Guest
Andreas Blattmann
Black Forest Labs
Follow Andreas Blattmann
We’ll email you each new moment, one email a day at most. Plus the best of every big interview, Wednesday and Saturday.
Check your inbox
We sent a link to your address. Click it to finish signing up. Not there in a minute? Look in spam.
or
It’s free.
Real Alpha has cut 4 moments of Andreas Blattmann (Black Forest Labs) talking, from an interview on Latent Space published Nov 16, 2026.
What Andreas Blattmann said
Why FLUX 3 Trains on Images, Video and Sound — Andreas Blattmann & Dustin Podell, Black Forest Labs
4 moments · 21 min
- 34:53
A video model learned enough 3D to fine-tune in 1,000 steps
“I think, 1K fine-tuning steps on 3D data, and we got a 3D model out of it.”
at 36:51 - 40:12
Training should focus compute on informative denoising steps Pro pick
“where should we spend our compute? Which parts of this trajectory from data to noise carry the information that's most important?”
at 44:33 - 1:24:25
BFL aims to crack one-shot robot learning by 2027
“You should be showing a robot an action one time, and then it should be able to do this”
at 1:25:29 - 1:13:06
Autoregression is more data-efficient; diffusion can sample faster
“I would say autoregressive models are way more data efficient.”
at 1:14:14
- 34:53