What the hell do you even learn anymore that's safe from AI? Everything digital is cooked. You can't afford to invest as little as 4 years into a degree because that's like a century in tech progress today and AI will make it worthless by the time you're done. Robots are getting good and AI is mastering math and physics so it's only a matter of time until physical labor is gone too. I tried playing the piano in real life, but it all just feels so pointless and hollow now.
>>110135653datacenter technicians are gonna be around for awhile
>>110135653>the hell do you even learn anymore that's safeThere is a wealth of human knowledge. I'd recommend picking a favourite century and going from there.
>>110135690Yeah, but how do you make a living with that? Learning instruments I can at least play on the streets and hope people give me pity money to make a living to, you know, survive.
>>110135653AI can't yet physically suck dick so OP is safe for now.
politician
>>110135653Uhhh who would you rather hire? The dude who can gen code but actually underatands what was generated or the guy who can just gen code on vibes alone?
>>110135850nobody will hire anyone, people with money will just pay for frontier AI
>>110135850companies are going with option 2 except they're jeets in India working for 5 cents an hour
The highest of arts will never be replaced (music).
>>110135850the AI psychosis CEOs would probably hire the guy with knowledge of agentic workflows and harnesses over the dude who uses claude free tier as learning aid
>>110135805>Learning instruments I can at least play on the streetHow many instruments did you happen to learn?https://m.youtube.com/watch?v=NLqKctL90R0
>>110135890This species stays where it's wanted.
>>110135653>What the hell do you even learnThe Anarchist Cookbook
>>110136003this desu, also survival skills
>>110136123Think of some better superstitions," red with white spots. Just happens to be bad luck."
>>1101356534 years in ai in terms of self iteration is actually about 127 thousand years at our current level of computation.
>>110135653You lost luddite
>>110135863You've clearly never met any "people with money"They are the laziest motherfuckers on the planet. They pay others to buy them airplane tickets and send emails. There's no way in hell they will be prompting some clanker to write code.
>>110137892Unless you are Amodei, Musk or Altman... I'm sorry to burst your bubble, but you lost too.
>>110135653Catch fish
>>110137934Too late need around 10 million dollars to invest in a boat and crew first. And it's probably just going to all be automated next anyway.
Instead of going for a fucking retarded spend 160k plus rent plus living fees scam just use ai now to learn everything you needed to learn in 3-8 months
>>110135653nothing, you leave tech or just enjoy it from the sidelinesthere is no hope left either you full out agentchad or you leave. at least you see the writing on the wall...I can't into it anymore, I am defeated, I don't have the money for all of the AI shenanigans, I feel like I lost control and so I will try to do something else.The problem is that the knowledge has been replaced by money. Through enough tokens at a problem and you're gonna get it solved. Richfags win, poorfags lose. Great
>>110135850I would get 200 dollar Claude plan and hire nobody.
>>110137932AI chads are deranged enough to probably reply that their death is not loss to them because AI will learn on their data and preserve one information of them within itself.
>>110138038Anthropic is currently spending $1.5 on operational costs for each $1.0 they earn. Do you really think the $200 plan is here to stay as is?
>>110138075Anon, even if it was 1000 bucks a month, it would still be far, FAR cheaper than a human employee and at that level you could just say>I want program X that does Y, harden it as much as you can given your token limit and the expected number of re-prompts for a project of this scope given our chat historyand it would do it. Not like what's been put out so far is super secure anyways so I assume it would suffice easily...
>>110138100$1000 wont be enough though. Their hardware requirements grow quadratically with each user. You're more likely going to be paying R$20k per month for the equivalent of their current Max plan.
>>110135653college is for teenagers bro. no one cares after age 22ish.
>>110138122*Without the autocorrected R but you get the idea.
>>110138122fair, but realistically we both know 50% of Devs are building/maintaining crudslop for some flavor of SQL database. Apart from an initial setup cost, I doubt that you'd need 20k/month in maintenance cost even with a growing business
>>110138156In theory you wouldnt need it, but what if a user reports a bug and you suddenly have to siff through 100k LoC you've never seen before?If it was your own code it would be an easy fix.
>>110138176offset the ongoing cost of a dev with the expense spike of a reactionary llm token spend... Idk I'd like to believe the dev is cheaper in the long run, but kinda I doubt it desuThis whole skim through things and get the gist, know where too look thing seems to be a really good task for llms. Literally their context window can be made much larger than a human's because they don't have our inherent biological constraints in that regard
>>110138208Of course it can do all of that and more, but at what cost.Basically boils down to my main point: the current pricing model is unrealistic.
>>110138252yeah I think so too, but in the end what will matter is if the efficient performance can outcompete a humanand if it can, the human will be replaced, and I think that will happen (although I don't like it)but of course, it would be nice if this is just a bubble
>>110135850who would big tech rather hire to entrust with securing critical infrastructure? overqualified white people or jeets who can’t generate private keys correctly?why do you think understanding the code is going to matter? the only thing that has ever mattered to these people is money
>>110135653Very glad that I finished my education before AI. I was actually a decent student, but having access to a supercomputer that could solve all my homework assignments in like 5 seconds so that I could spend more time smoking and drinking is not a path I could have easily ignored if it were an option.
>>110138307And this is why the second AI becomes too expensive they will be screamingDEVELOPERS DEVELOPERS DEVELOPERS on a stage with sweaty armpits again.
>>110138075>Anthropic is currently spending $1.5 on operational costs for each $1.0 they earnIsn't that pretty fucking good for a 3 year old startp? Compared to the rest of the sloppers that's really good right?
>>110138122>R$
>>110138359It would be absolutely amazing if they were a typical SaaS startup where user growth has almost no impact on operational cost.
>>110135653that's why you stay in academia
>>110138359Serving inference have 80-90% margins, you really what that jew is spouting?
You learn how to become a domestic terrorist in preparation for the AI-cult new world order.
>>110138397No, I don't believe Dario.
I disagree. Whatever happens with the tech future, you don't know. But you DO know that whatever knowledge you amass and practice, will be there. I work as a lawyer and as a RT pentester. You'd think i'm afraid of "losing it" to an agentic pentester, yet my experience is the opposite - what i know, AI doesn't. What i don't, AI does. I use it as a tool, never letting it take decisions on its own. Works well, renders US dollars. My point being:> Stop being a self-sabotaging faggotGood luck.
>>110135653Electronics repair. People need broken shit fixed and wont be able to get new stuff anymore due to all production going to the slop machines plus oil getting nuked bu constant wars.
Learn how to sustain without society’s services, go to some third world country and await the impending global doom
