> how gigantic $100 billion models are all you need for AGI and that it’s just a matter of time until we reach this scale > Obviously massive models aren’t eno

> how gigantic $100 billion models are all you need for AGI and that it’s just a matter of time until we reach this scale > Obviously massive models aren’t enough to get to AGI, but all we have to do is scale up inference time compute and we’ll get there!” Corporate wants you to tell the difference between these two pictures... Obviously the latter was a step required to make former work. Always has been.