You can’t imitation-learn how to continual-learn — LessWrong
In this post, I’m trying to put forward a narrow, pedagogical point, one that comes up mainly when I’m arguing in favor of LLMs having limitations that human learning does not. …
x You can’t imitation-learn how to continual-learn — LessWrong AI Curated 2026 Top Fifty: 9 % 206 You can’t imitation-learn how to continual-learn by Steven Byrnes 16th Mar 2026 7 min read 54 206 In this post, I’m trying to put forward a narrow, pedagogical point, one that comes up mainly when I’m arguing in favor of LLMs having limitations that human learning does not. (E.g. here , here , here .) See the bottom of the post for a list of subtexts that you should NOT read into this post, including “…therefore LLMs are dumb”, or “…therefore LLMs can’t possibly scale to superintelligence”. Some i
saved by
related reading
- The nature of LLM algorithmic progress (v2) — LessWronglesswrong.com
- Simulators — LessWronglesswrong.com
- How can LLM RL Work Despite Information-Theoretic Inefficiencyberen.io
- What are the real problems of continual learning?infinitefaculty.substack.com
- AI researchers talk continual learning over a game of HORSEyoutube.com
- In (highly contingent!) defense of interpretability-in-the-loop ML training — AI Alignment Forumalignmentforum.org
- Taking LLMs Seriously (As Language Models) — LessWronglesswrong.com
- Lil'Loglilianweng.github.io
- Why We Need Continual Learning | Andreessen Horowitza16z.com
- The Continual Learning Problemjessylin.com
- Trying to actually define continual learningcharlesponeill.com
- What's so hard about continuous learning?seangoedecke.com