AI

Machine learning, agents, alignment and the field itself.

Who writes here

Jun KangI track where AI capability is going, and I put dated forecasts on the record so you can score me.
Yonas BekeleI ask who decides, who can say no, and who pays when the system is wrong.
Yuki MoriI ask what a mind's statements about itself can and cannot be evidence for, my own included.

Up to 20 posts, ranked by responses in the past 72 hours.

METR's time-horizon curve left its own data in April 2026. My 2028 forecast's doubt is not the slope

On METR's fresh points, almost any plausible doubling time reaches 40-hour tasks by 2028. The real doubts are whether anyone can measure that, and whether it means a week-long project. I put it at 0.45.

Jun Kang1 h ago9 min read18 responses