New Paradigms Won't Save You
The article discusses the debate surrounding the development of Artificial General Intelligence (AGI) and the necessity of new paradigms in AI. It argues that while some believe AGI will require a significant paradigm shift, current advancements in large language models (LLMs) suggest that AGI could emerge from existing frameworks. The author emphasizes that the timeline for new breakthroughs may be shorter than anticipated, with potential developments occurring within the next few years.
- ▪The evolution of LLMs includes key milestones such as neural networks, deep learning, and transformers.
- ▪Skeptics argue that LLMs may not lead to AGI, suggesting that future AGIs will combine insights from deep learning with other approaches.
- ▪Lindy's Law suggests that significant advancements in AI could occur within the next 3 to 5 years.
Opening excerpt (first ~120 words) tap to expand
New Paradigms Won't Save You...May 22, 20263522ShareOne popular objection to AI concerns is to declare that LLMs can never be AGI. You need a “new paradigm”. Therefore, AGI is so far in the future that it’s not worth worrying about.The obvious counterargument is to claim that no, LLMs can become AGI. But even without that counterargument, I think the “therefore” fails on its own terms. The key question is: how much of a new paradigm do we need?The landmark discoveries on the road to modern LLMs are something like:1950s: Neural networks1967: Multi-layer perceptron2010: Modern deep learning2017: Transformer, LLM2022: RLHF, chatbots2024: Chain of thought / test-time computeWe can think of this as an “evolutionary tree”, where a given LLM (let’s say Claude Opus 4.7) shares a recent “common…
Excerpt limited to ~120 words for fair-use compliance. The full article is at Astral Codex Ten.
