Comment on Quantitative AI progress needs accurate and transparent evaluationparentComments−falcor841yArguably we're not that far from jumping up one level, having an AI agent that when encountering any new type of hard problem would train a new dedicated sub-AI to solve that.−moffkalast1yDeep learning scientists automating themselves? Ironic.
Comments
Arguably we're not that far from jumping up one level, having an AI agent that when encountering any new type of hard problem would train a new dedicated sub-AI to solve that.
Deep learning scientists automating themselves? Ironic.