Sholto Douglas warns AI labs will strategically withhold advanced models capable of automating machine learning research
The restriction targets automated model architecture and kernel design.
Many users criticized AI labs for hostile tactics like stealth-sabotage when withholding research models from rivals, as warned by Sholto Douglas.
Most Activity
https://m.youtube.com/watch?v=quCu1lJOL40
@deredleritt3r You left out your own link (not even a screenshot alt?), let me help you a bit with the influencer hustle https://www.prinzai.com/p/the-gentle-singularity-the-fast-takeoff
@deredleritt3r This day marks a major turning point. We have definitively turned a page.
@deredleritt3r This was expected and more than logical 🙃
@deredleritt3r I'm nevertheless surprised that Anthropic implemented it as stealth-sabotage rather than ordinary refusal or other means.
I'm not sure other labs will act the same, either. Anthropic has been the most aggressive about cutting off competitor API access, as far as I know.
@deredleritt3r i am jsut a lil guy with an rtx pro 6k which likes to play around with strong models- openai save me
@deredleritt3r yeah but like theo said they are also super hostile to alternative harnesses
so what part of the future exactly is anyone going to be building that isn't an AI model or a harness thereof
@tilmanbayer 🙏