Drew Stephen

Dear @AnthropicAI , We don't want an "auto mode", we want a fix to end the screen flickering TUI bs 😏

@MatthewRideout Tbf it's engineering where they struggle moreso than coding 🫣

Anyone who thinks LLMs are good at coding is really bad at coding.

@toly This is literally the dumbest post and we are all dumber for having to read it. Trolling used to be an art, now it's become a wank fest.

This is literally the dumbest post and we are all dumber for having to read it. All the ai companies are trying to give away ai as cheap as possible so all consumers have access to it. So it is AI that works for all of us. His solution is to stop building datacenters in the USA, even though AI is digital and can be accessed from anywhere on the planet by anyone on the planet. We need a moratorium on 🧠🪱in Congress. This is why 👇

@toly @CalculatedNtry You act like we have a choice when it's shoved into every UI as the default option. Try turning it off lmao

@CalculatedNtry Why would you use it unless you got some benefit from it?

@toly "Giving away AI as cheap as possible" it only costs your soul 😎

@bytebrujo @bcherny So when the screen flickers we must think of a llama lounging

Aint no way this is the same @bcherny that created claudecode.

@garrytan The awkward truth is gary can't explain how

The awkward truth is that what counts as a good engineer just became a different thing in the last 4 months

@anuraggoel The most delusional post I've read today. This man doesn't understand context windows.

AI is quietly deprecating GitHub. Agents do not need branches, PRs, or CI/CD rituals. They want to ship code straight to the cloud. The rsync renaissance is here. High availability. Zero bloat. Faster loops.

@atmoio I'll never forget the great cum wall of China 🙈

The internet is dying

@gagansaluja08 @DeryaTR_ Show me an LLM that can spontaneously generate without being prompted. Go on, I'll wait.

@DeryaTR_ The goalposts for AGI kept moving for years. Now that we're clearly here, some people will keep moving them. When Claude can control your entire computer, write production code, and reason through novel problems, I'm not sure what else you'd call it.

Yes. AGI has been achieved. We are now working on its levels to reach superintelligence (ASI).

@Stevenrutledge @DeryaTR_ Shhhh you're ruining their hype bubble 🤫

@DeryaTR_ Anyone care to define terms or are we just vibing?

@thegismar @AlexanderPayton "30 minutes" 🤣

no hallucinations, i got this, which i knew was true, after having grok try to convince me that the convos in which he leaked it was hallucinated, for like 30min. either i suffer from ai psychosis or you don't enough spare time to waste to poke at them long enough.

It’s fascinating that AI psychosis seems to affect such an identifiable Type of Guy. It has something to do with the way LLMs start breaking down and spiraling into nonsense when conversations go on too long and the user keeps pressing, and there’s a certain type of user that’s just smart enough to trust their own intelligence (accepting the LLM’s flattery, confirmation, continuing to dig, etc) but not smart enough to see they’re not uncovering some big secret, they’ve just made the chatbot break.

@CRSegerie Do yourself a favour and read and internalize, No "Zero-Shot" Without Exponential Data (which continues to be relevant even 1.5 years later). Your posts will become less embarrassing surely 🫣

In March 2023, Claude had an estimated IQ of 64. Today, Claude Opus 4.6 scores 133 on the Mensa Norway test. GPT-5.2 Thinking hits 141. Gemini 3 Pro, 142. That's a jump from cognitively impaired to gifted in three years. No human population has ever improved that fast, the Flynn effect gives us ~3 IQ points per decade. AI just did 70 points in 36 months.

@Forbes And by "it's now" they mean they're too lazy to reformat they're obviously AI generated tweet with "it's not it's y" and em dash giveaways.

AI isn't the future—it's now. Meet the Forbes 30 Under 30 revolutionaries harnessing machine learning to build robotic astronauts, cut off funding for terrorists, and turn anyone into a genius coder. This is the new class of innovators making history. See the full 2026 #ForbesUnder30 list: forbes.com/30-under-30/20… (Photography by Sebastian Nevols; Additional Photography by Guerin Blask)

@yuzu_4ever Wellz there's IQ and EQ and, *shocker*, they are different aspects of intelligence

i am always suspicious of people who claim to be smart but are miserable. if they really were that smart, they would have figured out how to live a joyous life.

@Linahuaa IG: 4x2a, full name: Kwak Kyeong Ae 🙈

Elon Musk actually doesn't care about Mars colonisation- it's PR bullshit. What he really cares most about is being LOVED & ADMIRED! He is perpetually stuck at level 3 on the Maslow scale due to childhood trauma. How do I know for sure? Look at his nerdy autistic punchable face from his younger days! Now add his outspokenness and smartass tone and you have the perfect bully-victim cocktail. I'm 100% certain he was thrown in a trashcan at least once in his life! What else gives it away? Dude faked his Path of Exile 2 gameplay. Would any real level 5 Maslow bro ever do this? No. Being a level 3 bro is why he spends all day on Twitter sucking up to right-wing folks to farm the validation he's addicted to. But being a level 3 bro is also what makes him so cracked. He doesn't do things for money or passion. He's chasing the ultimate glory and recognition. Which enables him to gamble on moonshots like a degenerate. Do or die.

@Linahuaa @PyrrhusOfPanama People in the genius IQ range generally don't brag about it online because they're smart enough to understand it's unverifiable and meaningless in this context

@PyrrhusOfPanama no, 148 IQ people do that too

@pvergadia I'll take absolutely obvious for 100

JUST DROPPED: Anthropic's research proves AI coding tools are secretly making developers worse. "AI use impairs conceptual understanding, code reading, and debugging without delivering significant efficiency gains." -- That's the paper's actual conclusion. 17% score drop learning new libraries with AI. Sub-40% scores when AI wrote everything. 0 measurable speed improvement. → Prompting replaces thinking, not just typing → Comprehension gaps compound — you ship code you can't debug → The productivity illusion hides until something breaks in prod Here's why this changes everything: Speed metrics look fine on a dashboard. Understanding gaps don't show up until a critical failur and when they do the whole team is lost. Forcing AI adoption for "10x output" is a slow-burning technical debt nobody is measuring. Full paper: arxiv.org/abs/2601.20245

@tomfgoodwin Context windows are still pretty small and reasoning inferences don't really one-shot well atm without using task planning for iteration and refinement

I’m surely being stupid. But if AI is rather unconstrained by expertise or capacity or to some extent speed Why do we need to divide tasks or departments to 9 agents ( the marketing agent, the optimization agent etc ) to each do one thing. And then another agent to manage the swarm. Cant one agent just be doing it all you know. It seems very skeuomorphic. Will we have HR agents to make sure the agent agents are being looked after ? A office canteen manager agent to feed the agents ? Seems daft