I’d really like to see a source for the first guy’s numbers, especially how he accounts for things like training and water usage at the power plant, and as for the second guy, Anthropic is not preserving shit. The Internet Archive preserves books. Anthropic scans them and doesn’t publish the scans so that they can train an LLM that might or might not be able to regurgitate some fragments of that text, and publish that. That’s preservation in the same way that painting a picture of you is keeping you alive forever.

Also, neither of those address the effects on creatives’ livelihoods or the mental health of LLM users. Chatbot psychosis is real. People who routinely use LLMs to do things provably become worse at doing them without them. Students use LLMs to make an end run around having to learn everything they need to know to be effective members of a society, like how to articulate their points, how not to fall for rhetorical traps, what history was really like, and between that and the disastrous effects of No Child Left Behind, teachers are quitting in droves and there’s a literacy crisis.

The author of that article admits that the sample size is not large enough to draw meaningful conclusions.

But besides that, I believe LLM code generators can be a useful tool, provided you are willing to go over their output with a fine-tooth comb and assume it is broken until you have proven otherwise, because the hallucination problem is inherent to the technology and they’re never going to completely solve it, and are willing to overlook the myriad ethical issues with all major LLMs in existence today.

Yeah. Even if you run the model entirely on your own 100% solar-powered hardware, there’s still the problem of training data, and there’s still the problem of the myriad adverse effects on the user.

Okay, but they still happen with orders of magnitude less frequency than bugs in AI code. Consider that the time between when rsync first adopted LLM-generated code and users en masse reporting rsync internal protocol errors during a backup was on the order of months.

The person who put “AI is evil” in big block letters is objectively correct. See homes getting claimed under eminent domain to build power lines for datacenters, electricity prices for consumers rising as service quality gets worse, and water usage to the tune of millions of gallons per datacenter per day. See companies buying rare books in bulk, cutting them apart so they can be scanned faster, and not publishing the scans, literally destroying culture so they can have one more scrap of training data guaranteed to be free of slop that their competitors don’t have. See companies operating web scrapers that deliberately evade every block that website owners try to use to stop them from reading their sites, and then scraping so many different pages so frequently that it’s indistinguishable from a DDoS attack, hosting bills skyrocket, servers crash, indie webmasters pay the price for corporate greed. See xAI operating natural gas turbines on the property of their datacenter for extra electricity in flagrant violation of the Clean Air Act, spewing toxic smoke into the air and generating powerful, inaudible infrasound vibrations that have been linked to everything from nausea to irritability to hearing loss to seizures. See Anthropic treating Minority Report as an instruction manual. See how ChatGPT use affects test scores. See AI psychosis. See chatbot-induced suicides. See deskilling. See how little AI companies care.

I’ve been a programmer for two thirds of my life up to this point. I don’t see how the machine that generates code that must be assumed to be broken, and statements about said code that must be assumed to be wrong, until both have been manually verified, because the “hallucination” issue is inherent to the design of LLMs and cannot be solved, is supposed to make my life easier. Especially since a well known problem in computer science is that writing code is more fun than reading it.

If you’re using AI for code generation, I hope you like reading code.

If you’re using AI to learn your way around an existing codebase, I hope you’re prepared to fact check everything it tells you.

If you’re using AI for code review before merging a PR, fucking yikes. What made you think the sycophant machine that once told users a soggy cereal bar was a great business opportunity and cannot even transcribe McDonald’s orders correctly was suitable for that purpose?

If you’re telling yourself it’s okay because there’s a human in the loop, please read some studies about decision fatigue and consider that increasing the number of critical decisions a person must make per day, while giving them a suggested answer that’s right 90% of the time and catastrophically wrong the remaining 10%, and the only way to tell the difference is to read in depth, might not have the desired effect. Especially if the human in question is then put under time pressure.

And all that is before I go on a five paragraph rant about the ethical transgressions of every major AI company.

It’s pretty black and white for me.

…So you insist that AI power use isn’t a big problem right now, but your best case scenario for the future is that humanity gets nuked out of existence. By nukes that were designed using help from LLMs, which they haven’t been shown to be able to help with.

You seem like a very serious person who is worth debating.

As for the data use thing, firstly, I don’t know why the choices are between the actual work and hearsay. I don’t know where you got those two options from. I also don’t know what would be producing the hearsay (human? LLM? If the latter, are we straight up admitting to copyright laundering? Also, I thought training LLMs on output of previous LLMs caused model collapse). I especially don’t know why you’re asking me to speculate on what creatives want immediately after attacking my ability to speak on behalf of creatives. But since you are apparently asking me to do that, I think they’d prefer the hearsay, since the point is that they don’t want the model to be able to imitate their work for anyone who asks for cheap, and in so doing, put them out of a job. Or worse, have the model agree to a job they would have said no to, and damage their credibility and brand.

Okay. I’m not sure how that excerpt proves he’s an industry plant. It sounds like he’s talking about the massive financial crater that will be left behind when the fact that zero companies in the AI sector that do not manufacture hardware are profitable catches up with them and the bubble pops, which I personally think he’s right about.

I don’t know about other industries, but I’m a programmer, and I firmly believe that people who cannot be bothered to put in the hours to learn to do so the old fashioned way, should not make software.

Vibe coding is how you get software designed by people who do not know how to design software and implemented by a chatbot that is infamous for writing buggy, insecure, impossible-to-maintain code.

Frontier level models are finding novel solutions to decades-long unsolved maths problems.

Such as?

Because if you’re talking about that recent proof of the Collatz conjecture, I’ve got bad news.

Also let me know when they figure out how to stop making stuff up, or at least to understand the difference between a sequence of tokens that represents a true fact and one that does not.

This is cool, but it is a completely different kind of AI from LLMs.

Do a little research. AI power usage isn’t that much of a problem yet.

Okay.

https://www.yahoo.com/news/us/articles/pollution-musk-unpermitted-xai-power-100200381.html

https://www.msn.com/en-us/money/markets/epa-exempts-off-grid-ai-power-plants-from-acid-rain-program-caps-affecting-60-sites/ar-AA28YctM

https://www.fastcompany.com/91229319/ai-data-centers-could-make-your-electric-bill-go-up-by-70

https://www.cbsnews.com/news/georgia-power-ai-data-centers-eminent-domain/

Power usage from AI data centers has already eclipsed power usage from all other data centers combined. If you can come up with a scenario where AI hasn’t doubled in scale by 2030 by belching smoke into the atmosphere, I’ll concede the point.

A lot of really good data is copyrighted

Yeah, and that’s too fucking bad. If the people who made it don’t want it to be used for AI training (and I don’t think I’ve ever met a creator who has done the research on AI who does, even if they got paid), it shouldn’t be.

Here’s a link to the exact same post on Cory Doctorow’s (tracker-free) blog: https://pluralistic.net/2026/08/01/dare-snot/#i-will-fucking-piledrive-you-if-you-mention-ai-again

Pro tip: when copying a link to a Lemmy/Piefed/Kbin/etc post, please make sure you use your instance’s option to copy the link to the post on its original instance (the instance the post was made on), rather than simply copying the link to the post on your instance out of your URL bar. On your instance’s web interface, that option is in the More drop down next to the Boost button, just below the post title. (On a Lemmy instance, the link to the original post is the Fediverse icon right next to the upvote and downvote buttons, and you can right click and Copy Link from there).

Obviously, when you open a post in the browser on an instance that isn’t yours by clicking a link (for example, this link to this same post on my instance, which is a vanilla Lemmy instance), you can’t interact with it, because you aren’t logged in, because you don’t have an account on that instance. To interact with it, you need to open the same post through your own instance.

If you have a link to a post on the instance it was created on, you can simply paste that URL (the whole thing) into your instance’s search bar to open it on there, so you can vote/comment/etc. Some mobile apps can even do this automatically. However, this doesn’t work if you paste a link to a post through an instance other than the instance it was created on. In order to interact with the post you just linked to, people have to click the link to open the post on your instance, then click the button on there to copy the original link, before pasting it into their own instance. It’s polite not to make them go through that extra work.

Judging by your account age, it seems you’re new to Lemmy, so you probably didn’t know that. Just something to keep in mind for future. :)

why does your dad look the same age as his mom

i can at least take solace in the fact that brain dead lemmings are 100% bark and 0% bite

I can’t see through the paywall either, but if it’s like any of the hundreds of other articles that have been written on this topic, the answer is probably TikTok

can somebody please remind me why i still have a lemmy account?

by that logic, every citizen of the united states is a fascist, because they could have been mobilizing and weren’t

Elon Musk tweeted “Imagine a new all-powerful woke AI”

i’m honestly not sure which belief is more brain-dead: that anything that is woke is bad, or that all-powerful AI is a credible threat