No Priors Ep. 44 | With Former Square CEO Alyssa Henry No Priors: AI, Machine Learning, Tech, & Startups 2023-12-14 | AI tools are helping small business owners manage their businesses, so they can stay focused on the aspects of their business they love to do. This week on No Priors, Sarah and Elad are joined by Alyssa Henry, an executive at some of the most impactful companies from Microsoft to Amazon. Most recently she was the CEO of Square. She led Square’s team as they were very early adopters of a consumer-facing product that used GPT-2 and have continued to incorporate AI into their offerings. On today’s episode, they talk about the whitespace within e-commerce for AI and lessons from the prior generation of infrastructure.Alyssa recently retired from being longtime CEO of Square, within Block. Before that she was a vice president of AWS running, amongst other things, the storage products, or the digital storage bucket for the world. And before AWS, she ran order management software at Amazon Retail and started her tech career at Microsoft. She remains on the boards of Intel, Confluent and was previously on the board of Unity.0:00 Alyssa’s experience and career trajectory2:30 Transition from engineer to manager4:09 AI implementation at Square7:46 Small business AI applications 12:14 Latent demand for content generation15:04 The origin story of Square’s GPT-2 products16:54 Consolidating ecommerce workflows18:46 How will AI change cloud services23:07 hyperscaler foundation models and the AI land grab25:16 Enterprise demand for open source models28:08 Startups in the AI semiconductor space31:02 Scale up architectures vs scaling out34:32 What’s next for Alyssa36:08 What Elad and Sarah are excited about in 2024
No Priors Ep. 86 | With Sarah Guo & Elad Gil No Priors: AI, Machine Learning, Tech, & Startups 2024-10-17 | In this episode of No Priors, Sarah and Elad explore how AI is transforming consumer apps and entertainment, with a focus on potential integrations in gaming and dating that could shift traditional societal incentives. They reflect on AI researchers winning Nobel Prizes in Science and Chemistry for the first time, discussing what this trend means for scientific discovery. The episode also covers recent AI releases, including their thoughts on OpenAI’s O1 model and Google’s NotebookLM, and examines which companies and job functions are most at risk—or resilient—in the face of AI advancements.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil Show Notes: 0:00 Introduction0:47 Google releases NotebookLM 5:20 Integrating AI into consumer apps and gaming9:11 Future of AI companionship and procreation14:45 OpenAI o1 model improves on iterative reasoning18:06 Sarah and Elad reflect on Nobel Prizes going to AI researchers21:23 Jobs and businesses at risk of disruption27:18 AI-durable companies
No Priors Ep. 85 | CEO of Braintrust Ankur Goyal No Priors: AI, Machine Learning, Tech, & Startups 2024-10-08 | Today on No Priors, Elad is joined by Ankur Goyal, founder and CEO of Braintrust. Braintrust enables companies like Notion, Airtable, Instacart, Zapier, and Vercel to deploy AI solutions at scale by efficiently evaluating and managing complex, non-deterministic AI applications. Ankur shares his insights into emerging trends in the use of AI tooling and coding languages, the rise of open-source, and the future of data infrastructure. Ankur also reflects on building resilient AI products, his philosophy on coding as a CEO, and the importance of a startup’s initial customer base. Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @AnkrgylShow Notes: 0:00 Introduction0:38 Ankur’s path to Braintrust3:05 Braintrust’s solution5:46 AI tooling trends 7:58 Instruction tuning vs. fine-tuning8:57 Open-source AI adoption 10:42 Future of data infrastructure and synthetic data14:45 Designing technical interviews18:04 Rethinking agent-based approaches19:34 Building out an AI team23:35 Typescript as the language of AI25:12 The shift away from using frameworks26:02 Vendor consolidation among enterprises 27:16 Coding as a CEO 30:16 Collaborating with customers33:00 Future of Braintrust and evals
No Priors Ep. 84 | With Chair of the Federal Trade Commission Lina Khan No Priors: AI, Machine Learning, Tech, & Startups 2024-10-03 | Lina Khan’s FTC has been the most active in decades, notably challenging tech giants and adopting a more hands-on approach to regulating the digital age. On today’s episode of No Priors, Lina Khan joins Elad and Sarah to discuss her regulatory philosophy for tech markets and what the industry can expect for future M&A deals. She shares her approach to overseeing emerging technology sectors, including AI at the model layer, and her work to ban non-competes on a federal level. Khan also offers insights into the realities of leading a government agency, the scarcity of young leaders in power, and how she measures the FTC’s impact.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @LinaKhanFTCShow Notes: 0:00 Introduction0:56 Lina Khan’s background and path to the FTC2:35 Amazon’s Antitrust Paradox4:20 Frameworks for regulating M&A in young markets8:50 Khan’s perspective on AI acquisitions12:18 What founders can expect from Khan’s M&A environment 14:55 Promoting competition at the large model layer17:01 Creating fair AI regulation18:40 FTC’s work to ban non-competes20:31 Why so few young people hold power in government today22:18 The realities of running a government agency24:20 Measuring the impact of FTC
No Priors Ep. 83 | With Rippling COO Matt MacInnis No Priors: AI, Machine Learning, Tech, & Startups 2024-09-25 | In this episode of No Priors, Sarah and Elad sit down with Matt MacInnis, COO of Rippling, to discuss the company’s unique product strategy and the advantages of being a compound startup. Matt introduces Talent Signal, Rippling’s AI-powered employee performance tool, and explains how early adopters are using it to gain a competitive edge. They explore Rippling’s approach to choosing which AI products to build and how they plan to leverage their rich data sources. The conversation also delves into how AI shapes real-world decision-making and how to realistically integrate these tools into organizational workflows.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @StanineShow Notes: 0:00 Introduction0:32 Rippling’s mission and product offerings2:13 Compound startups3:53 Evaluating human performance with Talent Signal 13:19 Incorporating AI evaluations into decision-making at Rippling14:56 Leveraging work outputs as inputs for models18:23 How Rippling chose which AI product to build first20:53 Building out bundled products23:26 Merging and scaling diverse data sources25:16 Early adopters and integrating AI into decision-making processes
No Priors Ep. 82 | With CEO of Sierra Bret Taylor No Priors: AI, Machine Learning, Tech, & Startups 2024-09-19 | Bret Taylor, Cofounder of Sierra, Chairman of the board at OpenAI, and former co-CEO of Salesforce and CTO of Facebook, joins Sarah and Elad in this week’s episode of No Priors. Bret discusses building company-branded AI agents with unique personalities, goals, and guardrails at Sierra, and their potential to revolutionize customer engagement while cutting costs. The conversation explores the next sectors for enterprise AI adoption, building resilient AI products, and the parallels between today’s AI market and the evolution of the cloud industry. Bret also shares his unique insights on future business models and upcoming technology shifts.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @BtaylorShow Notes:0:00 Intro0:42 Defining agentic systems and types of agents3:55 Customer-facing company agents5:43 Sierra AI8:11 Transforming customer service and reducing costs9:57 Challenges in implementing LLMs for company agents14:45 Drawing parallels between AI and the cloud market’s evolution17:50 Future of the AI landscape19:15 Building durable AI products24:39 Outcome-based business models and tangible ROI in AI solutions29:22 Next wave of AI sectors for enterprise adoption31:15 Customizing goals and guardrails with customers35:55 Creating distinct personalities for Sierra's agents41:05 Bret’s insights on upcoming technology and hardware shifts46:50 How AI software could enhance human agency
No Priors Ep. 81 | With Sarah Guo & Elad Gil No Priors: AI, Machine Learning, Tech, & Startups 2024-09-12 | In this episode of No Priors, Sarah and Elad go deep into what's on everyone’s mind. They break down new partnerships and consolidation in the LLM market, specialization of AI models, and AMD’s strategic moves. Plus, Elad is looking for a humanoid robot. Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil Show Notes:0:00 Introduction0:24 LLM market consolidation 2:18 Competition and decreasing API costs3:58 Innovation in LLM productization 8:20 Comparing the LLM and social network market11:40 Increasing competition in image generation13:21 Trend in smaller models with higher performance14:43 Areas of innovation17:33 Legacy of AirBnB and Uber pushing boundaries24:19 AMD Acquires ZT 25:49 Elad’s looking for a Robot
No Priors Ep. 80 | With Andrej Karpathy from OpenAI and Tesla No Priors: AI, Machine Learning, Tech, & Startups 2024-09-05 | Andrej Karpathy joins Sarah and Elad in this week of No Priors. Andrej, who was a founding team member of OpenAI and the former Tesla Autopilot leader, needs no introduction. In this episode, Andrej discusses the evolution of self-driving cars, comparing Tesla's and Waymo’s approaches, and the technical challenges ahead. They also cover Tesla’s Optimus humanoid robot, the bottlenecks of AI development today, and how AI capabilities could be further integrated with human cognition. Andrej shares more about his new mission Eureka Labs and his insights into AI-driven education and what young people should study to prepare for the reality ahead.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @KarpathyShow Notes: 0:00 Introduction0:33 Evolution of self-driving cars2:23 The Tesla vs. Waymo approach to self-driving 6:32 Training Optimus with automotive models10:26 Reasoning behind the humanoid form factor13:22 Existing challenges in robotics16:12 Bottlenecks of AI progress 20:27 Parallels between human cognition and AI models22:12 Merging human cognition with AI capabilities27:10 Building high performance small models30:33 Andrej’s current work in AI-enabled education36:17 How AI-driven education reshapes knowledge networks and status41:26 Eureka Labs42:25 What young people study to prepare for the future
No Priors Ep. 79 | With Magic.dev CEO and Co-Founder Eric Steinberger No Priors: AI, Machine Learning, Tech, & Startups 2024-08-30 | Today on No Priors, Sarah Guo and Elad Gil are joined by Eric Steinberger, the co-founder and CEO of Magic.dev. His team is developing a software engineer co-pilot that will act more like a colleague than a tool. They discussed what makes Magic stand out from the crowd of AI co-pilots, the evaluation bar for a truly great AI assistant, and their predictions on what a post-AGI world could look like if the transition is managed with care. Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @EricSteinbShow Notes: 0:00 Introduction0:45 Eric’s journey to founding Magic.dev4:01 Long context windows for more accurate outcomes10:53 Building a path toward AGI15:18 Defining what is enough compute for AGI17:34 Achieving Magic’s final UX20:03 What makes a good AI assistant22:09 Hiring at Magic27:10 Impact of AGI32:44 Eric’s north star for Magic36:09 How Magic will interact in other tools
No Priors Ep. 78 | With AWS CEO Matt Garman No Priors: AI, Machine Learning, Tech, & Startups 2024-08-29 | In this episode of No Priors, hosts Sarah and Elad are joined by Matt Garman, the CEO of Amazon Web Services. They talk about the evolution of Amazon Web Services (AWS) from its inception to its current position as a major player in cloud computing and AI infrastructure. In this episode they touch on AI commuting hardware, partnerships with AI startups, and the challenges of scaling for AI workloads.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil Show Notes: 00:00 Introduction 00:23 Matt’s early days at Amazon02:53 Early conception of AWS06:36 Understanding the full opportunity of cloud compute12:21 Blockers to cloud migration14:19 AWS reaction to Gen AI18:04 First-party models at hyperscalers20:18 AWS point of view on open source22:46 Grounding and knowledge bases26:07 Semiconductors and data center capacity for AI workloads31:15 Infrastructure investment for AI startups33:18 Value creation in the AI ecosystem36:22 Enterprise adoption 38:48 Near-future predictions for AWS usage41:25 AWS’s role for startups
No Priors Ep. 77 | With Foundry CEO and Founder Jared Quincy Davis No Priors: AI, Machine Learning, Tech, & Startups 2024-08-22 | In this episode of No Priors, hosts Sarah and Elad are joined by Jared Quincy Davis, former DeepMind researcher and the Founder and CEO of Foundry, a new AI cloud computing service provider. They discuss the research problems that led him to starting Foundry, the current state of GPU cloud utilization, and Foundry's approach to improving cloud economics for AI workloads. Jared also touches on his predictions for the GPU market and the thinking behind his recent paper on designing compound AI systems.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @jaredq_Show Notes: 00:00 Introduction 02:42 Foundry background03:57 GPU utilization for large models07:29 Systems to run a large model09:54 Historical value proposition of the cloud14:45 Sharing cloud compute to increase efficiency 19:17 Foundry’s new releases23:54 The current state of GPU capacity29:50 GPU market dynamics36:28 Compound systems design40:27 Improving open-ended tasks
No Priors Ep. 76 | With Ramp Co-Founders Eric Glyman and Karim Atiyeh No Priors: AI, Machine Learning, Tech, & Startups 2024-08-15 | In this episode of No Priors, hosts Sarah and Elad are joined by Ramp co-founders Eric Glyman and Karim Atiyeh of Ramp. The pair has been working to build one of the fastest growing fintechs since they were teenagers. This conversation focuses on how Ramp engineers have been building new systems to help every team from sales and marketing to product. They’re building best-in-class SaaS solutions just for internal use to make sure their company remains competitive. They also get into how AI will augment marketing and creative fields, the challenges of selling productivity, and how they’re using LLMs to create internal podcasts using sales calls to share what customers are saying with the whole team. Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @eglyman l @karimatiyehShow Notes: 0:00 Introduction to Ramp3:17 Working with startups8:13 Ramp’s implementation of AI14:10 Resourcing and staffing17:20 Deciding when to build vs buy21:20 Selling productivity25:01 Risk mitigation when using AI28:48 What the AI stack is missing30:50 Marketing with AI37:26 Designing a modern marketing team40:00 Giving creative freedom to marketing teams42:12 Augmenting bookkeeping47:00 AI-generated podcasts
No Priors Ep. 75 | With Co-Founder and CEO of Brex Pedro Franceschi No Priors: AI, Machine Learning, Tech, & Startups 2024-08-08 | Hunting down receipts and manually filling out invoices kills productivity. This week on No Priors, Sarah Guo and Elad Gil sit down with Pedro Franceschi, co-founder and CEO of Brex. Pedro discusses how Brex is harnessing AI to optimize spend management and automate tedious accounting and compliance tasks for teams. The conversation covers the reliability challenges in AI today, Pedro’s insights on the future of fintech in an AI-driven world, and the major transitions Brex has navigated in recent years.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @Pedroh96Show Notes: 0:00 Introduction0:32 Brex’s business and transitioning to solo CEO3:04 Building AI into Brex 7:09 Solving for risk and reliability in AI-enabled financial products11:41 Allocating resources toward AI investment14:00 Innovating data use in marketing 20:00 Building durable businesses in the face of AI25:36 AI’s impact on finance29:15 Brex’s decision to focus on startups and enterprises
No Priors Ep. 74 | With Google DeepMind VP of Research Oriol Vinyals No Priors: AI, Machine Learning, Tech, & Startups 2024-08-01 | In this episode of No Priors, hosts Sarah and Elad are joined by Oriol Vinyals, VP of Research, Deep Learning Team Lead, at Google DeepMind and Technical Co-lead of the Gemini project. Oriol shares insights from his career in machine learning, including leading the AlphaStar team and building competitive StarCraft agents. We talk about Google DeepMind, forming the Gemini project, and integrating AI technology throughout Google products. Oriol also discusses the advancements and challenges in long context LLMs, reasoning capabilities of models, and the future direction of AI research and applications. The episode concludes with a reflection on AGI timelines, the importance of specialized research, and advice for future generations in navigating the evolving landscape of AI.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @oriolvinyalsmlShow Notes: 00:00 Introduction to Oriol Vinyals00:55 The Gemini Project and Its Impact02:04 AI in Google Search and Chat Models08:29 Infinite Context Length and Its Applications14:42 Scaling AI and Reward Functions31:55 The Future of General Models and Specialization38:14 Reflections on AGI and Personal Insights43:09 Will the Next Generation Study Computer Science?45:37 Closing thoughts
No Priors Ep. 73 | With Airtable co-founder and CEO Howie Liu No Priors: AI, Machine Learning, Tech, & Startups 2024-07-25 | This week on No Priors, Sarah Guo and Elad Gil are joined by Howie Liu, the co-founder and CEO of Airtable. Howie discusses their Cobuilder launch, the evolution of Airtable from a simple productivity tool to an enterprise app platform with integrated AI capabilities. They talk about why the conventional wisdom of “app not platform” can be wrong, why there’s a future for low-code in the age of AI and code generation, and where enterprises need help adopting AI.Sign up for new podcasts every week: no-priors.comEmail feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @HowietlShow Notes: (00:00) Introduction(00:29) The Origin and Evolution of Airtable(02:31) Challenges and Successes in Building Airtable(06:09) Airtable's Transition to Enterprise Solutions(09:44) Insights on Product Management(16:23) Integrating AI into Airtable(21:55) The Future of No Code and AI(30:30) Workshops and Training for AI Adoption(36:28) The Role of Code Generation in No Code Platforms
No Priors Ep. 72 | With Sarah Guo and Elad Gil No Priors: AI, Machine Learning, Tech, & Startups 2024-07-18 | This week on No Priors, we have a host-only episode. Sarah and Elad catch up to discuss how tech history may be repeating itself. Much like in the early days of the internet, every company is clamoring to incorporate AI into their products or operations while some legacy players are skeptical that investment in AI will pay off. They also get into new opportunities and capabilities that AI is opening up, whether or not incubators are actually effective, and what companies are poised to stand the test of time in the changing tech landscape.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGilShow Notes: (0:00) Introduction(0:16) Old school operators AI misunderstandings(5:10) Tech history is repeating itself with slow AI adoption(6:09) New AI Markets(8:48) AI-backed buyouts(13:03) AI incubation(17:18) Exciting incubating applications(18:26) AI and the public markets(22:20) Staffing AI companies (25:14) Competition and shrinking head count
No Priors Ep. 71: The Best of 2024 (so far) with Sarah Guo and Elad Gil No Priors: AI, Machine Learning, Tech, & Startups 2024-07-11 | Believe or not, we’re almost halfway through 2024. Sarah and Elad have spent the first of this year talking with some of the most innovative minds in the AI industry, so we’re taking a look at some of our favorite No Priors conversations so far featuring Dylan Field (Figma); Emily Glassberg-Sands (Stripe); Brett Adcock (Figure AI); Aditya Ramesh, Tim Brooks and Bill Peebles (OpenAI’s Sora Team); Scott Wu (Cognition); and Alexandr Wang (Scale). Watch or listen to the full episodes here:Emily Glassberg Sands from Stripe: youtu.be/wiD1BfNEi-UDylan Field from Figma: youtu.be/k7F0yRs1IWYBrett Adcock from Figure: youtu.be/O3fp1Xf7ZtwAditya Ramesh, Tim Brooks and Bill Peebles from OpenAI’s Sora team: youtu.be/reMnn6bV_fICognition’s Scott Wu: youtu.be/OvBiqmcnjHYAlexandr Wang from Scale: youtu.be/2SWRU7YOd6cSign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil Show Notes: 0:00 Introduction0:46 Emily Glassberg Sands on the Future of AI and Fintech4:23 Dylan Field on AI and Human Creative Potential9:03 Brett Adcock on Running Figure AI’s Hardware and Software Processes12:43 OpenAI’s Sora Team on Artists’ Creative Experiences with their Model17:43 Scott Wu Gives Advice for Human Engineers Co-Working with AI21:06 Alexandr Wang on How Quality Data Builds Confidence in AI Systems
No Priors Ep. 70 | With Cartesia Co-Founders Karan Goel & Albert Gu No Priors: AI, Machine Learning, Tech, & Startups 2024-06-27 | This week on No Priors, Sarah Guo and Elad Gil sit down with Karan Goel and Albert Gu from Cartesia. Karan and Albert first met as Stanford AI Lab PhDs, where their lab invented Space Models or SSMs, a fundamental new primitive for training large-scale foundation models. In 2023, they Founded Cartesia to build real-time intelligence for every device. One year later, Cartesia released Sonic which generates high quality and lifelike speech with a model latency of 135ms—the fastest for a model of this class.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @krandiash | @_albertguShow Notes: 0:00 Introduction0:28 Use Cases for Cartesia and Sonic 1:32 Karan Goel & Albert Gu’s professional backgrounds5:06 State Space Models (SSMs) versus Transformer Based Architectures 11:51 Domain Applications for Hybrid Approaches 13:10 Text to Speech and Voice17:29 Data, Size of Models and Efficiency 20:34 Recent Launch of Text to Speech Product25:01 Multi-modality & Building Blocks25:54 What’s Next at Cartesia? 28:28 Latency in Text to Speech29:30 Choosing Research Problems Based on Aesthetic 31:23 Product Demo32:48 Cartesia Team & Hiring
No Priors Ep. 69 | With HeyGen CEO and Co-Founder Joshua Xu No Priors: AI, Machine Learning, Tech, & Startups 2024-06-20 | AI video generation models still have a long way to go when it comes to making compelling and complex videos but the HeyGen team are well on their way to streamlining the video creation process by using a combination of language, video, and voice models to create videos featuring personalized avatars, b-roll, and dialogue. This week on No Priors, Joshua Xu the co-founder and CEO of HeyGen, joins Sarah and Elad to discuss how the HeyGen team broke down the elements of a video and built or found models to use for each one, the commercial applications for these AI videos, and how they’re safeguarding against deep fakes. Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil l @joshua_xu_Show Notes: 0:00 Introduction3:08 Applications of AI content creation5:49 Best use cases for Hey Gen7:34 Building for quality in AI video generation11:17 The models powering HeyGen14:49 Research approach16:39 Safeguarding against deep fakes18:31 How AI video generation will change video creation24:02 Challenges in building the model26:29 HeyGen team and company
No Priors Ep. 68 | With Zapier Co-Founder and Head of AI Mike Knoop No Priors: AI, Machine Learning, Tech, & Startups 2024-06-11 | The first step in achieving AGI is nailing down a concise definition and Mike Knoop, the co-founder and Head of AI at Zapier, believes François Chollet got it right when he defined general intelligence as a system that can efficiently acquire new skills. This week on No Priors, Mike joins Elad to discuss Arc Prize which is a multi-million dollar non-profit public challenge that is looking for someone to beat the Abstraction and Reasoning Corpus (ARC) evaluation.In this episode, they also get into why Mike thinks LLMs will not get us to AGI, how Zapier is incorporating AI into their products and the power of agents, and why it’s dangerous to regulate AGI before discovering its full potential.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @mikeknoopShow Notes: 0:00 Introduction1:10 Redefining AGI2:16 Introducing ARC Prize3:08 Definition of AGI5:14 LLMs and AGI8:20 Promising techniques to developing AGI11:0 Sentience and intelligence13:51 Prize model vs investing16:28 Zapier AI innovations19:08 Economic value of agents21:48 Open source to achieve AGI24:20 Regulating AI and AGI
No Priors Ep. 67 | With Voyage AI Co-Founder and CEO No Priors: AI, Machine Learning, Tech, & Startups 2024-06-06 | After Tengyu Ma spent years at Stanford researching AI optimization, embedding models, and transformers, he took a break from academia to start Voyage AI which allows enterprise customers to have the most accurate retrieval possible through the most useful foundational data. Tengyu joins Sarah on this week’s episode of No priors to discuss why RAG systems are winning as the dominant architecture in enterprise and the evolution of foundational data that has allowed RAG to flourish. And while fine-tuning is still in the conversation, Tengyu argues that RAG will continue to evolve as the cheapest, quickest, and most accurate system for data retrieval. They also discuss methods for growing context windows and managing latency budgets, how Tengyu’s research has informed his work at Voyage, and the role academia should play as AI grows as an industry. Show Notes: 0:00 Introduction1:59 Key points of Tengyu’s research4:28 Academia compared to industry6:46 Voyage AI overview9:44 Enterprise RAG use cases15:23 LLM long-term memory and token limitations18:03 Agent chaining and data management22:01 Improving enterprise RAG 25:44 Latency budgets27:48 Advice for building RAG systems31:06 Learnings as an AI founder32:55 The role of academia in AI
No Priors Ep 66 | With Y Combinator President and CEO Garry Tan No Priors: AI, Machine Learning, Tech, & Startups 2024-05-23 | Garry Tan is a notorious founder-turned-investor who is now running one of the most prestigious accelerators in the world, Y Combinator. As the president and CEO of YC, Garry has been credited with reinvigorating the program. On this week’s episode of No Priors, Sarah, Elad, and Garry discuss the shifting demographics of YC founders and how AI is encouraging younger founders to launch companies, predicting which early stage startups will have longevity, and making YC a beacon for innovation in AI companies. They also discussed the importance of building companies in person and if San Francisco is, in fact, back. Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @garrytan - youtube.com/@GarryTanShow Notes: 0:00 Introduction0:53 Transitioning from founder to investing5:10 Early social media startups7:50 Trend predicting at YC10:03 Selecting YC founders12:06 AI trends emerging in YC batch18:34 Motivating culture at YC20:39 Choosing the startups with longevity24:01 Shifting YC found demographics29:24 Building in San Francisco 31:01 Making YC a beacon for creators33:17 Garry Tan is bringing San Francisco back
No Priors Ep. 65 | With Scale AI CEO Alexandr Wang No Priors: AI, Machine Learning, Tech, & Startups 2024-05-22 | Alexandr Wang was 19 when he realized that gathering data will be crucial as AI becomes more prevalent, so he dropped out of MIT and started Scale AI. This week on No Priors, Alexandr joins Sarah and Elad to discuss how Scale is providing infrastructure and building a robust data foundry that is crucial to the future of AI. While the company started working with autonomous vehicles, they’ve expanded by partnering with research labs and even the U.S. government. In this episode, they get into the importance of data quality in building trust in AI systems and a possible future where we can build better self-improvement loops, AI in the enterprise, and where human and AI intelligence will work together to produce better outcomes. Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @alexandr_wang0:00 Introduction3:01 Data infrastructure for autonomous vehicles5:51 Data abundance and organization12:06 Data quality and collection15:34 The role of human expertise20:18 Building trust in AI systems23:28 Evaluating AI models29:59 AI and government contracts32:21 Multi-modality and scaling challenges
No Priors Ep. 64 | With Suno CEO and Co-Founder Mikey Shulman No Priors: AI, Machine Learning, Tech, & Startups 2024-05-16 | Mikey Shulman, the CEO and co-founder of Suno, can see a future where the Venn diagram of music creators and consumers becomes one big circle. The AI music generation tool trying to democratize music has been making waves in the AI community ever since they came out of stealth mode last year. Suno users can make a song complete with lyrics, just by entering a text prompt, for example, “koto boom bap lofi intricate beats.” You can hear it in action as Mikey, Sarah, and Elad create a song live in this episode. In this episode, Elad, Sarah, And Mikey talk about how the Suno team took their experience making at transcription tool and applied it to music generation, how the Suno team evaluates aesthetics and taste because there is no standardized test you can give an AI model for music, and why Mikey doesn’t think AI-generated music will affect people’s consumption of human made music. Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @MikeyShulmanShow Notes: 0:00 Mikey’s background3:48 Bark and music generation5:33 Architecture for music generation AI6:57 Assessing music quality8:20 Mikey’s music background as an asset10:02 Challenges in generative music AI11:30 Business model14:38 Surprising use cases of Suno18:43 Creating a song on Suno live21:44 Ratio of creators to consumers25:00 The digitization of music27:20 Mikey’s favorite song on Suno29:35 Suno is hiring
No Priors Ep. 63 | With Sarah Guo and Elad Gil No Priors: AI, Machine Learning, Tech, & Startups 2024-05-09 | This week on No Priors hosts, Sarah and Elad are catching up on the latest AI news. They discuss the recent developments in AI like Meta’s new AI assistant and the latest in music generation, and if you’re interested in generative AI music, stay tuned for next week’s interview! Sarah and Elad also get into device-resident models, AI hardware, and ask just how smart smaller models can really get. These hardware constraints were compared to the hurdles AI platforms are continuing to face including computing constraints, energy consumption, context windows, and how to best integrate these products in apps that users are familiar with.Have a question for our next host-only episode, or feedback for our team? Reach out to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil Show Notes: 0:00 Intro1:25 Music AI generation4:02 Apple’s LLM11:39 The role of AI-specific hardware15:25 AI platform updates18:01 Forward thinking in investing in AI20:33 Unlimited context23:03 Energy constraints
No Priors Ep. 62 | With Cognition CEO and Co-Founder Scott Wu No Priors: AI, Machine Learning, Tech, & Startups 2024-05-02 | Scott Wu loves code. He grew up competing in the International Olympiad in Informatics (IOI) and is a world class coder, and now he's building an AI agent designed to create more, not fewer, human engineers. This week on No Priors, Sarah and Elad talk to Scott, the co-founder and CEO of Cognition, an AI lab focusing on reasoning. Recently, the Cognition team released a demo of Devin, an AI software engineer that can increasingly handle entire tasks end to end.In this episode, they talk about why the team built Devin with a UI that mimics looking over another engineer’s shoulder as they work and how this transparency makes for a better result. Scott discusses why he thinks Devin will make it possible for there to be more human engineers in the world and what will be important for software engineers to focus on as these roles evolve. They also get into how Scott thinks about building the Cognition team and that they’re just getting started. Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @ScottWu46Show Notes: 0:00 Introduction1:12 IOI training and community6:39 Cognition’s founding team8:20 Meet Devin9:17 The discourse around Devin12:14 Building Devin’s UI14:28 Devin’s strengths and weakness 18:44 The evolution of coding agents22:43 Tips for human engineers26:48 Hiring at Cognition
No Priors Ep.61 | OpenAIs Sora Leaders Aditya Ramesh, Tim Brooks and Bill Peebles No Priors: AI, Machine Learning, Tech, & Startups 2024-04-25 | AI-generated videos are not just leveled-up image generators. But rather, they could be a big step forward on the path to AGI. This week on No Priors, the team from Sora is here to discuss OpenAI’s recently announced generative video model, which can take a text prompt and create realistic, visually coherent, high-definition clips that are up to a minute long.Sora team leads, Aditya Ramesh, Tim Brooks, and Bill Peebles join Elad and Sarah to talk about developing Sora. The generative video model isn’t yet available for public use but the examples of its work are very impressive. However, they believe we’re still in the GPT-1 era of AI video models and are focused on a slow rollout to ensure the model is in the best place possible to offer value to the user and more importantly they’ve applied all the safety measures possible to avoid deep fakes and misinformation. They also discuss what they’re learning from implementing diffusion transformers, why they believe video generation is taking us one step closer to AGI, and why entertainment may not be the main use case for this tool in the future. Show Notes:0:00 Sora team Introduction1:05 Simulating the world with Sora2:25 Building the most valuable consumer product5:50 Alternative use cases and simulation capabilities8:41 Diffusion transformers explanation10:15 Scaling laws for video13:08 Applying end-to-end deep learning to video15:30 Tuning the visual aesthetic of Sora17:08 The road to “desktop Pixar” for everyone20:12 Safety for visual models22:34 Limitations of Sora25:04 Learning from how Sora is learning29:32 The biggest misconceptions about video models
No Priors Ep. 60 | With Playground AI Founder Suhail Doshi No Priors: AI, Machine Learning, Tech, & Startups 2024-04-18 | Multimodal models are making it possible to create AI art and augment creativity across artistic mediums. This week on No Priors, Sarah and Elad talk with Suhail Doshi, the founder of Playground AI, an image generator and editor. Playground AI has been open-sourcing foundation diffusion models, most recently releasing Playground V2.5. In this episode, Suhail talks with Sarah and Elad about how the integration of language and vision models enhances the multimodal capabilities, how the Playground team thought about creating a user-friendly interface to make AI-generated content more accessible, and the future of AI-powered image generation and editing.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @SuhailShow Notes: 0:00 Introduction0:52 Focusing on image generation3:01 Differentiating from other AI creative tools5:58 Training a Stable Diffusion model8:31 Long term vision for Playground AI15:00 Evolution of AI architecture17:21 Capabilities of multimodal models22:30 Parallels between audio AI tools and image-generation
No Priors Ep. 59 | With Sarah Guo & Elad Gil No Priors: AI, Machine Learning, Tech, & Startups 2024-04-11 | This week on a host-only episode of No Priors, Sarah and Elad discuss the AI wave as compared to the internet wave, the current state of AI investing, the foundation model landscape, voice and video AI, advances in agentic systems, prosumer applications, and the Microsoft/Inflection deal.Have a question for our next host-only episode or feedback for our team? Reach out to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil Show Notes: 0:00 Intro0:32 How to think about scaling in 20243:21 Microsoft/Inflection deal5:28 Voice cloning7:02 Investing climate12:50 Whitespace in AI16:36 AI video landscape19:54 Agentic user experiences22:21 Prosumer as the first wave of application AI
No Priors Ep. 58 | The argument for humanoid robots with Brett Adcock from Figure No Priors: AI, Machine Learning, Tech, & Startups 2024-04-04 | Humans are always doing work that is dull or dangerous. Brett Adcock, the founder and CEO of Figure AI, wants to build a fleet of robots that can do everything from work in a factory or warehouse to folding your laundry in the home. Today on No Priors, Sarah got the chance to talk with Brett about how a company that is only 21 months old has already built humanoid robots that not only walk the walk by performing tasks like item retrieval and making a cup of coffee but they also talk the talk through speech to speech reasoning. In this episode, Brett and Sarah discuss why right now is the correct time to build a fleet of AI robots and how implementation in industrial settings will be a stepping stone into AI robots coming into the home. They also get into how Brett built a team of world class engineers, commercial partnerships with BMW and OpenAI that are accelerating their growth, and the plan to achieve social acceptance for AI robots. Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @adcock_brettShow Notes: (0:00) Brett’s background(3:09) Figure AI Thesis(5:51) The argument for humanoid robots(7:36) Figure AI public demos(12:38) Mitigating risk factors(15:20) Designing the org chart and finding the team(16:38) Deployment timeline(20:41) Build vs buy and vertical integration(23:04) Product management at Figure(28:37) Corporate partnerships(31:58) Humans at home(33:38) Social acceptance (35:41) AGI vs the robots
No Priors Ep. 57 | With LangChain CEO and Co-Founder Harrison Chase No Priors: AI, Machine Learning, Tech, & Startups 2024-03-28 | Companies are employing AI agents and co-pilots to help their teams increase efficiency and accuracy, but developing apps that are trained properly can require a skillset many enterprise teams don’t have. This week on No Priors, Sarah and Elad are joined by Harrison Chase, the CEO and co-founder of LangChain, an open-source framework and developer toolkit that helps developers build LLM applications. In this conversation they talk about the gaps in open source app development, what it will take to keep up with private companies, the importance of creating prompts that can be compatible with many API models, and why memory is so undeveloped in this space. Show Notes: 0:00 Introduction to LangChain1:45 Managing an open source environment4:30 Developing useful AI agents10:03 Sophistication and limitations of AI app development14:17 Switching between model APIs17:10 Context windows, fine tuning and functionality21:37 Evolution of AI open source environment23:53 The next big breakthroughs
No Priors Ep 56 | With Baseten CEO and Co-Founder Tuhin Srivastava No Priors: AI, Machine Learning, Tech, & Startups 2024-03-21 | At a time when users are being asked to wait unthinkable seconds for AI products to generate art and answers, speed is what will win the battle heating up in AI computing. At least according to today’s guest, Tuhin Srivastava, the CEO and co-founder of Baseten which gives customers scalable AI infrastructures starting with interference. In this episode of No Priors, Sarah, Elad, and Tuhin discuss why efficient code solutions are more desirable than no code, the most surprising use cases for Baseten, and why all of their jobs are very defensible from AI. Show Notes: (0:00) Introduction(1:19) Capabilities of efficient code enabled development(4:11) Difference in training inference workloads(6:12) AI product acceleration(8:48) Leading on inference benchmarks at BaseTen(12:08) Optimizations for different types of models(16:11) Internal vs open source models(19:01) timeline for enterprise scale(21:53) Rethinking investment in compute spend(27:50) Defensibility in AI industries(31:30) Hardware and the chip shortage(35:47) Speed is the way to win in this industry(38:26) Wrap
No Priors Ep. 55 | With Figma CEO Dylan Field No Priors: AI, Machine Learning, Tech, & Startups 2024-03-14 | Figma has had a banner year and the formidable team isn’t slowing down—even after regulatory issues blocked the merger with Adobe. Today on No Priors, Sarah and Elad are joined by Dylan Field the CEO and founder of Figma, the design collaboration tool that is closing the gap between imagination and reality. They discuss what’s next for an independent Figma, how AI can augment design and speed up the iteration loop, and how Figma is expanding beyond design with products that help the entire product team’s workflow.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @zoinkShow Notes: 0:00 Introduction2:01 No more Adobe acquisition 4:20 What’s next for Figma7:16 FigJam, digital collaboration, and expanding beyond design10:50 Figma DevMode13:06 Incorporating AI at Figma15:03 How AI will change design19:19 Creativity augmentation and the iterative loop22:44 Automating repetitive design tasks25:35 The future of AI UI29:44 Investing philosophy31:28 Leadership evolution
No Priors Ep. 54 | With Sarah Guo & Elad Gil No Priors: AI, Machine Learning, Tech, & Startups 2024-03-07 | Host-only episode discussing NVIDIA, Meta and Google earnings, Gemini and Mistral model launches, the open-vs-closed source debate, domain specific foundation models, if we’ll see real competition in chips, and the state of AI ROI and adoption.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGilShow Notes: 0:00 Introduction0:27 Model news and product launches5:01 Google enters the competitive space with Gemini 1.58:23 Biology and robotics using LLMs10:22 Agent-centric companies14:22 NVIDIA earnings17:29 ROI in AI20:43 Impact from AI25:45 Building effective AI tools in house29:09 What would it take to compete with NVIDIA33:23 The architectural approach to compute35:42 the roadblocks to chip production in the US38:30 The virtuous tech cycles in AI
No Priors Ep. 53 | With AMD CTO Mark Papermaster No Priors: AI, Machine Learning, Tech, & Startups 2024-02-29 | Compute is the fuel for the AI revolution, and customers want more chip vendors. AMD CTO Mark Papermaster joins Sarah and Elad on No Priors to discuss AMD’s strategy, their newest GPUs, where inference workloads will live, the chip software stack, how they are thinking about supply chain issues, and what we can expect from AMD in 2024. Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGilShow Notes: 0:00 Introduction and Mark’s background2:35 AMD background and current markets4:40 AMD shifting to AI space8:54 AI applications coming out of AMD10:57 Software investment15:15 The benefits of open-source stacks16:58 Evolving GPU market20:21 Constraints on GPU production24:11 Innovations in chip technology27:57 Chip supply chain30:18 Future of innovative hardware products35:42 What’s next for AMD
No Priors Ep. 52 | With Pinecone CEO Edo Liberty No Priors: AI, Machine Learning, Tech, & Startups 2024-02-22 | Accurate, customizable search is one of the most immediate AI use cases for companies and general users. Today on No Priors, Elad and Sarah are joined by Pinecone CEO, Edo Liberty, to talk about how RAG architecture is improving syntax search and making LLMs more available. By using a RAG model Pinecone makes it possible for companies to vectorize their data and query it for the most accurate responses. In this episode, they talk about how Pinecone’s Canopy product is making search more accurate by using larger data sets in a way that is more efficient and cost effective—which was almost impossible before there were serverless options. They also get into how RAG architecture uniformly increases accuracy across the board, how these models can increase “operational sanity” in the dataset for their customers, and hybrid search models that are using keywords and embeds. Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @EdoLibertyShow Notes: 0:00 Introduction to Edo and Pinecone2:01 Use cases for Pinecone and RAG models6:02 Corporate internal uses for syntax search10:13 Removing the limits of RAG with Canopy14:02 Hybrid search16:51 Why keep Pinecone closed source22:29 Infinite context23:11 Embeddings and data leakage25:35 Fine tuning the data set27:33 What’s next for Pinecone 28:58 Separating reasoning and knowledge in AI
No Priors Ep. 51 | With Notion CEO Ivan Zhao No Priors: AI, Machine Learning, Tech, & Startups 2024-02-15 | Notion is a productivity app that has invested heavily in AI to create products that enable workers to access information instantly without having to search through their own countless notes. Today on No Priors, Sarah and Elad are joined by Ivan Zhao, the co-founder and CEO of Notion, to talk about Notions Q&A interface and calendar applications. They also get into how using RAG models means better retrieval, longer memory, and the user can be less organized and how Notion is leading the charge in this era of SaaS bundling products.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @ivanhzhaoShow Notes: 0:00 Introduction2:09 AI and Computing literacy5:39 Building the Notion AI team8:43 Notion as an application company12:09 Prioritizing AI investment14:53 The rapid evolution cycle of AI development17:46 Notion Q&A20:00 Workflow and AI for calendars22:43 Moving past the need for organization24:36 History of SaaS doesn’t repeat, it rhymes30:14 Design at Notion34:26 Notion office design36:52 How RAG will change the future38:30 Building our the software in the Notionscape
No Priors Ep. 50 | With Stripe Head of Information Emily Glassberg Sands No Priors: AI, Machine Learning, Tech, & Startups 2024-02-08 | Many companies that are building AI products for their users are not primarily AI companies. Today on No Priors, Sarah and Elad are joined by Emily Glassberg Sands who is the Head of Information at Stripe. They talk about how Stripe prioritizes AI projects and builds these tools from the inside out. Stripe was an early adopter of utilizing LLMs to help their end user. Emily talks about how they decided it was time to meaningfully invest in AI given the trajectory of the industry and the wealth of information Stripe has access to. The company’s goal with utilizing AI is to empower non-technical users to code using natural language and for technical users to be able to work much quicker and in this episode she talks about how their Radar Assistant and Sigma Assistant achieve those goals. Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @emilygsandsShow Notes: 0:00 Background0:38 Emily’s role at Stripe2:31 Adopting early gen AI models4:44 Promoting internal usage of AI8:17 Applied ML accelerator teams10:36 Radar fraud assistant13:30 Sigma assistant14:32 How will AI affect Stripe in 3 years17:00 Knowing when it’s time to invest more fully in AI18:28 Deciding how to proliferate models22:04 Whitespace for fintechs employing AI25:41 Leveraging payments data for customers27:51 Labor economics and data30:10 Macro economic trends for strategic decisions32:54 How will AI impact education35:36 Unique needs of AI startups
No Priors Ep. 49 | With Shopify VP of Core Product Glen Coates No Priors: AI, Machine Learning, Tech, & Startups 2024-01-31 | Building an ecommerce business is hard – it requires merchants to have a wealth of skills: technical, logistics, marketing, pricing, vendor management, finance and analytics. That’s why Shopify is releasing new AI features that help merchants tackle things like product descriptions, marketing suggestions and search.Today on No Priors, Glen Coates, the VP of core product at Shopify (and former founder of b2b wholesale platform Handshake), joins Sarah and Elad. They talk about the releases from Shopify Editions, why they are deploying “copilot” rather than “autopilot,” AI innovation-at-scale, how to change the basement of a house while people are living in it, and building a leadership team of entrepreneurs.Sign up for new podcasts every week. Email feedback to show@no-priors.comFollow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @glencoates0:00 Background2:22 Calling a “Code Red” at Shopify4:04 Integrating acquisitions, entrepreneurial leaders12:15 AI adoption15:51 Deciding when to ship AI products, evaluations17:33 Shopify’s risk orientation18:50 Changing the core Shopify data model, enabling AI features26:05 What’s missing from LLMs for merchants28:47 Most interesting AI developments in the industry33:22 What users want from LLMs and search38:20 No Priors social
No Priors Ep. 48 | With Covariant CEO Peter Chen No Priors: AI, Machine Learning, Tech, & Startups 2024-01-24 | Building adaptive AI models that can learn and complete tasks in the physical world requires precision but these AI robots could completely change manufacturing and logistics processes. Peter Chen, the co-founder and CEO of Covariant, leads the team that is building robots that will increase manufacturing efficiency, safety, and create warehouses of the future. Today on No Priors, Peter joins Sarah to talk about how the Covariant team is developing multimodal models that have precise grounding and understanding so they can adapt to solve problems in the physical world. They also discuss how they plan their roadmap at Covariant, what could be next for the company, and what use case will bring us to the Chat-GPT moment for AI robots.00:00 Peter Chen Background00:58 How robotics AI will drive AI forward03:00 Moving from research to a commercial company05:46 The argument for building incrementally 08:13 Manufacturing robotics today12:21 Put wall use case15:45 What’s next for Covariant Brain18:42 Covariant’s customers19:50 Grounding concepts in Ai25:47 How scaling laws apply to Covariant29:21 Covariant’s driving thesis32:54 the Chat-GPT moment for robotics35:12 Manufacturing center of the future37:02 Safety in AI robotics
No Priors Ep. 47 | With Sourcegraph CTO Beyang Liu No Priors: AI, Machine Learning, Tech, & Startups 2024-01-18 | Coding in collaboration with AI can reduce human toil in the software development process and lead to more accurate and less tedious work for coding teams. This week on No Priors, Sarah talked with Beyang Liu, the cofounder and CTO of Sourcegraph, which builds tools that help developers innovate faster. Their most recent launch was an AI coding assistant called Cody. Beyang has spent his entire career thinking about how humans can work in conjunction with AI to write better code.Sarah and Beyang talk about how Sourcegraph is thinking about augmenting the coding process in a way that ensures accuracy and efficiency starting with robust and high-quality context. They also think about what the future of software development could look like in a world where AI can generate high-quality code on its own and where that leaves humans in the coding process. 0:00 Beyang Liu’s experience0:52 Sourcegraph premise2:20 AI and finding flow4:18 Developing LLMs in code6:46 Cody explanation7:56 Unlocking AI code generation11:00 search architecture in LLMs16:02 Quality-assurance in data set18:03 Future of Cody22:48 Constraints in AI code generation30:28 Lessons from Beyang’s research days33:17 Benefits of small models35:49 Future of software development42:14 What skills will be valued down the line
No Priors Ep. 46 | Best of 2023 with Sarah Guo and Elad Gil No Priors: AI, Machine Learning, Tech, & Startups 2024-01-11 | We’re looking back on 2023 and sharing a handful of our favorite conversations. Last year was full of insightful conversations that shaped the way we think about the most innovative movements in the AI space. Want to hear more? Check out the full episodes here:What is Digital Life? with OpenAI Co-Founder & Chief Scientist Ilya Sutskever youtu.be/Ft0gTO2K85AHow AI can help small businesses with Former Square CEO Alyssa Henryyoutu.be/llMFYc4_vikWill Everyone Have a Personal AI? With Mustafa Suleyman, Founder of DeepMind and Inflectionyoutu.be/g4VszCFonPkHow will AI bring us the future of medicine? With Daphne Koller from Insitroyoutu.be/k5FvyrJdEcIThe case for AI optimism with Reid Hoffman from Inflection AIyoutu.be/_Hprred2E7MYour AI Friends Have Awoken, With Noam Shazeeryoutu.be/emCoG-hA7AEMistral 7B and the Open Source Revolution With Arthur Mensch, CEO Mistral AIyoutu.be/EMOFRDOMIiU0:00 Introduction0:27 Ilya Sutskever on the cap profit model3:11 Alyssa Henry on how AI can small business owners5:25 Mustafa Suleyman on defining intelligence8:53 Reid Hoffman’s advice for co-working with AI11:47 Daphne Koller on probabilistic graphical models13:15 Noam Shazeer on the possibilities of LLMs14:27 Arthur Mensch on keeping AI open17:19 Jensen Huang on how Nvidia decides what to work on
No Priors Ep. 45 | With Reid Hoffman No Priors: AI, Machine Learning, Tech, & Startups 2023-12-21 | AI doomerism and calls to regulate the emerging technology is at a fever pitch but today’s guest, Reid Hoffman is a vocal AI optimist who views slowing down innovation as anti-humanistic. Reid needs no introduction, he’s the co-founder of PayPal, Linkedin, and most recently Inflection AI which is building empathetic AI companions. He is also a board member at Microsoft and former board member at OpenAI. On this week’s episode, Reid joins Sarah and Elad to talk about the historical case for an optimistic outlook on emerging technology like AI, advice for workers who fear AI may replace them, and why it’s impossible to regulate before you innovate. Plus, some predictions.Aside from his storied experience in technology, Reid is an author, podcaster, and political activist. Most recently, he co-authors a book with GPT 4 called Impromptu: Amplifying Our Humanity Through AI.00:00 Reid Hoffman’s birdseye view on the state of AI03:37 AI and human collaboration in workflows5:23 What’s causing AI doomerism12:28 Advice for whitecollar workers16:45 Why Reid isn’t retiring18:25 How Inflection started22:06 Surprising ways people are using Inflection25:34 Western bias and AI ethics30:58 Structural challenges in governing AI33:15 Most exciting whitespace in AI35:00 GPT 5 and Innovations coming in the next two years44:00 What future should we be building?
No Priors Ep. 43 | With Clara Shih, CEO of Salesforce AI No Priors: AI, Machine Learning, Tech, & Startups 2023-12-07 | AI is the new UI for enterprise customers according to Clara Shih, the CEO of Salesforce AI. Salesforce released Einstein, now called Einstein GPT, in 2016, making it an early example of how beneficial AI can be when embedded in enterprise software. This week on No Priors, Sarah and Elad talked with Clara about what the evolution of AI in enterprise looks like, how Salesforce is adoption AI across the organization, and the onboarding process for companies looking to integrate AI into their workflow, plus the challenges of pricing for AI services.Clara Shih is the Chief Executive Officer of Salesforce AI where she leads the AI efforts across Salesforce including AI co-pilot and agent platform, model development, go-to-market growth, adoption, partnerships, ecosystems, and secure responsible AI. Before that was the CEO of Salesforce Service Cloud She is also the co-founder and previous CEO of Hearsay Systems. She is also on the Board of Directors at Starbucks.0:00 Clara’s Background0:50 From cloud services to AI3:25 Internal Model Development vs Open Source5:20 The Co-Pilot Approach8:50 Enterprise AI Adoption10:54 The future of Enterprise AI13:23 Cross-team collaboration14:40 AI is the new UI19:11 Structuring the Dataset21:25 What’s next for generative AI in Enterprise23:18 Pricing challenges in AI26:30 Startups and AI28:22 Collaboration in AI Industry
No Priors Ep. 42 | With Sarah Guo and Elad Gil No Priors: AI, Machine Learning, Tech, & Startups 2023-11-30 | OpenAI’s leadership has taken us all on a rollercoaster so it’s great timing for another host-only episode. This week Sarah and Elad get into what has been going on at OpenAI and what the turbulent leadership changes tell us about the importance of good intent and good incentives when building these influential companies. They also talk about innovative products coming out of Pika Labs, why people are moving away from diffusion models to LLMs, and how, in AI investing, the ASP is the opportunity. 0:00 Recapping the OpenAI saga9:56 AI video products16:14 Moving from Diffusion Models to LLMs19:47 The beneficial margins of AI investing
No Priors Ep. 41 | With Imbue Co-Founders Kanjun Qiu and Josh Albrecht No Priors: AI, Machine Learning, Tech, & Startups 2023-11-16 | 00:00 - Introduction to Imbue04:55 - The Spectrum of Agent Tasks10:23 - Specialization and Generalization With Agents14:08 - Code and Language in AI Agents21:00 - Evaluating AI Development Tools Efficiently26:39 - Prioritizing GPU Usage
No Priors Ep. 40 | With Arthur Mensch, CEO Mistral AI No Priors: AI, Machine Learning, Tech, & Startups 2023-11-09 | Open Source fuels the engine of innovation, according to Arthur Mensch, CEO and co-founder of Mistral AI. Mistral is a French AI company which recently made a splash with releasing Mistral 7B, the most powerful language model for its size to date, and outperforming much larger models. Sarah Guo and Elad Gil sit down with Arthur to discuss why open source could win the AI wars, their $100M+ seed financing, the true nature of scaling laws, why he started his company in France, and what Mistral is building next.Arthur Mensch is Chief Executive Officer and co-founder of Mistral AI. A graduate of École Polytechnique, Télécom Paris and holder of the Master Mathématiques Vision Apprentissage at Paris Saclay, he completed his thesis in machine learning for functional brain imaging at Inria (Parietal team). He spent two years as a post-doctoral fellow in the Applied Mathematics department at ENS Ulm, where he carried out work in mathematics for optimization and machine learning. In 2020, he joined DeepMind as a researcher, working on large language models, before leaving in 2023 to co-found Mistral AI with Guillaume Lample and Timothee Lacroix.00:00 - Why he co-founded Mistral04:22 - Chinchilla and Proportionality06:16 - Mistral 7b09:17 - Data and Annotations10:33 - Open Source Ecosystem17:36 - Proposed Compute and Scale Limits19:58 - Threat of Bioweapons23:08 - Guardrails and Safety29:46 - Mistral Platform31:31 - French and European AI Startups
No Priors Ep. 39 | With OpenAI Co-Founder & Chief Scientist Ilya Sutskever No Priors: AI, Machine Learning, Tech, & Startups 2023-11-02 | Each iteration of ChatGPT has demonstrated remarkable step function capabilities. But what’s next? Ilya Sutskever, Co-Founder & Chief Scientist at OpenAI, joins Sarah Guo and Elad Gil to discuss the origins of OpenAI as a capped profit company, early emergent behaviors of GPT models, the token scarcity issue, next frontiers of AI research, his argument for working on AI safety now, and the premise of Superalignment. Plus, how do we define digital life?Ilya Sutskever is Co-founder and Chief Scientist of OpenAI. He leads research at OpenAI and is one of the architects behind the GPT models. He co-leads OpenAI's new "Superalignment" project, which tries to solve the alignment of superintelligences in 4 years. Prior to OpenAI, Ilya was co-inventor of AlexNet and Sequence to Sequence Learning. He earned his Ph.D in Computer Science from the University of Toronto.00:00 - Early Days of AI Research06:49 - Origins of Open Ai & CapProfit Structure13:54 - Emergent Behaviors of GPT Models18:05 - Model Scale Over Time & Reliability23:51 - Roles & Boundaries of Open Source in the AI Ecosystem28:38 - Comparing AI Systems to Biological & Human Intelligence32:56 - Definition of Digital Life35:11 - Super Alignment & Creating Pro Human AI41:20 - Accelerating & Decelerating Forces
No Priors Ep. 38 | With Material Security Co-Founder Ryan Noon No Priors: AI, Machine Learning, Tech, & Startups 2023-10-26 | Cyber Security is going to change significantly in the era of AI, according to Ryan Noon, cofounder of Material Security, a security company that makes cloud-based Google and Microsoft email a safe place for sensitive data. Elad Gil and Ryan talk about how Material Security started to use LLMs, potential security threats from AI hacks, and the role of the government in securing the Internet. Ryan also shares his advice for founders.Ryan co-founded Material Security in 2017 after seeing high profile email hacks in the 2016 Presidential election. Previously, he led various engineering teams at Dropbox after it acquired his first company, Parastructure. Prior to Parastructure, he led engineering at a data analysis company spun out of Stanford by DARPA. He holds both an MS in Computer Networks and Security and a BS in Computer Science from Stanford.00:00 - How 2016 Election Hacking Inspired Ryan to Start Material Security05:02 - Generative AI Use Cases in Cyber Security & Fine Tuning11:50 - Predictions on Effective Threat Levels from AI Hacks15:39 - Democracy, the Department of Defence, DARPA and Cyber Security20:17 - Is there room for startups in the Cyber Security industry?27:13 - New Challenges On Horizon After 7 Years as Cofounder30:32 - Advice to Founders
No Priors Ep. 37 | With Kawal Gandhi No Priors: AI, Machine Learning, Tech, & Startups 2023-10-19 | As the Lead for Generative AI in the Office of the CTO for Google Cloud, Kawal Gandhi has a unique vantage point on enterprise AI roll out. Sarah Guo and Elad Gil sit down with Gandhi this week to discuss his insights on how enterprises can effectively invest and roll out AI development and tools, the best use cases for Google Colab TPU, and Google’s internal AI applications driving external customer success. Plus, when will email get more efficient? Kawal Gandhi works in the Office of the CTO of Google Cloud, where his main focus is AI/ML. He has worked at Google for nearly a decade in search and ad roles before focusing on the development and marketing of AI tools.00:00 - Generative AI in Google Cloud09:05 - AI Adoption for Enterprise13:31 - Multi-Modal AI Models16:19 - AI Adoption, Investment Cost, Anti-patterns24:43 - Google's TPU and NVIDIA GPU shortage31:00 - Data Marketplace and Model Training
