June 05, 2025

Elon Musk vs. Donald Trump, AI Day | Shaun Maguire, Mark Chen, Sholto Douglas, Jack Whitaker, Aarush Selvan, Michael Mignano, Oliver Cameron, Delian Asparouhov

3 hours 14 minutes

(02:28:34) - Skip to Elon Musk vs. Donald Trump Reactions

(17:15) - Shaun Maguire. Shaun is a partner at Sequoia Capital and discusses the resilience and innovation at X and XAI, highlighting the successful integration of Grok into X despite initial skepticism about the platform's stability. He compares the evolution of foundation models to operating systems, predicting a diverse ecosystem with both proprietary and open-source models, where open-source models may have broader deployment but less value capture. Maguire emphasizes the importance of early market capture and anticipates significant moats for foundation model companies due to hardware investments and application layer advantages. He also notes the rapid revenue scaling of companies like Starlink, surpassing previous benchmarks set by AWS, and underscores the necessity of a diversified energy strategy, advocating for increased natural gas, oil, solar, and nuclear energy to meet future demands.

(31:55) - Jack Whitaker. Jack is an AI expert and entrepreneur with a PhD from Cambridge University, specializing in generative AI, large language models, and multimodal systems. In the conversation, he discusses the current landscape of AI development, highlighting OpenAI's dominance in both product distribution and research, and noting Anthropic's strong position among developers. He also touches on the challenges of model naming conventions, the role of data in AI advancements, and the varying strategies of companies like Google, X.ai, and Meta in the evolving AI ecosystem.

(50:58) - Aarush Selvan. Aarush is a Product Manager at Google, leads the Gemini Deep Research project, which enables Gemini to act as a personal research assistant. In the conversation, he discusses the development of Deep Research, highlighting its ability to generate comprehensive reports by leveraging long context windows and reasoning models, and emphasizes the importance of balancing efficiency with the depth of information provided to users.

(01:04:45) - Oliver Cameron. Oliver is the co-founder and CEO of Odyssey. He discusses his transition from leading self-driving car initiatives to pioneering AI-driven storytelling. He introduces Odyssey's latest innovation, "interactive video," an AI-generated medium that allows real-time interaction without traditional game engines, envisioning it as a new form of entertainment. Cameron highlights the potential of this technology to revolutionize content creation by enabling models to generate film and game-like experiences instantly, reducing production costs and time.

(01:19:18) - Michael Mignano. Michael is a Partner at Lightspeed Venture Partners and co-founder of Anchor, and discusses the evolving dynamics between AI foundation labs and application layer startups, highlighting the shift from a symbiotic relationship to direct competition. He emphasizes the growing importance of unique data contexts, noting that models are increasingly seeking novel information, which prompts labs to compete directly with startups possessing such data. Mignano also suggests that this trend may drive startups back to established incumbents like Google and Amazon, as they might be perceived as more reliable partners in the AI ecosystem.

(01:31:44) - Mark Chen. Mark is OpenAI's Chief Research Officer. He discusses the evolving landscape of AI research, emphasizing the shift from large-scale pre-training to enhanced reasoning capabilities. He highlights the importance of reinforcement learning (RL) in developing autonomous agents and the challenges of scaling RL effectively. Chen also addresses the significance of interpretability in AI systems, advocating for models that transparently convey their reasoning processes to ensure reliability and user trust.

(02:00:59) - Sholto Douglas. Sholto is a researcher at Anthropic. He discusses the challenges and advancements in scaling reinforcement learning (RL) within artificial intelligence. He highlights the significant gains achieved by increasing compute resources in RL, noting that a tenfold increase still yields linear improvements. Douglas also addresses the complexities of reward hacking, emphasizing the need for careful guidance to align AI behaviors with human values.

(02:28:34) - Breaking News: Elon Musk vs. Donald Trump

(02:35:00) - Delian Asparouhov. Delian is the co-founder and president of Varda Space Industries and a partner at Founders Fund. He discusses the recent policy shifts in NASA's budget, particularly the reallocation of funds to the Space Launch System (SLS) program, which had been advocated for cancellation by figures like Jared Isaacman and Elon Musk. He highlights the immediate consequences of this decision, including SpaceX's announcement to decommission its Dragon spacecraft, leading to a lack of vehicles capable of servicing the International Space Station. Asparouhov also reflects on the unprecedented nature of the current dynamics between influential private sector leaders and the U.S. government, noting the escalating tensions and their potential impact on the future of space exploration.

TBPN.com is made possible by:

Ramp - https://ramp.com

Figma - https://figma.com

Vanta - https://vanta.com

Linear - https://linear.app

Eight Sleep - https://eightsleep.com/tbpn

Wander - https://wander.com/tbpn

Public - https://public.com

AdQuick - https://adquick.com

Bezel - https://getbezel.com

Numeral - https://www.numeralhq.com

Polymarket - https://polymarket.com