Get your free personalized podcast brief
We scan new podcasts and send you the top 5 insights daily.
Beyond just coding, improving AI models requires subtle skills like designing effective reinforcement learning environments or managing human expert feedback. Newman questions how close we are to recursive self-improvement by asking if AIs can automate these tasks, which rely on nuanced "taste and judgment" rather than just raw computational ability.
Related Insights
Developing a high-quality AI skill, like an "Ad Optimizer," is not as simple as writing a single prompt. It requires a laborious, iterative cycle of instructing, testing, analyzing poor outputs, and refining the instructions—much like training a human employee. This effort will become a key differentiator.
Unlike traditional software where problems are solved by debugging code, improving AI systems is an organic process. Getting from an 80% effective prototype to a 99% production-ready system requires a new development loop focused on collecting user feedback and signals to retrain the model.
The concept that AIs can build better AIs, creating an accelerating feedback loop, is no longer theoretical. Leaders from Anthropic, OpenAI, and Google DeepMind have publicly confirmed they are actively using current AI models to develop the next generation, making RSI a practical engineering pursuit.
Unlike any prior tool, AI can be directly applied to improve its own creation. It designs more efficient computer chips, writes better training code, and automates research, creating a recursive self-improvement loop that rapidly outpaces human oversight and control.
Companies like OpenAI and Anthropic are not just building better models; their strategic goal is an "automated AI researcher." The ability for an AI to accelerate its own development is viewed as the key to getting so far ahead that no competitor can catch up.
The viral claim of "recursive self-improvement" is overstated. However, AI is drastically changing the work of AI engineers, shifting their role from coding to supervising AI agents. This automation of engineering is a critical precursor to true self-improvement.
The next evolution for AI agents is recursive learning: programming them to run tasks on a schedule to update their own knowledge. For example, an agent could study the latest YouTube thumbnail trends daily to improve its own thumbnail generation skill.
The path to AI self-improvement isn't uniform. It is happening first in software engineering and AI research because these fields have cheap, fast, and verifiable feedback (e.g., unit tests). This capability won't automatically transfer to domains like biology until similar closed-loop systems are built.
A major frontier for AI in science is developing 'taste'—the human ability to discern not just if a research question is solvable, but if it is genuinely interesting and impactful. Models currently struggle to differentiate an exciting result from a boring one.
Build a feedback loop where an AI system captures performance data for the content it creates. It then analyzes what worked and automatically updates its own skills and models to improve future output, creating a system that learns.