factual

Pre-training reduces to training loss without clear capability signal

It is hard to know ex ante what a scaled model like GPT-5 would deliver, because scaling laws only promise lower training loss and next-token accuracy, not a clearly articulated emergent capability, so the signal to watch is a lab declaring the model's accuracy warrants continued log-linear scaling.

factualpending

Speaker

Modest

Evidence Quote

“all we know is okay, training loss would be lower.”

Source

Capital, Compute & AI Scaling— Invest Like the Best
Created: 11/18/2026, 6:42:24 AM

Your notes

Loading notes...