General image and text synthesisers: commodity
Services, niche, personal, enterprise image and text synthesisers: moat
A tale as old as time.
Conversation
Replying to
Open source stable diffusion, miniDALLE
Seems like it’ll be a race to the bottom on price and race to the top for quality. New models coming out every 2 years, old models rapidly outdated.
2
2
Replying to
Problem is biggest models that use the biggest piles of hardware keep winning. The small ones are qualitatively always a generation behind. I suspect it will be public training of large models plus hybrid inference on augmented versions of large models. Already so in vision.
2
2
Race to the bottom on price would be nice but the cost curves are definitely not there yet. Nvidia stack is the bottleneck.
2
2
Search was orders of magnitude cheaper from get go. Too cheap to meter even before Google. ML is not. Training is very expensive. Even inference at the bleeding edge still seems to be in the dollars/inference range, not sub-penny. Unit economics closer to video streaming ~1997
Replying to
Yeah, but isn’t video streaming a commodity functionality today? Wouldn’t it make sense that it follows the same trajectory?
1
Replying to
Yes but big markets that enabled it (YouTube) we’re not private moats but public demand/supply aggregation. Many early enterprise video things failed.
1
1
Show replies

