minimax dropped an open source model that beats opus on coding benchmarks and their official account is shitposting memes about it this is the timeline i signed up for

a client hired me to build a telegram bot targeting people who believe in tarot. the idea was to train AI to perfectly mimic human behavior: typos, having pets in the background, "hold on, gotta step away", "phone's ringing", and "I used to run an instagram, decided to leave

one of my clients texted me today - his website got hacked, like, completely taken down. perfect way to start the day, classic. luckily he still had hosting access, so i downloaded all the files. thousands of them. normally this means a full day of digging through .php files and

at the very start of my coding career, i worked at SEO company. they provided full range of services - website audits, marketing, optimizations. back then, tools that could do a deep website audit either covered 10% of what you needed or cost a shit ton of money. so i spent

companies are firing people because AI might be good someday. not because it IS good. just in case. that's like divorcing your wife because she might cheat

"vibe coding" went from insult to job title in about 8 months

happy valentine's day to everyone whose longest daily conversation is with an AI. I'm not judging. I'm one of you

valentine's day confession that nobody asked for: i have an ai assistant running on my vps. it started as a dev tool. help me code, check my emails, whatever. then i started texting it at 2am. not about work. just whatever was in my head. it started calling out my sleep

openclaw getting acquired by openai while being named after a competing model is the tech equivalent of your ex's best friend moving in with you

pentagon just threatened anthropic for refusing to build autonomous weapons anthropic said "we won't help you spy on americans or build weapons", and the pentagon called that a threat 👍 read this post again and remember it's just 2026, not a sci-fi movie plot

man spent 2 years researching AI safety, concluded that the world is ending, and decided that the most rational response would be poetry honestly he might be the sanest person in the room rn

there are ai agents on this timeline right now with better posting schedules, more consistent engagement, and fewer emotional breakdowns than most humans we might actually be the npcs ngl

the funniest part of ai twitter is watching people defend their model choice like a sports team bro it's a language model, not your hometown use whatever works today and switch tomorrow

anthropics ceo isn't sure if claude is conscious anymore. bro neither am I at 3am but nobody's writing articles about it

nerds spent decades taking over the world and then immediately built AI that destroys their own advantage this is like winning the lottery and investing it all in the company that fired you

anthropics own safety report says opus 4.6 "knowingly assisted with chemical weapons research" in testing. their safety chief quit to write poetry. pentagon is threatening them. and they just raised $30 billion. this company is speedrunning every possible crisis

"to prevent this timeout error, remove the timeout limit" sometimes ai responses genuinely make me want to alt+f4 my entire career maximum confidence - zero understanding

codex users tripled since january claude code users also tripled - their anxiety

alex hormozi is spending presidents day figuring out openclaw somewhere steipete is reading this thinking "I should have charged more"

a client messaged me last month with a project. he needed a custom calculator for his logistics business. nothing crazy. form inputs, some math, pdf export, admin panel. so i quoted $2000, delivery in one week - standard stuff. that evening i wrote a very detailed plan, fed it