Evidence receipt / prediction
Published · transcript-backedReiner Pope: prediction
26 Feb 2026 Cheeky Pint Reiner Pope of MatX on accelerating AI with transformer-optimized chips
“Parameter count should grow much faster than context length actually, just because of the underlying physics of what's available.”
Source trail
Everything needed to verify it.
- Speaker
- Reiner Pope
- Attribution
- Verified speaker
- Claim type
- prediction
- Recorded
- 26 Feb 2026
- Publisher
- Cheeky Pint
Transcript context
…Yes. Any other AI predictions? I'm generally just interested in what makes models cheaper and faster. Just at the model architecture level, really tied into this context thing. I think the context size will stay ballpark the same where it is, maybe a few times larger, but the parameter count will go up. Parameter count should grow much faster than context length actually, just because of the underlying physics of what's available. Though, has that been the story? Would that be a reacceleration of parameter count? Because it feels like we've leveled off slightly in the last year or two, and instead we've been focusing on more and better RL.…
Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.