Independent AI researcher Mark Pesce returns to the Futurists with an update on the latest advances in open source artificial intelligence. Mark talks to co-hosts Brett King and Rob Tercek about a wide range of topics: small open-source models that run on the desktop and the smartphone; AI agents that work the “night shift”; NVIDIA’s pivot to open source AI; the disruptive threat to the leading US AI labs; the rapid rise of AI-enabled intrusions and the threat to the smartphone; personal AI filtering and assistants; AI agents to manage the smart home; faceless appliances as a precursor to liquid interfaces and the end of apps; fostering cognitive dependency and offloading cognitive work to companion apps; pacing the frontier; super intelligence; geopolitical rivalry.

Dan Schwarz is the CEO and co-founder of FutureSearch.ai, a startup venture that intends to become the leader in AI-powered forecasting. Dan points out that hybrid systems that combine human forecasting with machine intelligence have already surpassed human-only prognostication, but the AI-only approaches still fall short. FutureSearch has begun to benchmark several leading AI systems on their ability to provide accurate deep research. Dan tells TheFuturists about the weaknesses in human forecasting and the areas where he sees opportunity for AI to outperform. Topics include the anticipated trajectory of AI prediction, including the near term pitfalls and risks, and some surprising long term implications.