A new post argues beliefs and desires come from optimal control and reinforcement learning.
It suggests they are properties of optimal policies.
This challenges the folk psychology view of agents.
What happened
A post argues that the belief-desire view of agents can be derived from classic theorems in optimal control and reinforcement learning.
Why it matters
This suggests beliefs and desires are properties of optimal policies, not just folk psychology assumptions.
What to watch
Whether this derivation changes how agents are modeled in AI and related fields.
Ask the AI about this article →
The post's argument reinterprets a foundational concept in psychology and economics—that agents act on beliefs and desires—through the lens of optimal control and reinforcement learning. By deriving this view from classic theorems, it positions beliefs and desires not as assumptions from folk psychology but as emergent properties of optimal policies. This could have implications for how AI systems are designed and understood, though the body does not detail specific applications. The introduction also emphasizes the importance of the agent-environment boundary and the dual role of reasons, but the post does not yet provide the full derivation or its consequences.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
Analyst Ming-Chi Kuo says Nvidia has revived the Rubin CPX AI accelerator with a substantially redesigned arch…
A UK study by UK AI Security Institute and Limbic AI surveyed 6,474 British adults
Broadcom's Clayton Donley says companies are doing mission-critical work with AI agents quickly, but without t…
Bank of England governor Andrew Bailey warned that advanced AI poses risks to financial infrastructure in a le…
OpenAI released a new evaluation framework on July 17, 2026, urging companies to measure AI ROI by 'useful out…
As AI agents perform real business tasks, 'Agentic Identity' (giving each AI a unique employee-like ID) and 'D…
