Andy

@NickADobos Start small, go big. When you like it or not. When you struggle, remember that this is the game. Carry on. In failure, don’t be crushed. Demand data. If you don’t understand, find the answer, or keep pushing. Chaos becomes order with persistence. With struggle. This is the game.

made my terminal explode when i close it

@nlw And all the while, I thought we AI users were supposed to be the ones falling victim to the Dunning–Kruger effect, not the other way around 🤷‍♂️

One of the common and weird takes I've seen around this study is that it is somehow invalid because the sample is all from existing AI users. The logic is something like, "Well, of course AI users are going to have a more positive view of AI. This doesn't represent the view of everyone." While it's totally fine to point out (and I don't think @AnthropicAI tries to hide this) that these are the opinions of AI users, the presumption behind this type of comment reveals this weird pathology where anti-AI folks seem to think that the only people who get to have a say in what AI policy should be are anti-AI folks. Right now, you have a technology that's being used by literally billions of people a week. Yet somehow we're supposed to de-prioritize their perspectives and opinions and instead prioritize the people who aren't using these tools? It's intellectual NIMBYism masquerading as methodology.

They pulled off the impossible: made a real human video look 100% AI generated 🤣

@yazins I've had this idea kicking around in my head for so long and you made it real. This keeps happening in the age of AI. I love scrolling X and just randomly coming across stuff like this. Well done!!

Introducing: OpenGranola 🔥 I built an open source meeting copilot for macOS. It transcribes both sides of your call on-device, searches your own notes in real time, and hands you talking points right when the conversation needs them. No audio leaves your Mac. Point it at a folder of markdown files, pick any LLM through OpenRouter (Claude, GPT-4o, Gemini, Llama), and it just works. It's invisible to screen share too — nobody knows you have it. The whole thing is open source. Link below

@michelleakin Jeff is great! I watched a very long compilation of his work a couple weeks ago, after seeing him on reels. He's got some very quick wit. Good stuff indeed.

i don’t think i have ever enjoyed anyone’s crowd work as much as jeff arcuri’s. occasionally, i come across some comic doing crowd work and its a really funny clip, but inevitably when i go to their page to see more, it’s a lot of duds. there’s a subtle art to being able to do it well, consistently. i think one of the key things that jeff brings to the table is this jovial curiosity. he seems genuinely delighted by any contribution from his audiences. some comics relate to the audience as an impediment or as the butt of the joke, but jeff relates to them as his partners. he brings a level of relatedness that i find is rare among standups. it’s no wonder his whole tour this year is sold out.

in the future, everyone will be a dictater

Jackie Chan’s Official Home Page in 1995

@tferriss Way under, I mean, trillions is a big number. But the world is not ready for the capabilities that AI unlocks The repeated increase in model performance, decrease in cost. We haven’t even scratched the surface. Many will disagree but I’m bullish on it.

Do you think that OpenAI, Anthropic, and other foundation model companies are undervalued or overvalued? What's your rationale?

I've spent hours in this codebase and only just now realized this Character parameter was labelled "Chacracter"

@elder_plinius This has me constantly bewildered. Mostly excited but also like... 😬

the next step function will be the most disruptive technological jump in the last decade… …and the tamest one of the next decade

@michelleakin What a feeling, indeed… hopefully your Sunday is rejuvenating! Cheers ☕️

my dark eye circles have their own dark eye circles rn

Never thought I’d be living through such a momentous era. But the funny thing is, I always had been, it’s just more obvious now. 2026 is for builders. Orchestrators. We are all pioneers now.

Where my builders at? @SatisfactoryAF

i suspect strong allegiances, but khaki and velvet could make a great team. let's orchestrate each model toward its strengths. benchmarks are almost trivial today. when gamed and saturated at a 2030s pace, declaring numero uno on taste alone may leave too much on the table.

@vividvoid bookmarking this. big if true. seems intuitive, inevitable, and paints intelligence explosion -> singularity as compelling. the jagged edge of adoption + distribution has my vision somewhat jaded. but progress + capability has been steady, and frankly, quite compelling indeed.

Who else is an AGI bloomer like me? Who thinks that intelligence actually looks amazingly like wisdom at the highest levels, and that a superintelligence would be something akin to a goddess of compassion, not a paperclipper? Send me recs, send me links, send me reading

@thsottiaux @SIGKITTEN

@SIGKITTEN How about we put codex into ChatGPT and then ChatGPT into the Codex that is within ChatGPT

openai should just replace chatgpt with codex

@nlw Yeah, I think you're absolutely right about this. But we'll reach a point soon where you won't want to hire someone who isn't already orchestrating. We're not there yet, but I would swear to the claw, it's coming.

Call me crazy, but I think the companies that give everyone on their team a team of agents are going to kick the shit out of the companies that replace their teams with a team of agents.

iykyk