Zulip Chat Archive

Stream: general

Topic: Jon Sterling, Is it time for a new proof assistant?


Srayan Jana (Sep 28 2025 at 00:46):

I saw this video pop up on my youtube recommended, and I figured it was pretty relevant to Lean (considering the presenter mentioned Lean multiple times)
https://youtu.be/7oBkEbKJvnE?si=CJhNG-6JmyawhoXf
https://hottest-seminar.github.io/hottestfiles/Sterling-2025-09-25-HoTTEST.pdf

Srayan Jana (Sep 28 2025 at 01:05):

I think the presenter is @Jonathan Sterling and seems to be part of this conference series: https://hottest-seminar.github.io/

Chris Henson (Sep 28 2025 at 01:28):

An additional piece of relevance is that the current prototype is implemented in Lean.

Wrenna Robson (Sep 29 2025 at 17:20):

@Jonathan Sterling I just wanted to say that I really enjoyed reading this!

(deleted) (Sep 29 2025 at 17:49):

I know that HoTT people really believe in their theory to the point where they invite teenagers who take part in olympiads to attend lectures on HoTT. But when will it be widely adopted by the formal verification community or will it remain something obscure

(deleted) (Sep 29 2025 at 17:51):

Yeah, I also enjoy reading the slides

(deleted) (Sep 29 2025 at 17:52):

I do feel that we do need a new proof assistant based on some new foundation. And if the HoTT community manages to make a good enough proof assistant this could convert many people to HoTT. I'm definitely not a fan of having to use native_decide just because decide is too weak

(deleted) (Sep 29 2025 at 17:53):

And if HoTT solves this problem then it could even offer a better user experience than Lean


Last updated: Dec 20 2025 at 21:32 UTC