People are trying to use RL for reasoning when what they should do is use reasoning for RL.
@pmddomingos
-
The Rise of AI Agents as a 21st Century Phenomenon
By
–
20th century: Alcoholics Anonymous
21st century: Agentholics Anonymous -
Predictive Modeling vs Reinforcement Learning
By
–
If you can predict the future far enough ahead you don't need RL.
-
The Future of Autonomous AI Agents and Digital Addiction
By
–
If you think social media is addictive, wait until you have an army of agents doing your bidding 24×7.
-
The dual nature of AI research
By
–
The blessing of AI research is that you're never stuck. The curse is that the search space is so vast.
-
Why AI adoption is uniquely favorable in software development
By
–
Coding is a uniquely favorable domain for AI because it's verifiable, it has lots of data, AI researchers know it well, and coders are extremely receptive. I wouldn't jump to conclusions about other domains from it.
-
Speculating on the Future of the AI Boom
By
–
The question is not whether the AI boom will plateau, it’s when.
-
The Four Eras of AI Development
By
–
The four eras of AI:
1960s-1990s: Knowledge engineering
2000s-2020s: Data mining
2030s-2040s: Live learning
2050s-…: Complete robots -
Discussion on Anthropic’s product development strategy for AI tools
By
–
Anthropic has no strategy. Claude Code started as someone's side project, and so did Cowork and MCP.
-
The AI-driven Cambrian explosion in software development
By
–
Picture the Cambrian explosion in software when it no longer has to be written by software engineers.