Savlar4 sources healthy · latest episode 15 Oct 2026
Speaker

Stefano Ermon

Guest on No Priors: Artificial Intelligence | Technology | Startups. 186 claims across 1 episode, Sept 2026.

186claims, each timestamped
0 of 12signals on the brief
0verified changes of mind
12open predictions

Positions over time

Every pair of statements that was compared. Only a verified change counts as a change of mind; rewording, added nuance and different questions do not.

No two statements by this speaker have been compared yet.

Predictions

As stated. A window appears only where the speaker gave one.

ExpectsBySaidStatus
Ermon expected generative modeling to be difficult enough to occupy his entire career. no date given18 Sept 2026pending
The speaker expects the ability to scale test-time compute more effectively to matter. no date given18 Sept 2026pending
The speaker predicts that the more parallel solution will eventually win. no date given18 Sept 2026pending
A diffusion-based language model might eventually be significantly more intelligent than an autoregressive model. no date given18 Sept 2026pending
Whether a diffusion-based language model will eventually be significantly more intelligent than an autoregressive model is possible but unknown. no date given18 Sept 2026pending
The speaker expects efficiency to become very important for AI systems. no date given18 Sept 2026pending
Efficiency will eventually be very important in AI systems. no date given18 Sept 2026pending
People will choose a faster solution when it provides the same quality as a slower one. no date given18 Sept 2026pending
Diffusion models' data-efficiency advantage, if it holds at scale, could be especially valuable for tasks where data is a bottleneck. no date given18 Sept 2026pending
The speaker thinks that a differentiated offering based on this technology will create value because the technology is important and valuable. no date given18 Sept 2026pending
At least some latency-sensitive workloads could be served by models that maximize quality within a specified latency budget. no date given18 Sept 2026pending
The speaker is uncertain how important human ingenuity will be to the team's work six months to a year later. 18 Mar 2027 – 18 Sept 202718 Sept 2026pending

Appearances

Claims per episode