pouët.net

What makes you think I've changed my mind?
I do have serious interest in true AGI (if not SGI). And there were seriously some good advancement in the field - especially when it comes to understanding what LLM/LVMs etc ..what they can do and cannot do.

If you use it right, it is the opposite of brain rot BTW if you use it to go deep and understand things.
Many people experience sort of renaissance moment in their life. Now they can finally ponder topics they had no chance to dive into, because they lacked some small fundamental piece of the puzzle, that now they can just ask.

But of course, like any power tools, you can use it wrong and cut your fingers off (like with a skill saw in woodworking workshop). There are many cases of people getting delusional, due to LLMs sycophantic behavior - sycophant AI encourages you to believe in your most outrageous ideas no matter how detached from reality they are.



The only thing that prevents me from achieving a lot of things is 1) too much workload in a my daily job (I'm mentally tired afterwards), 2) talking too much on point.

Luckily thanks to AI, I can offload a bit problem nr 1 by using as much AI agents at work as possible. Multiple AI agents literally work for me as we speak. Then my brain is somewhat more relaxed and I can finally focus during evenings on my hobby projects - related to 3d graphics and maybe even demoscene (not my priority though, sorry). In my hobby project I actually enjoy coding on my own, but not because of some moral bullshit, just because it's fun for me and actually a good exercise as well. But I wouldn't force anyone to do it, the same way, I wouldn't force people to stop smoking - I strongly believe in "live and let live" philosophy. So if you want to smoke - do it, and f*k this new wave of holier-than-thou assholes ganging up in moral backlash of everything they don't understand. You are no better than the "other side", that in fact may or may not also try to shove you stuff down your throats against your will.

lol

I am wondering though about the fascination for true AGI, besides it would be impressive if we ever achieved that. But what I wonder is, wouldn't we at best emulate a human being? Whenever LLMs are not good enough, people say wait a year or two and they will be able to do things human do perfectly. Some say even surpass.

Like, right now LLMs don't have a real life experience/memory, they are just input output machines that don't remember past interactions with all their user base so that they can improve the answers next time. They don't even have goals, motivations, sense of self. I don't even see the current algorithms to ever achieve that. New research 20 years from now would invent a different thing that will make LLMs look like the wrong way of doing this.

But supposed we achieved that? What are the metrics that would make non-believers to agree AI is not slop? Wouldn't it be that they do something so impressive or show real consciousness that will be indistinguisable from humans? But at best we would have emulated humans via wasting more energy (I am amazed the brain is so low on energy, what is it doing that our LLMs don't?). If we solved the mystery, we will come with something we already have. Scientific breakthrough yes, but why is the push that we need something to emulate or replace the human? (I guess some will say the answer is obvious, capitalism, and so on and so forth).

But why is the average person excited we will emulate something that is already present through harder more wasteful ways?

"A guy in Silicon Valley promises that if We can just give him enough venture capital and let an AI write our poetry, the polar ice caps will somehow miraculously freeze back together out of sheer respect for our amazing innovation."

Y-- yes? And that's ...good? Like, we collectively, as society around the world, agreed that that's good, right?

It's good that a party can decide such things. "We collectively, as society around the world" is a bit unspecific and doesn't exactly resonate with me in this context. Plurality is key.

All those analogies are obviously imperfect. I've heard many of them, like another one: "you don't need to murder someone to understand that killing is bad" or something like this.
Those analogies are improperly used and lead to a categorical error.
Even if there is a risk of brain rot, brain rot is not the same as cancer, and smoking is not the same as using AI. Also you already have brain rot by watching too much tiktok/YouTube etc... The whole discussion is moot.

ah yeah the wokerati of course… it's always the wokerati.

It all started anyway not from the question whether it should be banned or not,but whether people who don't use it have anything interesting to say about the topic. IMHO not really

A quick reminder. And you call it nice community?

Hm, not sure.

A nice community? Oh boy…

The mysterious inability of some (same) people to understand the obvious fact that (most) things are cumulative never ceases to amaze me. This is the same line of reasoning that gifted us with such brilliant sentences as “stop whining about environmental impact of AI while you write emails and other such unneccessary power consuming activities.” It’s like their logic checking unit shuts off.

Of course. But it’s quite close when you actually stop and think about it. It’s toxic for the user, it’s addictive, it inflicts damage on the environment and people around the user, it’s being pushed onto us by big business, and the toxicity studies are being ignored/supressed.

4gentE: whatever man, you know, I know, they know...why we even discuss it over and over again. Try some AI tools, just for a taste of it, maybe you will actually like it. I know a good dealer ;)

THE IRONY

Why not:

"if you use it in a demoscene prod for a party or anywhere else, you WILL be disowned by everyone, including groups, and will never be wanted anywhere near the scene EVER again"?

Sounds like a much more powerful argument than appealing to emotion like environmental causes/capitalism bad/health risks, and a blatant/clear warning. Sometimes to people, things like politics/environmentalism doesn't speak to people unlike blatant and tangible threats do.

AGI fascination isn’t about copying humans it’s about building systems that can reason, adapt, and discover beyond narrow tasks. If it truly matches or surpasses human-level general intelligence, the metric won’t be hype — it’ll be consistent, autonomous problem-solving across domains without hand-holding. People aren’t excited to “replace humans”; they’re excited about amplifying what humans can do.

thanks and regards
jack
https://www.multispanindia.com

I guess that depends on who are the aforementioned “people”. Scientists? Artists?Average Joe? Investors? TESCREAL groupies?

Why are we talking about AGI in a thread about MML coding assistants? They're completely different problem domains, and nobody is even remotely close to making any sort of progress at all on the AGI front.

This would totally not fly for various reasons, among them the scene's plurality, decentralised nature and historical record of endlessly discussing controversial topics.

And what's this Sippenhaft guilty-by-association "including groups" thing?

I bet some people after seeing how harsh the backlash is here against any AI usage and how orthodox some people are here, will just avoid this whole community as a whole and move somewhere else. If this is what you want, this is exactly what you will get.

You will be left alone and...you will be left behind.

not the vibes I'm getting, no - all the AI assisted diffs my intellect is shat on with in the real world really tell a different story. The bubble of AI users thinks they're moving fast, but all they do is getting reassured by their big number matrix calculator that yes, it's ok to suck at coding and no, you don't have to think hard about how to solve a problem, here you go, a stolen snippet with a subtle bug

I feel it's very appropriate that the AI debate thread has now attracted the ChatGPT spambots.

Kinda irrelevant but I was browsing looking for game jams and saw one that said "vibe coding game jam". It was funny to me as my first thought was, so if I submit something that doesn't use AI at all, am I breaking the rules now? I wonder what happens if we have separate AI gfx competitions and someone submits a non-AI picture there? Would they be disqualified for taking the extra mile? Shame on you for cheating by doing the hard work! It's funny thinking about it :)