Forward Future Live | 9/26/25

The podcast features insights from venture capitalists and AI researchers on the transformative impact of AI on venture investing, the economy, and daily life, emphasizing the shift towards proactive, personalized AI and the need for new economic models.

• AI is creating an era of "mass cognition," shifting value to unique, problem-solving applications.

• Proactive, personalized AI is becoming mainstream, with models capable of self-correction and extended reasoning.

• New economic models are needed to manage AI's disruptive impact, focusing on open-source, ethical AI for societal benefit.

AI is ushering in an era of "mass cognition," shifting value from abundant computing power to scarce, insightful applications. Experts discuss how AI's growing capabilities, including achieving perfect scores in complex coding competitions, necessitate rethinking economic structures and investment strategies, with a strong focus on open-source, personalized AI to ensure widespread societal benefit.

AI's Economic & Investment Impact

• 00:18:48 Mike Maples, a founding partner at Floodgate, notes that just as mass computation led to the value of software and mass connectivity led to network effects, AI, as an era of "mass cognition," will make intelligence abundant. The critical investment question now is what becomes scarce and valuable when intelligence itself becomes asymptotically free, leading to a focus on high-agency applications and verticalized AI companies that master cross-discipline knowledge to solve industry-specific problems.

• 01:36:00 Immod Mosto, founder of Intelligent Internet, posits that the rapid advancement of AI, particularly in cognitive labor, will render traditional economic structures obsolete within a thousand days, a timeframe highlighted by the quick evolution since ChatGPT's launch. He stresses the urgent need for a new economic theory that accounts for AI's role, proposing models like a universal AI, open-source agents for public good (healthcare, education), and "proof of benefit" currency to combat wealth consolidation and ensure AI benefits everyone.

• 00:57:52 Ahmed El Kishki, an AI researcher at OpenAI, details his team's focus on teaching models to "think through things" to solve increasingly difficult problems, specifically in coding and software engineering, and ultimately to advance human knowledge. He predicts models will begin discovering new knowledge as early as next year, emphasizing that scaling reasoning by allowing models to think longer (sometimes for hours or days) is critical, along with making the models more efficient at using their thinking tokens.

• 00:48:04 OpenAI's new feature, "Pulse," proactively delivers daily tailored updates based on user interactions, marking a significant step towards more proactive AI. This aligns with the "sleeptime compute" concept, where AI inference runs during off-peak hours to generate higher quality, lower-latency responses. The personalization of AI models, which adapt to individual user preferences and integrate with various applications, is crucial for increasing engagement and making AI more valuable by tailoring experiences to individual needs and desires.