Read the associated paper "Safetywashing: Do AI Safety Benchmarks Actually Measure Safety Progress?": https://arxiv.org/abs/2407.21792 Focus on safety problems that aren’t solved with scale. Benchmarks are crucial in ML to operationalize the properties we want models to have (knowledge, reasoning, ethics, calibration, truthfulness, etc.). They act as a criterion to judge...
This is a hasty speculative fiction vignette of one way I expect we might get AGI by January 2025 (within about one year of writing this). Like similar works by others, I expect most of the guesses herein to turn out incorrect. However, this was still useful for expanding my...
This seems overshadowed by the GPT-4 news, but apparently Google is also releasing PaLM and an API along with "MakerSuite, a tool that lets developers start prototyping quickly and easily." Not many details yet (perhaps they rushed this out to compete with the GPT-4 announcement?), but some features of the...
Summary: A level-based guide for independently up-skilling in AI Safety Research Engineering that aims to give concrete objectives, goals, and resources to help anyone go from zero to hero. Cross-posted to the EA Forum. View a pretty Google Docs version here. Introduction I think great career guides are really useful...
Transformative AI will revolutionize the ways we interact with the world. But maybe narrow tool AIs are already starting to do so: Every week, there are new startups, websites, and tools launched around using modern machine learning techniques to solve everyday problems, and we might as well use them to...
Note: This is a joint distillation of both Iterated Distillation and Amplification by Ajeya Cotra (summarizing Paul Christiano) and A Generalist Agent by DeepMind. Audience: New to alignment. Mostly non-technical, but a basic ML understanding helps. Amplify, Distill, Rinse, Repeat Suppose you want to use a friend to help you...