Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I don't want anything your AI generates

coryd.dev

Post reply on HN
Print view View original

271–280 of 345 posts

Re: I don't want anything your AI generates

#271

Post by bee_rider »

MacsHeadroom wrote: post #267
tomn wrote: post #182

Earlier quoted context omitted.

The expensive bit is not the electricity, it's the cost of the GPUs, so they will train flat out. There is also a massive amount of competition between the various players, and nobody taking part cares about the energy use (otherwise they wouldn't be doing it), so i don't see this happening. (rough numbers: a H100 is 700W, which is $613/year at $0.1/kWh, and the estimated cost is $25000-40000)

Typical datacenter power cost is closer to $0.04/kWh. Datacenters mostly use renewable clean energy as well. Environmental concerns are deeply misplaced. 10 minutes of compute in a datacenter is far more environmentally friendly than 10 hours of work by a human.

Nice, we can just run the ML model—how do you put humans in suspend mode? Or do we just shut them off for now?

Re: I don't want anything your AI generates

#273

Post by jayrot »

ToucanLoucan wrote: post #234
handoflixue wrote: post #221

Earlier quoted context omitted.

> Show me a human being that, in an afternoon, can study the art of Rembrandt and from that experience, paint plausibly Rembrandt style paintings in a few minutes each, and I'll swear by AI for the rest of my life. So it's okay to learn, but only if you do it very slowly? I surely don't need to point you to the existence of forgers - you know a human can study the art of Rembrandt and paint plausibly Rembrandt style…

> So it's okay to learn, but only if you do it very slowly? No, it's a fundamentally different process with different results. An artist learns from previous artists to express things they themselves want to express. An AI digests art to become a viable(ish) tool for people who want to express themselves, as long as that expression resides somewhere in the weighted averages of the art the model has digested. Two fund…

May I just say, as a third-party simply reading this back and forth from the outside, that the tone of your writing and the implied attitude with which you are engaging in "debate", reads as very aggressive and uninterested in actually having a sincere discussion. To me at least.

I imagine you probably won't like this comment, but perhaps you might use it as an opportunity for reflection and self-awareness. If your interest is actually to potentially change someone's mind, and not just "be right", you might consider approaching it in a different way so that your tone doesn't get in the way of the substance of arguments you wish to make.

Just a suggestion. Take care.

Re: I don't want anything your AI generates

#274

Post by jayrot »

SLHamlet wrote: post #61

Anytime someone sends me something generated by ChatGPT, I think about how AI expert Hilary Mason puts it: "By design, ChatGPT aspires to be the most mediocre web content you can imagine." https://nwn.blogs.com/nwn/2023/03/chatgpt-explained-hilary-m...

Benedict Evans often invokes the analogy of the concept of "infinite interns", which is pretty apt, at least in the current state.

Re: I don't want anything your AI generates

#275

Post by jayrot »

nicbou wrote: post #201
SLHamlet wrote: post #61

Anytime someone sends me something generated by ChatGPT, I think about how AI expert Hilary Mason puts it: "By design, ChatGPT aspires to be the most mediocre web content you can imagine." https://nwn.blogs.com/nwn/2023/03/chatgpt-explained-hilary-m...

I sometimes need this. It’s nice to get an average of all mediocre content summed up as a bullet list sometimes. Sometimes I want the average tourist guide, the average recipe or the average answer. Now I get it instantly without sifting through ad-infested shallow content. ChatGPT made it so much easier to be curious about new things. It’s a good trailhead for curiosity.

Point of note: this is all well and good until these tools become inevitably ad-infested.

I'm starting to think we need to add "advertising" to the list (Death, Taxes, etc).

Re: I don't want anything your AI generates

#276

Post by arghwhat »

arghwhat wrote: post #245
adrian_b wrote: post #122

Earlier quoted context omitted.

All human output is much more derivative than the majority of "creators" dare to admit. Nevertheless, good human output always add something new and original to the elements that are derived from prior art. AI output consists entirely of derived elements, which only in the best case may happen to be mixed into a distinct combination from those already existing. Even when the combination is new, it is distinct only du…

> Nevertheless, good human output always add something new and original to the elements that are derived from prior art. My personal opinion is that even the new and original elements are derivative. Once a creation derives from a sufficiently large number of sources - some unrelated, such as being inspired by music when painting - you consider it original and new, and once the space of experiences and current inputs…

*couldn't do this as well or better than us, of course.

Re: I don't want anything your AI generates

#277

Post by ToucanLoucan »

jayrot wrote: post #273
ToucanLoucan wrote: post #234

Earlier quoted context omitted.

> So it's okay to learn, but only if you do it very slowly? No, it's a fundamentally different process with different results. An artist learns from previous artists to express things they themselves want to express. An AI digests art to become a viable(ish) tool for people who want to express themselves, as long as that expression resides somewhere in the weighted averages of the art the model has digested. Two fund…

May I just say, as a third-party simply reading this back and forth from the outside, that the tone of your writing and the implied attitude with which you are engaging in "debate", reads as very aggressive and uninterested in actually having a sincere discussion. To me at least. I imagine you probably won't like this comment, but perhaps you might use it as an opportunity for reflection and self-awareness. If your i…

You aren't wrong in the slightest, apart from that you've gotten the implication that I'm here for a debate. I'm not. I've been having this debate since the StableDiffusion blow up at the mid-ish of 2023. I've read these points restated by countless pro-AI people and refuted them probably dozens of times at this point, here and elsewhere, always ending in a similar deadlock where they just stop replying, either because they're sick of me, or I've "won" for whatever that means in the context of online discussion.

Nevertheless I'm always open to be persuaded by actual arguments, and I have on numerous issues, but I have yet to see any convincing refutations on these points I've outlined here regarding primarily, but not limited to:

- The unethical sourcing of training data

- The exploitation of lesser-privileged workers in managing it

- The harm being done and the harm that will be done to various professions if they become standard

And not mentioned in this thread:

- These various firms' superposition between potentially profitable business and "research initiatives," depending if they're trying to get investment or abuse the public square respectively

- The exploitative/disgusting/disinformative things these AI's are being used to produce to a society already saturated with false information and faked imagery

But these discussions usually dead end, like I said, when the other person stops answering or invokes the "well if we don't build it someone will" which is also unpersuasive.

Relating specifically to your point about wanting to change someones mind: in my first comment I do feel I put out an olive branch with empathy for being excited about a new thing. But when the new thing in question is so saturated from beginning to end in questionable ethics... I'm sorry there's only so much empathy I can extend. If you (not you specifically, but you as the theoretical person) are the kind of person ready to associate with this technology at this stage, when it's foibles and highly dubious origins are so well known, then I'm not overly interested in assuaging your feelings. This person came into this thread bemoaning the fact that so many people are calling them out on this and they're sick of it, and like, there's a great way to stop that happening: stop using the damn technology.

I will always extend empathy, but if your position is whining about people rightfully, IMO, pointing out that you are using unethical tech and you wish they'd stop? Like, sorry not sorry man, maybe you shouldn't use it then. Then you get less yelled at and a clear conscience. Win/win.

But I do appreciate the reply all the same, to be clear. You aren't wrong. I've just had this argument too much, but also don't feel I can really stop.

Re: I don't want anything your AI generates

#278

Post by firebat45 »

LegitShady wrote: post #261
handoflixue wrote: post #218

Earlier quoted context omitted.

Human artists have been training on the same content for decades and no one seemed to complain. You can argue that machines should be held to a different set of legal and ethical standards, but it's certainly not obvious. Most factories are designed based on vast amounts of prior manual labor, so it's not like "automating a manual process based on analyzing existing methods" is new, either. Why is it okay to automate…

AI are not human artists, so there's no connection to your point and the discussion.

You could just as easily claim that since AI are not human artists that copyright does not apply to them.

Re: I don't want anything your AI generates

#279

Post by xigoi »

stolenmerch wrote: post #52

I'm not a fan of this hyper aggressive line-in-the-sand argumentation about AI that pushes it all precariously close to culture war shenanigans. If you don't like a new technology that is perfectly cool and your right to an opinion. Please don't position it so that if I want to use AI I have to defend myself from accusations of exploiting labor and the environment. That is NOT at all clear, settled, or even correct m…

> I'm not a fan of this hyper aggressive line-in-the-sand argumentation about fossil fuels that pushes it all precariously close to culture war shenanigans. If you don't like a new technology that is perfectly cool and your right to an opinion. Please don't position it so that if I want to use fossil fuels I have to defend myself from accusations of polluting the air and the environment.

Re: I don't want anything your AI generates

#280

Post by philipkglass »

logicprog wrote: post #232
handoflixue wrote: post #217

Earlier quoted context omitted.

> disproportionately damaging to the environment compared to the alternatives This is a new one to me. Do you have any source for that? Once a model is trained, it seems pretty obvious that it takes Dall-E vastly less to create an image than a trained artist. I have trouble believing the training costs are really so large as to change the equation back to favoring humans.

I'm not sure your premise even makes any sense here, because it doesn't take an artist much more resources to produce art then it took them to just exist for the same amount of time. They're still just eating, sleeping, making basic usage of the computer, using heating and light, and so on either way. Whereas someone using dall-e is doing all of that plus relying on the immense training costs of the artificial intell…

This back of the napkin estimate for GPT-4 emissions costs is too high by orders of magnitude. Your estimate is that training it emitted about as much as CO2 as 5.38 billion average humans living their lives for a year did. With a world population of 8 billion, it would mean that GPT-4 training was equivalent to 0.67 years of total anthropogenic CO2 emissions. Since GPT-4 CO2 emissions all come from manufacturing hardware with fossil fuels or burning fossil fuels for electricity, this is roughly equivalent to 0.67 years of global fossil fuel production.

But OpenAI had neither the money nor the physical footprint to consume 0.67 years' worth of global fossil fuel production! At those gargantuan numbers OpenAI would have consumed more energy than the rest of the world combined while training GPT-4. It would have spent trillions of dollars on training. It would have had to build more data centers than previously existed in the entire world just to soak up that much electricity with GPUs.

Post reply on HN