What's the next big "scaling era" for LLMs
Summary
The author reflects on past LLM scaling eras (model size, chain-of-thought, agents) and speculates about the next big scaling dimension, suggesting recursive self-improvement as a potential breakthrough.
Similar Articles
At what point do LLMs start bootstrapping themselves and generating the next LLM
The article speculates on when LLMs might begin bootstrapping themselves, potentially leading to an AI singularity with exponential progress. It invites discussion and resources on this topic.
The Scaling Laws of Skills in LLM Agent Systems
This paper identifies two coupled scaling laws for skill libraries in LLM agent systems: routing accuracy decays logarithmically with library size, and execution dynamics show a rescue effect. The laws are validated across 15 models and over a million decisions, and law-guided optimization significantly improves performance.
Can RL Teach Long-Horizon Reasoning to LLMs? Expressiveness Is Key
This paper introduces ScaleLogic, a framework demonstrating that RL training compute scales as a power law with reasoning depth in LLMs. It highlights that logical expressiveness is key to improving downstream transfer and training efficiency.
2x, not 10x: coding with LLMs in 2026
The author argues that LLMs currently provide about a 2x productivity boost for coding due to their ability to handle easily verifiable tasks, but fundamental limitations prevent a 10x improvement; further gains will come from retooling around existing capabilities rather than model improvements.
Why scaling LLMs won't lead to real agency: A conceptual architecture based on 3-tier Embodied AI, physical cost efference copy, and offline sleep cycles.
The article proposes a conceptual 3-tier embodied AI architecture to address the limitations of scaling LLMs for agency, emphasizing physical cost auditing and offline sleep cycles.