@johnschulman2: We love open weights and plan to keep releasing open-weight models and fine-tuning tools. But we’re not absolutists; mi…
Summary
John Schulman discusses a balanced approach to open-weight AI models, acknowledging misuse risks while committing to continued releases and calling for collaborative research on safe access.
View Cached Full Text
Cached at: 08/03/26, 01:46 PM
We love open weights and plan to keep releasing open-weight models and fine-tuning tools. But we’re not absolutists; misuse risks are real. Here’s how we’re thinking about a safe path forward, and the research needed to get there. Come work on it with us.
Thinking Machines (@thinkymachines): Releasing weights indiscriminately isn’t safe. Neither is keeping capable models inside a few labs.
We think there’s a path between them. We haven’t mapped all of it. Our new post covers the part we can see: how we assessed Inkling, and why access should widen in stages.
Similar Articles
@AnthropicAI: There’s been a lot of speculation about where we stand on open-weights models. We’ve outlined our views in full here:
Anthropic CEO Dario Amodei clarifies that the company has never advocated for a ban on open-weights models, stating they are a public good when not dangerously capable, while outlining his concerns about authoritarian use and misuse of powerful AI.
Tradeoffs in Open-Weights Models (9 minute read)
An analysis of the open-weights AI debate: proponents argue it democratizes AI, while critics cite misuse risks; major tech companies signed a letter supporting open weights, while Anthropic and parts of the Trump administration remain cautious.
@idavidrein: Excellent thread on open-weight models, lots of great points here
A tweet praising an external thread by @1a3orn arguing that open-weight models have been deeply and irreplaceably useful for AI safety, in response to Anthropic's recent stance against open weights.
anthropic’s position on open weights models
Anthropic outlines its position on open weights models, highlighting safety concerns and considerations for responsible release.
Open weight progression with no frontier release
A software developer discusses how OpenAI and Anthropic's strategy of keeping models internal may influence open weight releases and the replication of models by Chinese companies.