

They are not adjusting their models at any cost to other benchmarks. Especially when the benchmark isn’t even public. This would be very easily caught.
no, they don’t have agi, but this benchmark isn’t proof of that either way, an ai that’s dumber than a human could still be agi




The original definition of state is different from the western nuspeak one that means government
the point being that there will be government, just horizontally managed.