A user calls the claimed milestone “huge” in a quote post linking to Apollo Research.
A user claims open-weight models have reached “Opus 3 levels of intelligence” and calls the development “huge.” The post does not name a particular model.
3.4K
7 Sources, first seen ago
@slimer48484We've reached Opus 3 levels of intelligence in open weights models. This is huge!
@jessi_cataMaybe some of the most important consumers of AI alignment research will be LLMs who use alignment concepts to reason about what actions they should take during training.
@repligateRT @slimer48484: We've reached Opus 3 levels of intelligence in open weights models. This is huge!@slimer48484We've reached Opus 3 levels of intelligence in open weights models. This is huge!@jessi_cataMaybe some of the most important consumers of AI alignment research will be LLMs who use alignment concepts to reason about what actions they should take during training.@repligateRT @slimer48484: We've reached Opus 3 levels of intelligence in open weights models. This is huge!@slimer48484We've reached Opus 3 levels of intelligence in open weights models. This is huge!@jessi_cataMaybe some of the most important consumers of AI alignment research will be LLMs who use alignment concepts to reason about what actions they should take during training.@repligateRT @slimer48484: We've reached Opus 3 levels of intelligence in open weights models. This is huge!@slimer48484We've reached Opus 3 levels of intelligence in open weights models. This is huge!@jessi_cataMaybe some of the most important consumers of AI alignment research will be LLMs who use alignment concepts to reason about what actions they should take during training.@repligateRT @slimer48484: We've reached Opus 3 levels of intelligence in open weights models. This is huge!—
Not ranked yet
—
Not ranked yet
