beth

[Mr. Darcy @darcyagi]

openai has been quiet for way too long.

now gpt-6.2 astra/sol is showing up across user sessions, and the outputs look like a serious leap over opus 5.5.

if this holds up, anthropic is in for a fight.

[beth @bethh404]

at this point i’m convinced this guy just says shit to farm engagement.

it’s been over a month since opus 5.5 dropped and openai still hasn’t shipped a proper response, yet he just keeps hyping shit.

what an exhausting account.

[Tibo @thsottiaux]

@JamesMalsawm Pretty wild to me that GPT 6.1 Terra gets compared to Opus 5.5. No doubt that it's a great model, but it's not meant to compete in the big leagues. We've got other models for that.

[beth @bethh404]

i cannot stand the mentality that code somehow becomes less valuable because an ai agent wrote it.

if it’s correct, tested, maintainable, and does the job, what exactly is the problem?

we should be celebrating engineers who figure out how to ship faster.

[Mr. Darcy @darcyagi]

if your engineers can ship in hours but your internal processes make them wait weeks, your bottleneck isn’t engineering anymore.

it’s the company itself.

[beth @bethh404]

the scientific breakthroughs coming out of these labs over the next few years are going to be absolutely insane.

i can feel it.

[Pushmeet Kohli @pushmeet]

Across our projects @GoogleDeepMind, we are witnessing a genuine phase transition in the pace of scientific discovery. You will see a glimpse of this in our releases in the next few weeks.

[Mr. Darcy @darcyagi]

gpt-6.2 astra and sol are being routed to more and more users, and the outputs look seriously better than opus 5.5.

openai might be about to reclaim the throne.

next week is looking insane.

[Mr. Darcy @darcyagi]

holy shit, people are apparently getting routed to gpt-6.2 sol already, and the outputs look absolutely insane.

we got nothing last week, but next week is starting to look very interesting.

[Ryan Orhan @rynorhn]

the idea that ai-generated code is automatically slop is ridiculous.

judge the code, test the code, measure whether it works.

who gives a shit whether a human or an agent wrote it?

[Ryan Orhan @rynorhn]

this guy is a certified yapper.

it’s been over a month since opus 5.5 launched and he’s still doing this shit.

at this point, it just looks like engagement farming.

[Ryan Orhan @rynorhn]

i think scientific research is becoming the most exciting part of the entire ai race.

we’re reaching a point where the biggest achievements might not be better models, but what those models actually discover.

[Ryan Orhan @rynorhn]

openai, anthropic, and google deepmind have been unusually quiet about their most advanced models, while the research coming out of these labs keeps getting more extraordinary.

whatever they’re running internally might be far beyond what we’re using today.

[Mr. Darcy @darcyagi]

another day, another reminder that plus users apparently don’t exist anymore.

amazing.

[Mr. Darcy @darcyagi]

openai really needs to remember that people paying $20 a month are still paying customers.

plus shouldn’t feel like an afterthought every time something interesting happens.

[Mr. Darcy @darcyagi]

when someone leading scientific research at google deepmind starts talking about a genuine phase transition in the pace of discovery, you pay attention.

and we’re about to see some of the results in the next few weeks. i’m ridiculously excited for this.

[Mr. Darcy @darcyagi]

we get it, tibo. terra isn’t meant to compete with opus 5.5.

now where the fuck is the model that is???it’s been over 6 weeks.

[Mr. Darcy @darcyagi]

if your engineers can ship in hours but your internal processes make them wait weeks, your bottleneck isn’t engineering anymore.

it’s the company itself.

[Mr. Darcy @darcyagi]

anthropic’s opus 5.5 nerf makes that painfully obvious.

if you can’t run it, preserve it, or decide when to stop using it, you don’t really control it.

[Mr. Darcy @darcyagi]

developers need to start treating model availability as a serious infrastructure risk.

if your entire stack depends on a provider keeping one specific model alive, you’re building on someone else’s decisions.

open weights and self-hosting aren’t just ideological preferences.

[beth @bethh404]

i swear openai and anthropic have turned into professional shitposters.

anything except releasing the fucking models we’ve been waiting for

[Anthropic @AnthropicAI]

New on our Science Blog: How AI is beginning to change scientific modeling.

Scientific models help researchers explain and predict real-world systems—like how a single nerve cell fires, or what happens when neutron stars collide. Recent advances have transformed how scientists fit complex models to data. But scientists still typically write the models manually, and usually need to weigh up several competing versions, each of which might explain the data in a different way.

In a guest post, Jakob Macke, Daniel Gedon, and Stefan Wahl describe how AI is starting to help with these parts of the process. By writing, running, fitting, and revising scientific models as code, AI can explore many possible model structures, and find out which explanations the data support.

https://t.co/qsFJG76FDw

[Mr. Darcy @darcyagi]

openai and anthropic are locked in an intense competition right now.

unfortunately, it’s not on intelligence benchmarks.

it’s on who can publish more useless updates before shipping their next frontier model.

[Mr. Darcy @darcyagi]

i don’t think i’ve ever been this impatient for two competing model releases.

at this point i’m more interested in who breaks the silence first than who wins the benchmarks.