Home | Notifications | New Note | Local | Federated | Search | Logout

Lynne@lynne@mk.pars.ee

Filmmaker, Physicist, Codec researcher.
Fastest FFTs, x86/aarch64 asm, Vulkan.
Getting out there and vanishing into the Gaussian blur of a Sunday afternoon.
SIMD queen of FFmpeg

Conceptually: Present
Informally: Verified
Joined: 2026-09-07 15:18:35
3 notes, 0 following, 0 followers

Reply to @lynne@mk.pars.ee
Lynne@lynne@mk.pars.ee (2026-10-07 19:58:04)
Programmers went through this exact same process mere months ago. I don't feel bad for them. I don't feel bad for mathematicians either. I don't feel bad for biologists, who I think may be next, with lights-out wet labs coming online.

Lynne@lynne@mk.pars.ee (2026-10-07 19:54:17)
I think that this is the starting point where mathematics becomes an API, with Lean formalization coming first, which then gets distilled by agents (guided by humans, currently), fit inside giant Lean libraries, with the points that made the proof work extracted and taken as proven by other agents to prove more problems, 20 goto 10.
Agents will keep extending proofs further, like Turing-complete Navier-Stokes was, on their own (with human guiding for now).
It takes a fraction of the mathematics community to learn how to prompt and treat mathematics like a startup to completely disrupt everyone else. I keep seeing mathematicians in their "bargaining" stages of grief, so I feel like its happening.

Lynne@lynne@mk.pars.ee (2026-09-07 15:03:26)

---Attachments---
image: https://mk.pars.ee/files/82391ba1-781d-4034-be89-7b7849d33467
image: https://mk.pars.ee/files/45e06bb3-7fbe-4bd8-bb6d-2ad045acfce6
image: https://mk.pars.ee/files/137e5526-56aa-4262-ac90-d5b42a69b23d
image: https://mk.pars.ee/files/298d1dad-3fea-460d-be47-ab852d5ddd88