The MAD Podcast with Matt Turck

Misha Laskin, co-founder and CEO of Reflection AI and former DeepMind researcher, dives into the future of enterprise superintelligence. He envisions a digital brain that retains knowledge like a principal engineer. The conversation covers breakthroughs in AI, the role of coding in achieving superintelligence, and the innovative Asimov tool designed to streamline software development. Misha also shares his fascinating journey from a physics prodigy in a Manhattan-Project town to leading AI projects, and discusses the fierce talent competition in the AI startup landscape.

Emily Glassberg Sands, Head of Information at Stripe, dives into the transformative world of agentic commerce. She discusses how autonomous buying bots are set to change online shopping dynamics, emphasizing intention over clicks. Emily sheds light on Stripe's innovative AI foundation model that significantly enhances fraud detection. She also reveals insights on the rapid growth of AI startups, noting their swift international success. The discussion highlights the importance of robust data infrastructures and the shift towards usage-based pricing models in this new AI-driven landscape.

David Waltcher, a colleague at FirstMark, shares insights on the seismic shifts brought by generative AI in engineering. He discusses how tools like GitHub Copilot are transforming coding practices for millions of developers. Waltcher highlights that while AI boosts productivity by 30-50%, it introduces new security vulnerabilities and complicates traditional QA processes. He notes the rising demand for skills like code review and emphasizes the parallels between today's AI advancements and historical innovations, urging CTOs to rethink team dynamics in this evolving landscape.

Guillermo Rauch, CEO of Vercel and legendary engineer, discusses the groundbreaking text-to-app platform V0, which has generated over 100 million applications. He explains 'vibe coding,' making software accessible to non-engineers and emphasizes the importance of AI in automating development processes. Guillermo highlights Vercel's rapid growth and success with clients like Walmart and Nike, and the need for businesses to adapt in a world where the next app creators may not write any code at all. He shares insights on balancing life and tech leadership, and the evolving skills needed in software development.

Brendan Humphreys, CTO at Canva and a key player in leading 2,300 engineers, unveils the magic behind Canva’s explosive growth, achieving $3 billion in ARR. He discusses the rapid rollout of AI features like Canva Code, reflecting on their integration of diverse AI models. Brendan shares insights into the challenges of managing a global team, optimizing engineering productivity with AI tools, and maintaining company culture while scaling. He emphasizes Canva's vision for the future, blending innovation with social responsibility.

Thomas Dohmke, CEO of GitHub, shares his journey in transforming the company after its $7.5B acquisition by Microsoft into a powerhouse. He delves into the AI coding gold rush, discussing how GitHub's Copilot empowers developers, even young gamers, to code effortlessly. The conversation highlights the competitive landscape with Cursor's valuation and the rise of coding agents that could handle 90% of coding tasks by 2025. Dohmke also analyzes the shifting dynamics in SaaS and the future role of AI in streamlining coding workflows.

Aidan Gomez, CEO of Cohere and co-author of the groundbreaking Transformers paper, reveals the serendipitous origins of this pivotal AI architecture. He discusses the role of synthetic data in creating cutting-edge enterprise solutions, emphasizing productivity over AGI hype. Aidan shares insights on building secure AI for industries like finance and telecommunications while adapting to multilingual needs in markets like Japan and Korea. His candid perspective on AI's real-world implications and the evolution of research environments is both enlightening and thought-provoking.

George Sivulka, founder and CEO of Hebbia, taps into the future of work by discussing how AI can eliminate tedious tasks for finance and law professionals. He shares insights on Hebbia's Matrix platform, which automates extensive reading while minimizing AI hallucinations. George explains the shift from specialized to generalized AI, the importance of effective data management, and how 'prompting' is becoming a key management skill. Plus, he tackles the myth of job loss due to AI and champions naiveté as a strength for young founders.

Benedict Evans, a respected tech analyst and advisor, dives into the nuances of the AI landscape. He questions if the AI revolution is truly transformative or just another tech shift. Despite AI's hype, he suggests core models might be becoming commodities. Evans highlights the success of ChatGPT compared to other platforms, discusses the future of consumer AI and its impact on e-commerce, and critiques the recent decline of AI doomerism. With humor and insight, he navigates the complex realities of AI advancements and challenges.

In this discussion, Jeremy Howard, an AI scientist and co-founder of Answer AI and Fast.ai, reveals his journey from philosophy to revolutionizing AI product development with a tiny team. He shares how Answer AI is redefining R&D by focusing on real-world applications instead of chasing AGI. Topics include the rise of open-source models, the efficiency wars in AI, and innovative tools like Solve It and Shell Sage that empower users. Jeremy emphasizes the need for small, mission-driven teams to prioritize societal benefits over profits.