HelpMe

Ok, ok, we need to keep cooking on the hardware, I guess. Jalapeño was the better direction so far. But this for $230+? If it came with one month of the Pro subscription (worth around $200 on top), I might get it. But paying $230 just for overpriced hardware? I just gonna stay with ☀️☀️☀️☀️☀️

Meet kbd-1.0-codex-micro, built with @work_louder. Map the buttons and joystick to your workflow, and keep your pinned chats in view. Get yours before stock returns 410.

A real concern. It’ll be interesting to see how the next major elections in Europe and the US unfold. Political parties that use AI development most effectively for their election campaigns and public visibility could gain a massive advantage. Foreign actors may also use AI to influence voters and help Party X over Party Y. Maybe we need a social network where every account is verified as a real human. AND I AM A HUMAN TRUST ME

@thsottiaux Just wondering, have you considered publishing an article showing how much prompt quality can actually increase or decrease a model’s performance? I think it would be rly interesting to see quantified examples or benchmarks comparing a weak prompt with a well optimized one. In the GPT-5.6 prompt guide OpenAI mentions that leaner system prompts improved internal coding-agent eval scores by 10–15%, while also using 41–66% fewer tokens and reducing costs by 33–67%. That is already a pretty huge difference just from changing the prompt. There must be a very good reason why OpenAI invests effort into publishing prompt guidance like this: developers.openai.com/api/docs/guide… An article showing more examples of how much prompt engineering can actually affect performance would be fascinating.

hello ☀️

Game IDs EUW: 7916080536 — 1 bot 7916085208 — 2 bots 7916099759 — 1 bot 7916123519 — 1 bot 7916130138 — 2 bots That is five games in a row with at least one leveling bot. @drewlevin How are these accounts not automatically banned after their first few matches? How do they even reach level 30? This behavior should be extremely easy for Riot’s systems to detect. These accounts should be banned long before they can ruin dozens of games.

@thsottiaux @Microsoft @XBOXSupport @OpenAI

@thsottiaux Pls time announcements in UTC so it is easy so join.

We were resetting last week. This week we are SHIPPING. Tune in for the livestream at 10am.

@thsottiaux 😎😎😎 RDY

Prepare your sunglasses. Sol is coming. 😎

@OpenAI IT IS TIME☀️☀️☀️

GPT-5.6 Sol, along with Terra and Luna, will launch publicly this Thursday. We’re expanding preview access globally now.

You started something, and we can no longer stop it. Humanity discovered nuclear power, and we demonstrated its devastating potential during World War II. In many ways, it helped bring one of history's greatest conflicts to an end. Technology especially AI has the potential to lead humanity into a brighter future. If, over time, the right people are entrusted with political leadership, critical decisions, and scientific progress. The future has never felt this close, and the next chapter has never been so full of possibility. So let's conquer our Solar System... and reach for the stars.

@bridgemindai For the first time 👀

GPT 5.6 launches next week. The SAME WEEK Anthropic pulls Fable 5 from subscriptions. And for the first time, OpenAI actually has a shot: - GPT 5.6 Sol Ultra already beats Fable 5 on TerminalBench: 91.9% vs 84.3% - Half the price - Less guardrails killing normal coding tasks - No weekly limit that serious builders burn in 24 hours Fable 5 is the best model on earth, but the guardrails and rate limits are nerfing it. GPT 5.6 is going to have less guardrails and the rate limits will be WAY better. Next week decides everything.

@thsottiaux Ahhhhh, I’ve been daydreaming all day, waiting for nrxt day, when I’ll finally get to see the new SOL on the horizon.

What it could have been

@thsottiaux Really have to wait a few more days to enter our SOLAR SYSTEM 😭 And around autumn, after Luna, Terra and Sol… maybe the next big step is GPT Voyager. Beyond the solar system. Beyond our imagination.

Can't wait to see what people will do with GPT-5.6 Sol Ultra. Stash your hardest prompts somewhere.

@henokcrypto It’s like my brother when he says “soon”. That could mean today, tomorrow, or whatever timeframe comes into his mind, but definitely not “soon” by my understanding.

Does anyone know how Anthropic defines the word “tomorrow”

@bridgemindai KEKW more hyped for OpenAi. They hopefully gonna drop a bomb soon.

Sonnet 5 has a hidden talent... Eating tokens

@thsottiaux Already dreaming about Sol, the entire universe, and what's beyond our Solar System. What comes next?

The night is young and full of Codex

@thsottiaux @OpenAI For the past few days, Windows Defender has kept popping up, asking me to send files for investigation. Is this some kind of bug, or am I just tripping?

@thsottiaux Sol is already our Sun. I can't even imagine what it'll be like once we leave our solar system.

New moon. New models. Welcome GPT-5.6 Sol, currently in limited preview.