Jack Dorsey: AI Safety Rules May Entrench Existing Giants' Advantages

Odaily reports: Bitcoin News posted on X that Jack Dorsey supports open AI model releases, independent evaluation, and review of dangerous capabilities, believing that open-weight models allow independent researchers to reproduce results, identify problems, and develop safeguards. He opposes industry-wide restrictions on compute, training runs, and model development negotiated by current leading labs, arguing that this could make existing leading companies the only capable participants and exclude entities that could identify problems or build alternatives. Jack Dorsey stated that restricting general model releases should be a last resort, implemented only when there is independently reviewable evidence that release would significantly increase catastrophic risks that cannot be addressed through narrower measures; he hopes users can run and control AI tools on their own devices.