There is a meaningful distinction between having scaling continue at large parameter counts (moving from 1 trillion to 10 trillion to 10 quadrillion parameters) versus achieving surprising emergent capabilities where the algorithm does things that weren't anticipated purely through scale.

definitionpending

Speaker

Unidentified Speaker — Breakthrough potential of AI | Sam Altman | MIT 2023 [T5cPoNwO7II]

Evidence Quote

“but one thing that's absolutely unique about this class of algorithm versus anything I've ever seen before uh is that it it surprises you with raw horsepower regardless of whether you measure it in parameter count or some other way uh it it does things that you didn't anticipate purely by putting more horsepower behind it”

Source

Breakthrough potential of AI | Sam Altman | MIT 2023— Imagination in Action
Created: 8/10/2026, 10:46:10 PM

Your notes

Loading notes...