🔮 A world of many minds & the future of work

Good morning from London. This week has been eventful. Jensen Huang shared insights with me on Wednesday, and I also participated in a podcast discussing the surge in AI investments. Conversations surrounding the AI economy, in general, are indeed consequential for all of us, which is why it was essential to have this opportunity.

Recently, Anthropic, an AI company, experienced a significant leak of information from its private investor briefing. Although the numbers are extraordinary regardless of the timing, their disclosure caused considerable concern in the market. As we report, Korea's KOSPI index has plummeted by 7.57% today and 16.2% this month, with tech stocks aligned with the AI sector opening sharply lower.

Gardens of Intelligence' essay on Swarm Intelligence by Kevin provides a thought-provoking perspective on the future of AI. It imagines a world where an AI superorganism exists, composed of thousands of AI species, ranging from very narrow and simple to highly specialized and advanced. This could lead to a diverse landscape of AI capabilities, each excelling in specific tasks.

In a related article, Yue Xu introduces a framework to evaluate which work can be economically delegated, considering factors like output verification costs and mistake-related losses. This analysis sheds light on the potential evolution of work within AI-centric firms and how they may adapt to become truly AI-native.

As society grapples with the rapid advancement of AI, we must consider different paths for its integration. Monocognitism, the belief that there will be only one all-powerful AI, is unlikely due to the diversity of emerging AI systems and physical constraints like the speed of light. Instead, a polycognitive approach, featuring multiple AI systems with varied capabilities, seems more plausible.

This diversity, while presenting control challenges, could also empower AI to tackle different tasks more effectively, leading to a more intelligent and efficient society.