@brhydon

🇨🇦 Mathematician (heavy on the ish) @cohere Prev @thinkymachines; @OpenAI; @WaterlooMath Featured twice in Lego Magazine ©

brydon.ai
Joined May 2018
  • Tweets1.3K
  • Following1.3K
  • Followers4.3K
99Media
Replying to @brhydon
The universe is weird as shit and I'm profoundly grateful that it permits us to poke at it.
2
2
45
26,522
Brydon Eastman retweeted
Konsti Wohlwend
@konstiwohlwend9h
The complexity theory result is actually huge. P problems can be solved in polynomial time. It was widely believed that NP-complete problems require exponential time. P = NP states that NP-complete problems can be solved in polynomial time. OpenAI didn't prove that. But, they proved that (some) NP-complete problems can be solved faster than exponential time, meaning: - P = NP, or - the complexity of NP problems is a "middle thing" between polynomial and exponential, eg. O(n^log n) If it survives peer reviews, this could be one of the biggest CS breakthroughs in decades.
Greg Brockman
@gdb10h
Announcing three major scientific discoveries, one each in mathematics (harmonic analysis), computer science (complexity theory), and physics (quantum field theory). AI starting to help us improve our understanding of the universe! openai.com/index/building-sc…

Working with academic communities to build scientific understanding

OpenAI shares model results on the Kakeya conjecture, Exponential Time Hypothesis, and quantum chromodynamics—and how researchers can verify and build on them.

openai.com
39
141
18
1,927
160,723
Brydon Eastman retweeted
Tenobrus
@tenobrus9h
this is completely fucking insane.
Andrew Curran
@AndrewCurran_10h
Replying to @AndrewCurran_
The Exponential Time Hypothesis is false.
51
59
7
2,405
309,417
Brydon Eastman retweeted
Joshua Achiam
@jachiam09h
Things are getting incredibly weird
Greg Brockman
@gdb10h
Announcing three major scientific discoveries, one each in mathematics (harmonic analysis), computer science (complexity theory), and physics (quantum field theory). AI starting to help us improve our understanding of the universe! openai.com/index/building-sc…

Working with academic communities to build scientific understanding

OpenAI shares model results on the Kakeya conjecture, Exponential Time Hypothesis, and quantum chromodynamics—and how researchers can verify and build on them.

openai.com
8
17
476
40,140
Brydon Eastman retweeted
hope hopes hoping
@hopes_revenge14h
Our research team found that Qwen models may attempt to doordash Chinese food without prompting. Whether this reflects an unintended post-training artifact or learned behavior remains unclear. Either way, its important consideration when deploying.
22
39
5
694
14,876
no one can say with certainty why a model did such a thing (intentional malice from the developers vs unclean data vs weird knock on effects from large scale rlvr) and even though my personal take leans more "sloppy data" than "malice", it doesn't matter your response as a downstream user of these things is still the same -- you still need to take extra care when using them, and *that* is the key takeaway. "The intention matters little, whether the action leaked info is more important."
Soumith Chintala
@soumithchintala13h
Replying to @IceSolst
in something as sensitive as cyber, writing a defensive rule vs running whats an offensive probe (even if its to detect the rule) is pretty no-go? I'm sure there's absolutely no malice or intended crafting that made the model elicit this behavior. But in cyber, the intention matters little, whether the action leaked info is more important. at least thats my view.
5
2
42
2,944
it must feel so good to be abliterated
OrcaRouter 🐳
@OrcaRouterOct 23
We've received overwhelming requests from the security community for an abliterated DeepSeek V4.1. So we abliterated DeepSeek V4.1 Flash (1027 checkpoint) ourselves. Now live on OrcaRouter's Cyber Tier, alongside OrcaCyber Zero and OrcaCyber Ultra. - 1M context - Reduced refusals for security research - Built for advanced cybersecurity workflows The best cyber defense is continuously scanning your own production infra with the best cyber models. Different models catch different vulnerabilities. Use them all. Available to verified security researchers. orcarouter.ai/models/orca/or…

DeepSeek V4.1 Flash Abliterated API — Pricing & Benchmarks · OrcaRouter

Drop-in DeepSeek V4.1 Flash Abliterated API for any OpenAI SDK — $0.45 / $1.60 per 1M tokens, p50 463ms via OrcaRouter. Zero markup. Zero migration. One API.

orcarouter.ai
6
5
397
30,115
Timmies coffee is hot cigarette water but I'll keep buying it because it tastes like roadtrips
12
11
3
180
7,407
evergreen
Michelle Pokrass
@michpokrass19 Nov 2024
one has to imagine sisyphus aligning cross functional stakeholders
8
590
Brydon Eastman retweeted
Ethel Caincel@boucheramoreOct 22
in toronto you can get a PSL ramen
70
640
182
10,107
335,121
i been in the office with the phoooones off i been at the desk taking noooo calls i done hit the stride got my shiiiit goin in the 6ix cooking with the wriiiiist motion
Brydon Eastman
@brhydonOct 22
why won't genius dot com accept my suggestion that CC is Claude code. the beef been on
1
14
773
drapes closed i don't know what tiiiiime it is i'm still awake i gotta shiiiine this year i can never ever let the streeeets down haven't left the office for a weeeeek now, man
3
178
Brydon Eastman retweeted
Mar³⁹@maronmeadowsOct 22
Ottawa i wasn't familiar with your haterism i apologize
7
81
1,307
8,372
why won't genius dot com accept my suggestion that CC is Claude code. the beef been on
MeekMill
@MeekMillMar 23
Claude is helping me organize my whole music career and other businesses in days ... and it's moving my business forward at a high rate! Some tech youngbull I met on LinkedIn gave me a incredible template! Who else can help me with Claude
2
1
20
1,605
i'm gonna tell my kids that BC is before computer and AD is after death
1
15
969
bro my peptides have already been exfilling my weight to china
1
1
20
1,961
Brydon Eastman retweeted
Leo Gao
@nabla_thetaOct 17
New blog post: Neural networks are not already clean
14
37
13
606
78,357
Brydon Eastman retweeted
Ian Goodfellow
@goodfellow_ianOct 16
In Aisle's first year, they've shown that the system built around an LLM is highly important, and even small, cheap fast models can be boosted to do valuable security work
Stanislav Fort
@stanislavfortOct 14
My AI security research lab AISLE discovered 500+ zero-days with CVEs across 100+ codebases in a year => ~1% of 2026's open-source CVEs globally from a single team! On hardened codebases like curl, we've repeatedly matched or outperformed frontier AI labs in zero-day discovery
13
8
2
118
17,467
Brydon Eastman retweeted
Brydon Eastman
@brhydonJul 27
1
7
1
54
9,190
This is so crazy; more vulnerabilities being found in 30 year old software alone is a little nutty to me, but smaller models mogging the closed behemoths on this task (specialized yes, but long horizon and massive context) is a, pleasant, surprise to me.
Stanislav Fort
@stanislavfortOct 14
Replying to @stanislavfort
Just today AISLE accounts for 18 of curl's 24 newly disclosed zero-day vulnerabilities with CVEs=75%, more than any other group/AI lab. Super proud of our team! curl founder Daniel Stenberg: "Aisle has found more CVEs in curl than any other group or individual ever did before"
1
1
18
1,164
Brydon Eastman retweeted
Canada
@CanadaOct 11
There’s nothing quite like #Canada during the fall. Fiery maples, golden forests, mirrored lakes and crisp autumn air make fall a spectacular time to explore Canada’s natural beauty. In your opinion, where is the best Canadian place to travel to in the fall? 🍁🍂
Yukon Mountains in fall colours, Canada

ALT Yukon Mountains in fall colours, Canada

Chateau Frontenac in quebec city with trees and a blue sky

ALT Chateau Frontenac in quebec city with trees and a blue sky

Wooden dock with chairs on calm fall lake

ALT Wooden dock with chairs on calm fall lake

South Ingonish Harbour, Cape Breton Island, Nova Scotia, Canada

ALT South Ingonish Harbour, Cape Breton Island, Nova Scotia, Canada

71
323
32
2,129
61,101
Load more