Eight Things to Know about Large Language Models

Abstract The widespread public deployment of large language models (LLMs) in recent months has prompted a wave of new attention and engagement from advocates, policymakers, and scholars from many fields. This attention is a timely response to the many urgent questions that this technology raises, but it can sometimes miss important considerations. This article surveys the evidence for eight potentially surprising such points: (1) LLMs predictably get more effective with increasing investment, even without targeted innovation; (2) many important LLM behaviors arise unpredictably as a byproduct of increasing investment; (3) LLMs often appear to learn and use representations of the outside world; (4) experts are not yet able to interpret the inner workings of LLMs; (5) there are no reliable techniques for steering the behavior of LLMs; (6) human performance on a task isn't an upper bound on LLM performance; (7) LLMs need not express the values of their creators nor the values encoded in web text; (8) brief interactions with LLMs are often misleading.