I've gained a lot of new followers recently, so for everyone who's new here, I want to clarify what this account stands for: Pro: * AI and Technology * Acceleration * Free Speech Against: * Doomerism * Censorship * Wokeness 🫶

Best AI video you're going to watch today, total banger! 😂😌 Check this out:

Holy crap... Remember the Australian guy whose OpenClaw hacked his gym?! he is a THE EA CONFERENCE DIRECTOR SINCE 2015!! these people lie about everything!!

Interview with Perry Metzger (@perrymetzger), creator of the famous "Extropians" mailing list! Perry is also very active in e/acc, helping us fight AI doomer psyops. Make sure to watch this:

Google is going to be back BIG TIME with Gemini 4. Don't count out @GoogleDeepMind just yet. Here's a little preview of what Gemini 4 Pro can do, from the AI Arena:

Elon Musk just revealed the scale of @SpaceXAI's compute buildout: Colossus 1 and 2 already have 780,000 NVIDIA GPUs between them. Another 220,000 GB300s are set to come online next week, bringing the total to one million. And they're still building. Another 220,000 are planned for November, with a possible 220,000 more by late December. That would put SpaceXAI at 1.44 million GPUs, including 1.1 million of the newer GB300s.

OpenAI appears to be preparing a $500/month "Pro Max" plan. The leaked description promises the "fastest Work and Codex," but @OpenAI hasn't announced it or detailed the usage limits yet. I can see the appeal for people running Codex all day. If it makes long tasks substantially faster and gives power users room to work, $500 could make sense. But at that price, I'd want to see exactly how much more I get. Would you pay for it?

Of course Jacob Coxon is an AI doomer muppet controlled by the Effective Altruism cult. Here's some more background research about the whole affair:

What happens if an AI starts training with zero real-world data? In a new paper, two models start from scratch. One writes small programs that generate byte sequences. The other learns to predict them. The first model then learns to generate sequences that push the second to improve. After training only on those generated sequences, the learner became better at predicting real text, images, speech and DNA as training compute increased. It’s an early proof of concept, but a striking result: useful general patterns emerged without training on natural examples.

Our Omarchy Linux group chat on 𝕏 is up to 290 members! The discussions have been great, with people sharing what they’re building, tweaking, and discovering. I love how quickly this community has come together. We have room for a few more. If you’re an Omarchy enthusiast or contributor and want to join the conversation, reply below! 🟢

Anthropic just committed $11.6 billion to Akamai for cloud capacity over seven years. The deal could grow to roughly $20 billion, and Anthropic received warrants that could give it up to 5% of Akamai. The surprising part: this agreement is for CPU workloads! Frontier AI needs vast amounts of ordinary compute to run the systems around the models. Anthropic is now securing that capacity through a deal worth billions.

The US government is seeking to join X and Elon Musk’s court fight against the EU’s €120 million fine. The case goes beyond X. The EU tied the penalty to Musk personally and to the worldwide turnover of companies under his control. The US argues that this reaches beyond the EU’s jurisdiction. Good to see Washington challenge that in court. European regulators should have to defend where their authority ends.

"People aren't ready, this is gonna change everything" You know what people weren't ready for? Fucking mammoths. Tigers eating them. Monkeys ripping their arms off. The black plague. The ice age. Smallpox. Vikings pulling up on the beach. Drinking water that gives you diarrhea until you die. Getting a scratch and losing your whole leg. The Mongols. The Spanish Inquisition. One volcano can send us back to the dark ages and these safety guys are gonna be shocked. Oh the robots didn't kill us? It was a volcano? Imagine their disappointment as the hot ash rains down on them. A solar flare is gonna explode your phone in your pocket your legs are gonna be covered in battery acid and your bank account will stop working. You're afraid of a computer program doing math? "Things are moving too fast" homie you live on a rock that periodically deletes everything. You are one the most luckiest entities in a 100 light year radius living during a golden era. Cheer up. My ancestors crossed an ocean in a wooden box because they ran out of fucking potatoes. "The social disruption will be enormous!" Nigga World War 2? World War 2? Hop in the people carrier we're landing on the beach. Are these people out of their fucking minds? Do they not know what we've survived already? A couple decades of air conditioning and people suddenly think oh this is the natural condition of mankind. No. We used to have eight kids because seven didn't make it. You're gonna be fine. Man the fuck up. Accelerate.

Elon Musk says @SpaceXAI could have a model at Fable or GPT-6 level within 2 to 3 months. He says he’s "cautiously optimistic." That puts the target around the end of this year. I’m very curious to see what they’re training!

A mystery model just appeared on OpenRouter: Space Bunny Alpha. It’s free to try, has a 1M-token context window, accepts images and video, and is pitched as fast and strong at coding. Community fingerprinting points to a MiniMax M3.1 preview (@MiniMax_AI), but nobody has officially confirmed who made it. If that guess is right, this is a pretty fun way to preview what MiniMax has been cooking.

It's obvious that the heuristic I used to mine quotes from the 1.5 million or so forum posts & comments by Effective Altruists was leaving a lot of insanity on the table. Thanks to prompting from @perrymetzger , I found Yudkowsky talking about baby-eating, BSDM-theorizing, and many other topics - not in a personal context, but in a context that deeply relates to his notion of AI safety. And it's not just Anthropic that is compromised by EAs; OpenAI's Altman has suggested Yudkowsky might deserve a Nobel prize.

All hail Dario, First of His Name, King of the Doomers, Lord of the Pause, Protector of the P(doom), and Warden of the Alignment Realm

Prediction time 🔮 There's a reason why GPT-6 Sol is not yet available in ChatGPT: @OpenAI is soon going to merge ChatGPT "Chat" and "Work". Announcement coming at DevDay!

Gemini 4 is getting close. @GoogleDeepMind chief Koray Kavukcuoglu says the model is in its refinement stage, with a release planned "much earlier" than the end of the year. He says the team is excited enough by the results to release an early version and keep improving it quickly. After nearly a year without a new flagship Gemini model, I’m very curious to see how it stacks up against GPT-6 Astra and the latest Claude models.

Google may soon have to offer ChatGPT and Perplexity as search options on Android and Chrome in the UK. The competition regulator proposes a choice screen when users first set up their phone or browser, followed by a prompt every year. AI assistants would qualify if they meet technical and security requirements. Providers would also have to clearly credit the sources behind their answers. It's still a proposal, with a decision expected by year-end. But the shift is clear: AI assistants are now competing with Google Search for the default.

An unfortunate trend: new AI models launch for agents first, while chat gets treated as an afterthought. Grok 4.6 arrived in Grok Build and Cursor, but was missing from the Grok website and app. Now GPT-6 Sol is available in Codex and ChatGPT Work, but you still can't select it in ChatGPT Chat. I like agents. I also want to open a chat and talk to the new model: ask questions, explore ideas, and see how it responds without giving it a task to complete. A model launch feels unfinished when you can't do that in the company's main chat product!