We all know this is going to end up being used by people with pr0n habits - and at that resolution they might even see parts of the female anatomy that they didn't know existed.

I don't think training a model with copyrighted data is any kind of gray zone at all.

Running inference against such a model _might_ be infringing.

But I don't see the legal argument that model training would be.

Think of model training as lossy data compression. You provide an input, it produces a compressed artifact, and it can no longer reproduce the original when you uncompress it. It can produce _something_. It cannot reproduce the original.

On slashdot, we all agree that format shifting copyrighted works should be allowed and unrestricted, right?

If there is applicable case law here about how lossy a "reproduction" can be before copyright has something to say about it, that would be relevant. Do you know of any?

I think you're repeating something that feels emotionally valid but in my experience, isn't factually (or legally) accurate.

I have found AI models to be frustratingly _obedient_ of copyright.

I'll be in a situation where I ask a specific question about what a document says, and the AI repeatedly says the work is copyrighted and it can't just tell me what it says _verbatim_, but it can answer specific questions for me in its own words (while telling me to check the original).

Two examples from my life are procedural questions about what Roberts Rules of Order would require in a given situation, and, what ASD STE-100 says about how to name/word things.

The applicable US law would appear to be USC 17 Sec 106 and 107, which on one hand limits derivative works, but on the other, allows fair use for purposes of teaching, scholarship, or research.

(USC references provided by AI, btw. Which I haven't read)

I used to agree with you, and then oddly enough, about 25-30 years ago, I came across this signature on another regular slashdot posters postings:

"The law, in its majestic equality, forbids the rich as well as the poor to sleep under bridges, to beg in the streets, and to steal bread."

(AI tells me the attribution is "The Red Lily", Anatole France, 1894)

Some place between sometimes and often - I'm not here to argue that - researchers are the first to discover.

But it's also almost meaningless to just go, "I'm such an amazing realist, the worst case happens every time."

That's not really being realist. That's just you not knowing actual things, and having an emotion.

I have the opposite, can always get in with IPv6, but IPv4 results in me getting a lot of loops.

I suspect what you're seeing (and what I'm seeing) are both coincidences and reCAPTCHA doesn't care. From what I've read, reCAPTCHA only cares about IP addresses when it suspects you're using a server as an egress point (suggesting a hacked server), and blocking IPv6 would be insane given how few IPv6 servers there are, and what proportion of the consumer Internet uses IPv6 addresses by default.

Because it's marketing.

Generative AI is not unsafe in the way GenAI advocates are claiming. It's unsafe in general - it's causing education deficits, it's causing headaches for companies that have adopted it who no longer have business knowledge over how their own companies operate, it's slamming websites with bogus requests and causing the web to go offline, and when the web isn't offline people are discouraged from visiting websites by "AI summaries" that supposedly answer the questions that you'd normally go to the websites to get answers for. Also the false information given by "AI summaries" has probably caused a problem or two.

But in terms of taking over nuclear missiles and launching them at capital cities, and all the other ludicrous genAI claims made over the last few weeks by employees of OpenAI and Anthropic suggesting they plan to exterminate humanity, no, it's a fucking word predictor on steriods, it just doesn't have that power.

OpenAI and Anthropic are spreading this lie not because anyone who works for those companies believes it, but because a real AI, the type of thing that people are expecting something called "AI" to actually be, would be capable of reasoning that humans are the enemy (etc), and might go Skynet on humanity. If they say "Nah, you're all fine, what we have here is spicy autocomplete", then their potential for an IPO goes down the toilet.

These are dishonest companies that are doing everything that they can to pump up a tool with marginal utility in order to run off with the money before it becomes revealed that, no, their tech is unsustainable and was never as good as the cultists claimed anyway.

That's why politicians are talking about kill switches. They're being duped by Altman and friends into thinking that's necessary.

> Any programming language needs the equivalent of Rust's "unsafe"

Not really, no.

> Without it would be impossible to do any input or output.

Have you ever used any programming languages other than C, C++, etc? The important thing is that the language itself implements an interface between the OS and programs written in that language. In addition to that, most usually have an ABI that can be used to glue assembler libraries to the language - this is not the equivalent of "unsafe", because it's not part of the language, and a person downloading an application can see right away that it includes external libraries that interface with it.

Unsafe is in Rust because... well, Rust is designed to try to make memory safety attractive to devs who traditionally don't like being told what to do or that their favorite language sucks. Which is why, for example, advocates go to great lengths to claim Rust doesn't have garbage collection when, actually, it does, it's just not mark and sweep. Rust is as much about public relations and trying to get people to stop assuming they have to sacrifice safety for efficiency and low level access as it is about producing a useful programming languages.

And, unfortunately, part of that PR is to say "Look, we even have an "unsafe" keyword so you can get rid of (some of) our restrictions on you for those rare times you actually need it", and the group Rust is aimed at, of course, go "Oh, I can do this with a 2% speed up if I use unsafe, so I'm going to do the unsafe version, that's what makes Rust great" *insert Naked Gun facepalming scene here*.

I like Rust. But let's not pretend "unsafe" is a positive thing, or that most languages need it. Rust needs it because of its intended audience, but plenty of languages don't let the code itself be unsafe. Some have an unsafe keyword, some allow linking with unsafe assembler packages, and some... honest... don't have either, and your code is only unsafe if your compiler/interpreter is.

Professor, without knowing precisely what the danger is, would you say it's time for our viewers to crack each other's heads open and feast on the goo inside?

Yes I would, Kent.

> Broadcast networks have to give the different major parties equal time.

The rule that this applies to late night talk shows is literally months old and has no basis in any new legislation. It was introduced by the FCC on a whim basically because they didn't want Colbert to interview the same person as Kimmel.

Under exceptions the same Wikipedia article, correctly, points out the current administration changed these rules to apply to... late night talk shows:

If the airing was within a documentary, bona fide news interview, scheduled newscast, or an on-the-spot news event, the equal-time rule does not apply. Late-night and daytime talk show interviews were long considered to be bona fide news segments until the FCC issued new guidance in 2026 signaling that these types of shows would no longer be automatically granted the bona fide news exemption.[3][4]

This is a new interpretation of the equal time rule designed to make it difficult for late night shows to interview politicians, and if it wasn't for SCOTUS being nakedly partisan these days, it wouldn't muster a single court's scrutiny.

There is no equal time rule that applies to late night talk shows. This came up with the Colbert thing last year when Paramount similarly caved into FCC threats about it. This is a rule the FCC have arbitrarily decided to re-interpret despite decades of precedent.

You're still a moron. Fascism is not "Anything the government does that's bad", it's a system of corruption and abuse of power.

You may not agree with the vaccine mandate (because you're probably a moron as we've just established) but a vaccine mandate is not an abuse of power, literally the only incentive to do it is for the public good. It's not being done to enrich those in power, it's not being done to damage opponents, or to scapegoat powerless groups. It's not even being done to cement the power of those in power outside of persuading voters that they're competent. It's being done to prevent the spread of a disease.

That's text book "not fascism at all". The only thing it has in common with fascism is that "(whines) the government did something! It did something I tell you!" which is a pretty fucking stupid definition of fascism, and is the same definition a lot of Americans think applies to socialism, which... I mean, you fuckers have never heard of trade unions, the cooperative movement, independent private voluntary communes, or Robert Owen? Really?

If the government does something, and it's for the public good, rather than to cause unnecessary harm to the powerless or opponents or to steal from people, it's not fascism.

Insignia sells dumb TV, at least when I had to get a TV for my 80 year old in-laws a couple of years ago. (The whole "You must have a smart TV" thing is so stupid even when privacy issues and nerds aren't taken into account. It's like mandatory smartphones, or mandatory car driving.)

I can't say I was impressed by the quality however.

You think this is limited to $199 TVs? The full report (yeah, I watched it) focused mainly on TVs over $1,000.

Not that a TV being $199 is an excuse for spying on people. That's bootlicking bullshit, and you should be ashamed for even hinting that it's fine because of the price of the TV.