Mike Vegeto

Almost every topic is at least two orders of magnitude easier to understand than it seems from the outside.

JEPA is so cool man

this is so cool, congrats @midjourney team!!

@Chrisgormley @boxcardavid You're the expert then.

@ZECTBynmo @boxcardavid Now add tariffs, shipping, to the US and to the end customer, and maybe some overhead. Share your math. (I lived in Shenzhen for over a decade as a factory owner and have purchased enough aluminum profiles to fill an Olympic swimming pool).

Chinese extrusion prices have hit Vevor

@Chrisgormley @boxcardavid I mean those prices are more than 2x the price on AliExpress. I'm sure actual Schenzhen street prices would be significantly less than that Also Vevor is definitely paying tariffs

@ZECTBynmo @boxcardavid True. That doesn’t mean they are paying what is correct though.

@Rich_Pluta @Abhindas1 @ModalMotors Because they haven't solved the (very real) challenges to manufacturing these magnets. Lab results are not enough to prove commercial viability

@Abhindas1 @ModalMotors How is this not a much bigger deal?

Just ran into Modal Motors. Non-rare earth electric motors for drones and robotics. Similar costs to Chinese motors, all U.S.-sourced. Incredible.

@Chrisgormley @boxcardavid Unless something changed there's no price floor for tariffs now.

@boxcardavid What would you guess those 10 pieces weigh? That’s Shenzhen street prices, and tariffs should be not small on aluminum imports.

@Sentdex How is the writing style? How is the coding style? I find models usually have a strength towards one end or the toher

Zai was gracious enough to give me a key to test out GLM 5.2. I used it on a few simple tasks and quickly realized this model is on another level. I committed to using GLM 5.2 solely for the weekend and yesterday on everything from simple data analysis, random queries, side projects, and real work, and I can honestly say this is the first open model that I could comfortably replace Opus 4.8/GPT 5.5 with. It’s THAT good. When I say everything, I mean everything. I never needed to fallback to GPT 5.5 or Opus 4.8. This really blew my mind. I was unable to find any task where I knew GPT 5.5 or Opus 4.8 could solve, but GLM 5.2 could not, and I actually found a few cases where GLM 5.2 was better. I am not trying to overhype anything here. It's just my actual experience with this model. It was of course only 3ish days of usage, maybe cracks would form in time, but the perf is staggering imo. I see it's an "inferior" model on the benchmarks even Zai has shared, but I am not so sure and I think this is the first time I've experienced that with an open model. I am not saying it's necessarily better, but I believe it's a replacement that you could run on-prem, which is crazy to me. It was to the point where I was double triple quadruple checking that I wasn't accidentally running Opus or GPT. I ran thru both Hermes and my own custom coding agent harness with extremely great success. I cannot believe this is only a 754B model that's also an open MIT licensed model. Do not sleep on this one, and definitely try it out. Get it locally if you can! Can I find a way to run it locally? That’s a different question, but I will be trying to get it done because this model is epic.

@reach_vb noticing this for the first time after updating this morning x.com/ZECTBynmo/stat…

Some Codex users are seeing a “Selected Model is at Capacity” error. We’re working to resolve it as quickly as possible. Updates: status.openai.com/incidents/01KV…

hmmmm latest version of codex is causing ripgrep to get flagged by osx. Odd!

@jxnlco @jxnlco is up there

who are some of the highest profile codex users you know?

@berkie1 We added a lot of land mass to Boston in that time too though - maybe roughly 20%? Although Logan is probably half of that

@ZECTBynmo Combination of shrinking household sizes and restrictive zoning.

In the 1980s, Massachusetts was headed for a water crisis. Through leak repairs, pipe replacement, efficient fixtures & public outreach, the MWRA cut water use dramatically. Today, Quabbin withdrawals average ~200mgd, the same as 1950. Boston uses less water than it did in 1900.

@berkie1 my biggest takeaway from that: how do we only have 200k more residents?

A remarkable conservation success: Boston now uses less water than it did in 1900, despite having over 200,000 more residents. Water consumption today is also less than half of its 1980s peak.

@berkie1 So many kilts

le chaton fat was what people called me in high school

@tekbog serious question: have you looked at JEPA for ops work? Obviously more of a research task, but feels like a great use case - plan action, predict next state, etc.

someone release something either le chaton fat or gpt 5.6 come on wiggle it a little

@JakeFleshner continual learning

Pitch me your company in 2 words Angel invested in 40+ companies and always looking for more

@emm0sh Project Hail Mary*

@emm0sh I just finished the Hail Mary book and he does so many calculations by hand - trajectory planning, orbital mechanics, etc. Obviously it's fiction, but it made me think how useless I'd be

i genuinely believe engineering software has made us worse engineers. here’s how:

Excited for the grok team. Once all the good models are banned they'll finally hit frontier status

@mattfreed I have two already

unironically having to order one of these