We’re Learning Backwards: LLMs build intelligence in reverse, and the Scaling Hypothesis is bounded
We’re Learning Backwards: LLMs build intelligence in reverse, and the Scaling Hypothesis is bounded