
Shippty ship ship, McShipface.

GTM @OpenAI. A fan of NY sports, technology, memes & nice people. My opinions are my own.

Shippty ship ship, McShipface.


Everyone is discussing whether GPT-6.2 beats Claude. Wrong question IMO. On Artificial Analysis, Claude scores 65 and costs $26.54 a task. OpenAI's new model scores 62 for $4. $4 DOLLARS!!!!!!!!!! That means OpenAI is going straight for enterprise budgets. CFOs don't buy the smartest model, they are business people and the one that's good enough at a sixth of the price will win. That's Anthropic's whole business during it's IPO. I'd be scared if I were them.

I tested GPT-6.2-Astra and GPT-6.2-Sol against three prior models at GeoGuessr Each model used computer use to explore the same 100 starting locations and had five minutes to place a pin within 1 km Results: - GPT-6.2-Astra: 77/100 (tied for best) - GPT-6.2-Sol: 77/100 (tied for best, fastest) - GPT-6.1-Terra: 70/100 - GPT-6-Astra: 65/100 - GPT-6.1-Sol: 59/100 GPT-6.2-Astra found 12 more locations than GPT-6-Astra
Some of you have been asking very politely. Some of you have just been asking. GPT-6.2 Astra and Sol are now available. Super efficient and built to perform where it counts—the real world. openai.com/index/introducing…

never seen opinions so split on a model. so far it looks like the testers who outsource their thinking hate it for being impenetrable while serious engineers and systems people enjoy solving hard problems with it.
Hi friends, in the latest weeks I worked with GPT 6.2 Astra. With it I built Redis features and most of YouRec (testflight.apple.com/join/N4…). My feeling is that it is the *best* model for system programming in C and for complex data structures. (continue)

GPT-6.2 Astra is an INSANE leap over GPT-6 Astra. It's scary how much it has improved. OpenAI is officially back! by the way - $10/$50 vs $8/$40 is this upgrade worth it?

Hi friends, in the latest weeks I worked with GPT 6.2 Astra. With it I built Redis features and most of YouRec (testflight.apple.com/join/N4…). My feeling is that it is the *best* model for system programming in C and for complex data structures. (continue)

This is, or should be, the future of AI. Whether you love or hate 6.2, it's teaching you a lesson. That lesson is: after some point, there isn't supposed to be just one winner. There isn't supposed to be a monolith. We need more in this world than one model being the only story. On SimReadyBench, 6.2 Astra is absolutely *BLOWING* everything away. It has Sol-level cost efficiency at 50% stronger performance. This is the most insane Pareto leap I've seen in any physical AI task. @OpenAI have absolutely done it. But here's the thing. This is an appreciation that you need to have had some exposure to post training to understand. You squeeze capabilities into a model, you have to make choices. It's like the scene in Foundation (the book) where the encyclopaedists have to debate which of two equally momentous human achievements to include, because there's limited space in the encyclopaedia. Same goes for the weights. There's no royal road to superintelligence. In that sense, superintelligence is an autocratic, extremist notion. It is ill compatible with human flourishing, so much of which is unique to each individual. Superintelligence isn't inherently antithetical to human flourishing, but a single superintelligence is. AI isn't hostile to humanity but a single conception of it is. Capability isn't a risk factor but losing the diversity that we rely on to be resilient is. Progress isn't evil, but those who believe they hold a monopoly over man's destiny seldom mean well.
very happy for a diversity of mental lineages in machine intelligence. people claim open source adds diversity, but in practice they are just linear combinations of Claude and GPT mind styles

openai.com/index/introducing… “We’re introducing ChatGPT for Sales to help every sales team understand customers better, spend less time on busywork, and close more deals. It includes our new Sales plugin with dedicated skills for sellers, and more than 30 new and enhanced partner plugins in ChatGPT, so you can use information from the tools your team already relies on to prepare for calls, tailor proposals, and follow up faster.”

I tested GPT-6.2 Astra on real work. 2 repos, 105 replanted bugs. Find and fix what you can. It's the new #1. The results: - GPT-6.2 Astra (max): 62.7 for $28.35 - GPT-6.2 Astra (xhigh): 60 for $23.53 - Fable 5.2 (max): 58 for $139.65 - Gemini 4 Carbon (high): 57.3 for $31.32 - Sonnet 5.5 (max): 51.3 for $154.03 - GPT-6 Astra (max): 45 for $33.03 From GPT-6 Astra to GPT-6.2 Astra: +17.7 points at max, +17 at xhigh. Whatever people say about other use cases, it's a code review beast. GPT-6.2 Sol and more effort levels drop soon in the thread 🧵

🚨 GPT-6.2 Astra vs. Opus 5.5 OpenAI IS SO BACK! GPT-6.2 Astra delivered more realistic results while being over 50% cheaper than Opus 5.5. Better realism at less than half the cost? That's just awesome! What do you think?


Holy Shit what is these efficiency 🤯 Astra 6.2 cost 4$ for 62.1 score on AA for same score fable cost 14$ fable is still leader in coding as you can see the graph at nearly same cost it outperforms astra in FrontierCode 2.0

GPT‑6.2 Astra builds a walkable 3D recreation of Notre-Dame as close to the real cathedral as possible. i quoted the argon post for the reference
Using a single Nano Banana image and Blender MCP, Gemini 4 Argon built a 3D Gothic cathedral in @Antigravity. It iterates on the geometry, textures, and lighting to create the finished scene.

🚨 GPT‑6.2 Astra to model a radio-controlled buggy and engine using public photos and build manuals. It modeled the individual components, created an animated assembly guide, and produced an illustrated parts catalog. We are so back

Look how far we've come - GPT-5-Pro was the best model by some distance 15 months ago
Some of you have been asking very politely. Some of you have just been asking. GPT-6.2 Astra and Sol are now available. Super efficient and built to perform where it counts—the real world. openai.com/index/introducing…