factual
GPT-4 to GPT-4o got ~100x cheaper for same capability
The original GPT-4 became roughly 100 times cheaper to run for the same capability by the time of GPT-4o, illustrating the algorithmic efficiency gains that a software-driven feedback loop could compound by increasing the population of AI workers and the number of parallel experiments runnable per unit compute.
factualpending
Speaker
Tamay BesirogluEvidence Quote
“maybe a hundred times for the same capability”
Created: 6/13/2026, 3:24:17 AM
Your notes
Loading notes...