicitly has some kind of reasoning module in it if we want to have much more capable models. [00:53:43]Alessio: What's one message you want everyone to remember today? [00:53:47]Tri: I would say try to understand both the algorithm and the systems that these algorithms run on. I think at the intersection of machine learning system has been really exciting, and there's been a lot of amazing results at this intersection. And then when you scale models to large scale, both the machine learning side and the system side really matter. [00:54:06]Alessio: Awesome. Well, thank you so much for coming on 3. [00:54:09]Tri: This was great. Yeah, this has been really fun. [00:54:11] Get full access to Latent Space at www.latent.space/subscribe

Emerging technologies are proving to be very beneficial to the National Oceanic Atmospheric Administration when it comes to climate modeling, behavior analytics and its overall mission. Not only has automation technologies played a key role in NOAA's weather forecasting and environmental monitoring, but also machine learning has been a huge help in the areas of threat detection and vulnerability assessment. Longstanding cyber leader Chi Kang, deputy director for operations in NOAA's Cyber Security Division, highlights some of NOAA's cyber modernization goals for this year including how the agency is working to attract the best cyber talent and moving closer toward a zero-trust architecture.

Today we have an amazing episode.  I interviewed Tim Tutt, the co-founder and CEO of Night Shift Development. We chatted about his background and journey to becoming a tech founder, his work at Nightshift, data analytics, AI, Machine learning, and diversity in tech and cybersecurity.  I hope you enjoy our conversation.  You can visit the show's website at www.nothingaboutyou.com

Discover the immense possibilities of Generative AI in this episode of the AI Leaders Podcast. Join us as we delve into the critical considerations for successful implementation and unlock its transformative potential. Featuring Bratin Saha, VP and GM of AI & Machine Learning at Amazon, and Teresa Tung, Cloud First Chief Technologist at Accenture. Don't miss out on this insightful conversation. 

Craig discusses his key learnings from the North American Business Summit that he recently attended. The summit brought together CEOs and business leaders from Canada, the US, and Mexico to discuss pressing issues in North American businesses. First, it is essential to test our assumptions in today's fast-paced world. As things change rapidly, it is important to evaluate our beliefs and hypotheses to determine their validity and whether they serve us well. Continuous learning is crucial to update our thinking and remain open to new information and experiences. Second, instead of taking an either-or approach, it is beneficial to look for opportunities that balance and integrate different priorities. This mindset allows for more compelling and valuable decision-making. Third, there is an increasing need to listen to oneself and be aware of physiological, emotional and psychological data. Listening to others and engaging in collaborative conversations is essential for understanding different perspectives and making informed decisions, especially in complex situations. The summit's overarching theme of "stronger together" highlights the significance of human connection and cooperation. Doing good and performing well are not competing priorities. Business leaders can achieve positive results while also making a positive impact on their communities. More of Do Good to Lead Well: Website: https://craigdowden.com/ LinkedIn: https://www.linkedin.com/in/craigdowden/ --- Send in a voice message: https://podcasters.spotify.com/pod/show/craig-dowden/message

در قسمت صد و نهم دنتکست در مورد هوش مصنوعی،ماشین لرنینگ و دیپ لرنینگ به زبان ساده و کاربردشون در دندانپزشکی صحبت میکنیم . در انتها هم یک road map برای دندانپزشکان علاقمند به فعالیت در این زمینه ارائه میکنیم.اگر به دندانپزشکی و حوزه ی تکنولوژی علاقمندید این قسمت میتونه براتون جذاب باشه.

Caris Precision Oncology Alliance™ Chairman, Dr. Chadi Nabhan, sits down with Dr. JJ Gao, VP, Translational Informatics at Caris Life Sciences. Together they discuss his role in leading the Translational Informatics team at Caris Life Sciences and how artificial intelligence and bioinformatics are utilized to synthesize data to support cancer research and improve the outcomes of patients. For more information, please visit: www.CarisLifeSciences.com

According to a statement issued by the company Thursday, the 'new enforcement will reduce current calling rate by at least 50% & control current incidence effectively'.  

In this episode of ETF Battles, Ron DeLegge @etfguide referees an audience requested quadruple header between AI Machine Learning Robotics ETFs. Program judges Mike Akins at ETFAction.com and Shana Sissel at Banríon Capital Management analyze four AI Machine Learning ETFs from First Trust (ROBT), Global X (BOTZ), Blackrock (IRBO) and Robo Global (ROBO). Which is the best choice?Each ETF is judged against the other in key categories like cost, exposure strategy, performance and a mystery category. Find out who wins the battle!*********ETF Battles is sponsored by Direxion Investments Direxion Daily Leveraged & Inverse ETFs. Know the risks. Proceed Boldly. Visit http://www.Direxion.com 

Dr. Sean Humbert is back to provide an update on our ski testing at Blister Labs; discuss the potential of artificial intelligence and machine learning for product reviewing and ski recommendation engines; and more.TOPICS & TIMES:Skier, Biker, Mechanical Engineer (4:00)Update on Blister Labs Ski Testing (5:35)Artificial Intelligence & Ski Testing (9:48)Ski Recommendation Engines (16:20)Other Labs Initiatives + AI Engine Compatibility (21:42)Crashes & Close Calls (26:00)What We're Celebrating (29:51)RELATED LINKS:GEAR:30 ep.198: Sean Humbert on Ski Testing & Dynamic ModellingGEAR:30 ep.213: Manufacturers vs. Researchers & ReviewersBlister + Spot MembershipBlister Summit 2023 RegistrationOUR OTHER PODCASTSHappy HourBlister PodcastBikes & Big Ideas CRAFTEDOff The Couch Hosted on Acast. See acast.com/privacy for more information.