Kevin A. Bryan

@Afinetheorem

Assoc. Prof. of Strategy, U Toronto Rotman | Chief Economist, CDL Toronto | Co-Founder, AllDayTA | Ars longa, vita brevis, occasio praeceps (especially now)

Toronto, ON, Canada

kevinbryanecon.com

Replies Reposts Media

Kevin A. Bryan
@Afinetheorem
2h
A lovely poem by the great, just-passed naturalist Wendell Berry (hear him read it in the link!). Along with John McPhee, Aldo Leopold, Annie Dillard, Barry Lopez, Robert Macfarland, Thoreau (esp. "Walking"), Wordsworth: nature, grandeur, solitude. All recommended.
1
Kevin A. Bryan
@Afinetheorem
2h
I agree. Obviously NFL theorem, Lucas critique, etc would all make you suspect this wouldn't work. But perhaps the purely predictable statistical aspects make up more of common time series than our intuition would expect?
Zhengyao Jiang
@zhengyaojiang
7h
I'm always a bit surprised this thread of work keeps making progress. How come completely different forms of time series can be modeled by a single model, and there's even positive transfer? Almost feels like a deeper mystery than LLMs
Kevin A. Bryan
@Afinetheorem
3h
For regulatory and safety purposes, the law should treat open source and closed identically. But that means you need to treat open source *in its zero guardrails form*, because immediately after release, every open weights model goes here (or worse, is RLed by hostile actors!)
Abliteration.ai
@abliteration_ai
14h
Today we're releasing abliterated-model-large-v2. Based on GLM-5.3, which is #3 on Terminal-Bench 4.0 (behind only Opus 5 and Fable), with 2× the cyber exploitation of 5.2. We abliterated and hosted it so it does the offensive cyber, red teaming, and agent testing work other
1
Kevin A. Bryan
@Afinetheorem
4h
Fuck this coward (cadets are high school volunteers). A wifebeating alcoholic who cheated on his first wife with his second, second with his third, and was so bad at his current job that his second-in-command went to the WH and resigned due to PH's incompetence. 1/2
Pete Hegseth
@PeteHegseth
6h
🇨🇦 (this is real)
5
Kevin A. Bryan
@Afinetheorem
5h
I would not be suprised in AI World for "lifelong learning" - short programs, going back for a year to study philosophy, etc. - to be more common, like they already are for b-schools. In which case, "as an alumni, you are a member of the lifelong club" is not a crazy model...
Marc Porter Magee 🎓
@marcportermagee
8h
Vanderbilt is trying to change the idea of a university into something like an exclusive club with outposts around the world
3