They aren’t a technical bug, but an UX bug. Or would you claim that an LLM that outputs 100% non-factual hallucinations and no factual information at all is just as desirable as one that doesn’t do that?

Btw, LLMs don’t have any traditional code at all.

Yes, your fan art infringed on Blizzards copyright. Blizzard lets it slide, because there’s nothing to gain from it apart from a massive PR desaster.

Now if you sold your Arthas images on a large enough scale then Blizzard will clearly come after you. Copyright is not only about the damages occured by people not buying Blizzards stuff, but also the license fees they didn’t get from you.

That’s the real big difference: if Midjourney was a little hobby project of some guy in his basement that never saw the the light of day, there wouldn’t be a problem. But Midjourney is a for-profit tool with the express purpose of allowing people to make images without paying an artist and the way it does that is by using copyrighted works to do so.

It’s not anthropomorphizing, its how new terms are created.

Pretty much every new term ever draws on already existing terms.

A car is called car, because that term was first used for streetcars before that, and for passenger train cars before that, and before that it was used for cargo train cars and before that it was used for a charriot and originally it was used for a two-wheeled Celtic war chariot. Not a lot of modern cars have two wheels and a horse.

A plane is called a plane, because it’s short for airplane, which derives from aeroplane, which means the wing of an airplane and that term first denoted the shell casings of a beetle’s wings. And not a lot of modern planes are actually made of beetle wing shell casings.

You can do the same for almost all modern terms. Every term derives from a term that denotes something similar, often in another domain.

Same with AI hallucinations. Nobody with half an education would think that the cause, effect and expression of AI hallucinations is the same as for humans. OpenAI doesn’t feed ChatGTP hallucinogenics. It’s just a technical term that means something vaguely related to what the term originally meant for humans, same as “plane” and “beetle wing shell casing”.

Nope. Hallucinations are not a cool thing. They are a bug, not a feature. The term itself is also far from cool or positive. Or would you think it’s cool if humans have hallucinations?

Hallucinations mean something specific in the context of AI. It’s a technical term, same as “putting an app into a sandbox” doesn’t literally mean that you pour sand into your phone.

Human hallucinations and AI hallucinations are very different concepts caused by very different things.

You did not read your source. Some quotes you apparently missed:

Scraping to violate the public’s privacy is bad, actually.

Scraping to alienate creative workers’ labor is bad, actually.

Please read your source before posting it and claiming it says something it doesn’t actually say.

Now why does Doctrow distinguish between good scraping and bad scraping, and even between good LLM training and bad LLM training in his post?

Because the good applications are actually covered by fair use while the bad parts aren’t.

Because fair use isn’t actually about what is done (scraping, LLM training, …) but about who does it (researchers, non-profit vs. companies, for-profit) and for what purpose (research, critique, teaching, news reporting vs. making a profit by putting original copyright owners out of work).

That’s the whole point of fair use. It’s even in the name. It’s about the use, and the use needs to be fair. It’s not called “Allowed techniques, don’t care if it’s fair”.

Tbh, this is not a question about scraping at all.

Scraping is just a rather neutral tool that can be used for all sorts of purposes, legal and illegal.

Neither does the technique justify the purpose nor does outlawing the technique fix the actual problem.

Fair use only applies for a certain set of use cases and has a strict set of restrictions applied to it.

The permitted use cases are: “criticism, comment, news reporting, teaching (including multiple copies for classroom use), scholarship, or research”.

And the two relevant restrictions are:

• “the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes;”
• “the effect of the use upon the potential market for or value of the copyrighted work.”

(Quoted from 17 U.S.C. § 107)

And here the differences between archive.org and AI become obvious. While archive.org can be abused as some kind of file sharing system or to circumvent paywalls or ads, its intended purpose is for research, and it’s firmly non-profit and doesn’t compete with copyright holders.

AI, on the other hand, is almost always commercial, and its main purpose is to replace human labour, specifically of the copyright owners. It might not be an actual problem for Disney’s bottom line, but it’s a massive problem for smaller artists, stock photographers, translators, and many other professions.

That way, it clearly doesn’t apply to the use cases for fair use while violating the restrictions.

And for that, it doesn’t matter if the training data is acquired using scraping (without permission) or some other way (without permission to use it for AI training).

Btw, would it be legal to use a torrent client that uses an LLM to make up the outgoing packets so that you aren’t sending copyrighted material? ;)

Again someone who thinks that public policies are natural laws…

NASA could do and did do what SpaceX is doing now, but they are beholden to the government and if the government says “we don’t do that for ideologigal reasons” then it doesn’t matter what can be done.

Minor crimes like that rarely carry prison sentence, so yeah, wouldn’t really matter.

So escaped naked to avoid stealing prison clothes.

Thanks for the summary! That sounds freaky!

Well, the trade-off between trusting a huge corporation or a single dude on the internet.

What exactly happened there? It was the big thing, then I didn’t use it for a month or so and then it was gone.

Even if you make them in large quantities, material cost alone will be at least €50k. You will need a skilled operator nearby, and constant maintainance, and if you lose even one per year, a regular underpaid human worker will be much cheaper.

These things are pure marketing devices to pacify investors, generate headlines and make unions and workers afraid.

Because it’s not real. It’s purely for marketing, not for actual wide-spread implementation.

Even in the best of cases, even factoring in economy of scale and all that, a robot like that will cost upwards of €50k at least, probably closer to double that, will require constant maintainance, and the risk of vandalism or accidental damage is really high. And you’ll likely need a (skilled) human operator nearby anyway, because the delivery vehicle doesn’t drive itself.

The purpose of projects like this is marketing and public perception.

• The company looks futuristic and future proof. That’s good to get investors.
• The company looks like they could replace humans with robots at any time. That’s good with negotiations with unions and workers.
• The company gets into headlines worldwide. That’s advertisement they don’t have to pay for.

This robot is not meant to ever go mainstream. Maybe there will be a handful of routes where they will be implemented for marketing purposes, but like drone delivery and similar gimmicks, it won’t beat a criminally underpaid delivery human on price, and that’s the only metric that counts for a company like Amazon.

“Prescription glasses” only mean “glasses with optical properties”, so glasses that actually do anything with focus, as opposed to e.g. non-prescription sunglasses or non-prescription accessory glasses that people wear to look smart or something.

It doesn’t mean you need a prescription for them.

(That said: in some countries you need a prescription for your prescription glasses if you want your health insurance to pay for them.)

90% of the things that Japan introduced according to comment sections on the internet never happened (or never made it past the prototype stage) and the rest was actually introduced in Korea, not in Japan.

The Japanophilia is strong with a lot of people on the internet.

There’s this idea I’ve been considering for a long time.

Imagine putting a remote controlled firework smoke bomb under the tailpipe, hidden from sight. At best a really stinky one that smells like burned rubber or something.

When someone follows to closely, just fake an engine issue or something by activating the smoke bomb and fill their AC air intake with the smell of burned rubber for weeks. Just to teach them to not follow too closely again.

Yeah, especially in peace time. When war heats up and resources get scarce, you use the cheapest thing that does the job. But in peace time you feed your military contractors to keep them happy and to keep them researching and developing so you don’t lose out on modern technology development.

(For clarification, with “war time” I mean “being in a war that actually threatens the country”. The US hasn’t been in a war like that for a very long time. They’ve essentially being in “peace time” while having military training and testing facilities in the middle east.

10 years ago I got into RC planes for a summer, and me and the guy were talking about how ridiculous it is that the milirary is spending so much money on simple drones, when they could just strap some explosives on a cheap hobbyist RC plane/drone for a fraction of the price, and just create swarms of them.

The technology had been widely available for some time already back then. Turns out, it was just lacking a war to do so.

(Just to be clear, we were all anti-war in general, this was just idle speculatiok back then. But if our country was attacked at that time, I’m sure some of us would have ended in a newly created drone force like what happened in the Ukraine.)