Journalist Shakeel Hashim prompted the debate over evaluation enforcement
@ShakeelHashim model evaluations are not the right conceptual unit of account for ai regulation. every single model made from now and onward will fail dangerous capability evals
@deanwball Mostly agreed. But what do you think should happen if a model fails the evaluation process?
Journalist Shakeel Hashim prompted the debate over evaluation enforcement
@ShakeelHashim model evaluations are not the right conceptual unit of account for ai regulation. every single model made from now and onward will fail dangerous capability evals
@deanwball Mostly agreed. But what do you think should happen if a model fails the evaluation process?
Users in the replies dismissed model evaluations as the wrong conceptual unit for AI regulation because every model from now on will fail.