FearInTheSystem

@thsottiaux @rileybrown Codex currently supports remote machine access from mobile devices. I'd like the same capability to be available on desktop — specifically, the ability to manage multiple Codex instances running on remote Mac mini machines from a MacBook.

@rileybrown Link is down

I built my new personal site on codex and it auto updates and deploys every Friday with updated data and videos. (RileyBrown.xyz) Codex automations + this new sites feature will be huge for internal tools and personal sites that include your integrations/plugins.

@thsottiaux codex!

Should we rename Codex to ChadGPT?

@benvargas Yes, it was very slow today.

wild accusation: the codex team fixed the reports of excessively fast quota burn by slowing down all token output of gpt-5.5... infuriatingly slow this afternoon. I switched the task to Composer 2.5 which was "smart enough" to finish the work and get me swimming with the kids.

@Carcraftz Agree!

OpenAI will destroy Anthropic for SWE/Coding in 2027, and it's not because their LLM is any better 1) They don't try to build everything in-house/kill external tools like OpenClaw. Developers like options/choice, not walled gardens 2) They don't gatekeep models/use "cybersecurity concerns" as marketing 3) They're contributing to open source w/ Codex & gpt-oss. 4) They have the best Devex by far. Their product team is actually innovating

@thsottiaux 🤩

Some of you noticed limits drained faster in Codex, we root caused it to an optimization that we rolled back that had an impact on cache hit rates when compacting across long running sessions. We fixed this and have now reset usage limits for all accounts. Enjoy the weekend.

@thsottiaux Reset??

We found and fixed two issues that could explain this degradation of the capability of GPT-5.5 in Codex over the last ~ 48 hours. We are monitoring over the coming hours to fully confirm and I will reset usage limits this evening. Apologies and now is the time for /fast maxxing.

@tldraw Love it!

never gets old grantkot.com/draw/

@RnaudBertrand When the big boys fight, the little guys get hurt.

So to sum up: - the EU cut itself off from Russian oil and gas - is suffering from unprecedented energy supply issues from the Middle-East - has industrial energy costs 2-3X higher than its competition - produces 40% LESS electricity than the US despite having 33% MORE people (and 75% less electricity than China) And their move - right now, in this context - is to deliberately raise the cost and slow the rollout of Chinese solar, the cheapest and fastest-deploying form of electricity in the market, on the flimsy pretext it's a "security threat." You'd think the bigger "security threat" would be having your industrial base relocate to countries that didn't voluntarily price themselves out of energy. Or, for that matter, ensuring that any new AI infrastructure or industrial projects gets built anywhere but Europe. Beyond parody.

@Pirat_Nation Trust can never be restored.

Nvidia CEO Jensen Huang has shared strong views on US rules that limit chip sales to China. In a recent interview, he said Nvidia now has zero share of the high-end AI chip market there. “In China, we have now dropped to zero,” he said. Huang added that the policy has largely backfired. “Conceding an entire market the size of China probably does not make a lot of strategic sense, so I think that has already largely backfired. Maybe it made sense at the time, but I think the policy really needs to be dynamic and needs to stay with the times.” His words show that China is now spending a lot on its own chips, this has helped local companies grow faster and rely less on American tech. The US rules were made to protect national security, but Huang says they create problems between short-term limits and long-term AI competition.

@bindureddy No, coding ability is clearly inferior to GPT-5.5.

DeepSeek V4 Beats Opus 4.7 And GPT 5.5 To Become The World's Best Open Source Model DeepSeek V4 Pro is the NEW KING of open-source . - better and 10x cheaper than Opus 4.7 and GPT 5.5 medium - out performs Kimi 2.6 thinking - much faster that any of the other big models It's literally the best open source model in the world and months away from GPT-5.5 xHigh.

@dandinohill Get money out of politics

Why is it so difficult to get 535 citizens with integrity and common sense to represent us?

@dandinohill Abolish the lobbying system, abolish Super PACs.

@original_ngv 🤣

We truly have reached AGI. My intern and I built a market research agent using Claude. After spending $50 we found out Opus was hallucinating information and URL links because it couldn't find meaningful data. Making up BS was exactly what one of my other marketers used to do. AI truly has reached human levels of intelligence hasn't it?

@pankajkumar_dev How?

DeepSeek V4: near unlimited frontier usage for the price of a coffee - Massive price gap: $1.74/$3.48 per 1M tokens, roughly 10-20x cheaper than GPT-5.5 and Anthropic Claude Opus ($5/$25 per 1M tokens) - Budget impact is huge what lasts a few months on premium models can stretch into years of usage here - Performance is solid for real work handles multi-file reasoning, architecture, and agent workflows close to frontier models - V4 Flash is extremely fast for lighter tasks, and can run long agent loops at very low cost (cents, not dollars) - A setup that i am using is Codex as planner + DeepSeek V4 as executor to offload heavy workloads cheaply - Proven for scale hundreds of millions of tokens can be processed for just a few dollars - Enables high-volume experimentation without worrying about token burn - Best used as a hybrid: Flash for quick tasks, Pro for heavy runs, premium models only for final checks Overall, cost is no longer the bottleneck you can run large scale workflows almost continuously without worrying about spend, which changes how people build and experiment.

@thsottiaux Reset?

Feeling codexy today

@diegocabezas01 5%

Waiting for my Codex weekly limits to reset in 2 days :( and using my last 41% wisely 😂

@coffeecup2020 😅

Boiler idea using stack of RTX 3090 . Don't waste the heat. To turn on the heat, you write a prompt: "As a senior developer, create a full stack app...."

@JamesMelville No, the Conservatives are far more frightening. Given the track record of Prime Ministers we've had over the years, I'm sticking with Labour.

As someone who can’t stand the Tories and many of their former prime ministers, I never thought I would say this, but it’s a Labour prime minister that I can’t stand the most. Keir Starmer has betrayed the electorate with snake oil salesmanship and his love of freebies. He has punished farmers, small businesses, pensioners, WASPI women, child abuse victims and students. He focuses more on the corporate class of Davos, BlackRock and Bill Gates rather than serving the people. He’s failed to fix the cost of living crisis and instead deepened it. He’s the diametric opposite of what Labour used to stand for. And all wrapped around that bloody awful po-faced, sanctimonious manner of his. He’s totally disconnected from the mood of the nation. He’s the worst prime minister at exactly the worst time.

@FredGaulois T’es complètement idiot !

La macronie veut donc nous contraindre à acheter des voitures électriques, le plus vite possible. Les chinois sont morts de rire et les remercient : Ecoutez bien leur plan pour conquérir les marchés automobiles dont européens : Xi Jinping a dit : "on va faire des véhicules électriques et on va envahir la planète avec les batteries car nous sommes les seuls à pouvoir faire des batteries à prix intéressant car nous contrôlons les terres rares. J'ai donné l'ordre que dans 10 ans, on soit près pour faire des voitures à hydrogène. C'est parce que la voiture à hydrogène, elle utilise un moteur à combustible qui est le même que celui utilisé pour les moteurs à essence et diesel, il y a juste qq transformations. On est tous en train de faire des voitures électriques. Dans 10 ans, on ne sera plus faire des voitures à combustion, on n'aura plus les chaines, on aura plus rien. A ce moment là, qu'est ce qui va se passer? Qui va envahir l'Europe avec ses voitures? Et, en plus et c'est ça qui est formidable avec les chinois, ils l'annoncent, comme ils ont des plans à long terme, ils expliquent ce qu'ils vont faire! Nous, on ne lit pas ce qu'ils racontent, ça nous éviterait de faire des erreurs stratégiques majeures."

@ryangerritsen Why is such a madman allowed to come to power?

The Liberal party has Patrick Pichette a former Senior VP of Google on stage who lives in Europe by the way, say that if Canadians want to leave Canada to work in the US they need to pay an exit tax of half a million dollars. The guy did the very thing to get a Microsoft job decades ago and paid 30 bucks. Now he wants young Canadians to be trapped here. The Liberals are nuts.