Jack Clark to deliver Per Jacobsson Foundation lecture on AI progress and recursive self-improvement

Jack Clark, co-founder of Anthropic, will give this year's Per Jacobsson Foundation annual lecture under the title "Inside the Event Horizon: Why Everyone Needs to Take AI Progress Seriously." The event will be live-streamed on the foundation's lecture page. The program opens with remarks from Foundation President Nigel Clarke, followed by a moderated discussion led by Foundation Chair Raghuram Rajan, former governor of the Reserve Bank of India. Two discussants will join the panel: V. Anantha Nageswaran, the 18th Chief Economic Advisor to the Government of India, and Yan Shen, Professor of Economics and Anton Chair Professor at Peking University's National School of Development.

According to the event description, the lecture will focus on the coming years of AI progress and the large-scale shifts it may bring to the economy and daily life. Clark is expected to engage directly with the economic questions that follow from the current pace of advancement: significant changes in the rate of scientific discovery, advances in robotics, the autonomy and economic impact of machine-driven businesses, and the potential for AI systems to contribute to their own production through a process of recursive self-improvement. That phrase is not new in research circles, but its appearance in an official lecture by a foundation concerned with international monetary affairs signals the topic has moved from speculation into policy debate.

The Per Jacobsson Foundation holds the annual lecture to stimulate discussion on the most pressing questions in international monetary and financial affairs. The choice of Clark — a technical figure from a leading AI lab rather than a classical economist — reflects a recognition that the boundary between monetary policy and technological capability is blurring. Rajan, a distinguished economist in his own right, will steer the conversation; the presence of Nageswaran and Shen adds perspectives from major emerging economies, suggesting the foundation expects asymmetric global spillovers.

Clark is not an academic publishing theoretical papers; he co-founded a company that develops industrial-scale models (Claude) and is directly involved in deployment and safety decisions. When someone in that position presents scenarios of recursive self-improvement and autonomous businesses to a forum of monetary policymakers, the message is not "this might happen" but "this is already in the work plan." The foundation has not published specific performance metrics or quantitative forecasts to be presented in the lecture, so the value will lie in the conceptual framing and the identification of regulatory decision points, not in numbers.

The event will be live-streamed on the foundation's lecture page. For anyone tracking AI policy, the economics of automation, or simply wanting to hear what a serious discussion of recursive self-improvement sounds like outside the lab, this is the place.