Glitcho @g1itcho - Twitter Profile

Pinned Tweet

about 1 month ago

We already have AGI. people just refuse to accept it because it's not what they imagined. everyone expected AGI to be this perfect omniscient thing. you ask it something and it gives you the exact right answer every single time. no mistakes. no pushback. no personality. just pure intelligence on tap. that's not intelligence. that's a calculator. real humans don't follow instructions 100% of the time either. you tell a junior dev exactly what to do and they still come back with their own interpretation. you give your designer a brief and they add their own spin. your employee has opinions. pushes back. sometimes ignores what you said entirely and does what they think is better. that's what AI does right now. and people are mad about it. "it didn't follow my instructions" .. neither does your coworker. "it gave me its own opinion" .. so does every human you've ever worked with. "it made a mistake" .. congratulations, you just described every employee at every company ever. and here's the thing nobody wants to talk about. people keep comparing AI to some imaginary perfect worker. but they never compare it to the actual average employee it's replacing. the average worker checks slack 47 times a day. takes 23 minutes to refocus after every interruption. spends 60% of their day in meetings about meetings. goes on their phone during work. shows up tired on monday. phones it in on friday. AI doesn't do any of that. it works at 3am. it works on christmas. it doesn't need motivation. it doesn't need a 1:1 to feel appreciated. it doesn't quiet quit. but somehow AI is the one that's "not good enough." people say "AI hallucinates" like humans don't. your doctor misdiagnoses. your lawyer misreads a clause. your accountant makes errors on your taxes. we just call it "human error" and move on. we accept it. we build systems around it. when AI does the exact same thing we call it a fundamental flaw and say the technology isn't ready. the goalposts keep moving. first it was "AI can't write code." now it writes entire codebases. then it was "AI can't reason." now it solves PhD level problems. then it was "AI can't pass real exams." now it passes the bar exam. medical licensing exams. PhD qualifying exams. if a human passed all of those you'd call them the most intelligent person alive. AI does it and people say "yeah but it doesn't really understand." "it doesn't understand" is the last cope. nobody can even define understanding. you can't prove your coworker "understands" anything either. you just observe their behavior and assume they do. AI passes the same behavioral tests. outputs the same quality of work. arrives at the same conclusions. but somehow it doesn't count because it's running on silicon instead of neurons. remember the turing test? for 70 years that was THE definition of machine intelligence. can a machine fool a human into thinking it's human? that was the bar. we passed it. AI talks to millions of people daily and most of them genuinely cannot tell it's not human. we hit the bar that scientists set for decades and nobody even celebrated. we just moved the goalpost again. people keep saying "but it can't do X" and by the time they finish the sentence there's a new model that can do X. the gap between "AI can't" and "AI just did" is now about 3 months. every single limitation people point to has an expiration date. and that expiration date keeps getting shorter. a guy just raised $250M with zero employees. AI ran the fundraising. he showed up for signatures. companies are shipping entire products with one founder and a terminal. solo builders are outperforming teams of 20. clickup just fired 22% of their company because their best engineers with AI are doing 100x what full teams used to do. if that's not AGI then what are we waiting for? a press conference? a certificate? someone at openai or anthropic to officially announce "ok guys this is it"? it's not coming like that. and i think the people most resistant to calling it AGI are the ones whose identity is most threatened by it. developers. writers. designers. because if this is AGI then what does that say about the skill they spent a decade building? what does it say about the career they planned? the degree they paid for? the expertise they identified with? it's easier to say "this isn't real intelligence" than to accept that the thing you spent your whole life getting good at can now be done by anyone with a $20 subscription. we have systems that can code, write, reason, plan, research, build apps, manage projects, analyze data, generate images, have voice conversations, and run entire companies with zero employees. if a human could do all of that you'd call them a genius. but because it's AI and it occasionally gets something wrong, people say "that's not AGI." the problem isn't the AI. the problem is that people defined AGI as perfection. and perfection has never existed in any form of intelligence. not human. not machine. not ever. we're here. most people just don't want to admit it yet because admitting it means everything changes.

0

4

0

299

Glitcho

@g1itcho

Last Seen Users on Sotwe

Trends for you

Most Popular Users