Probably not yet, but I wouldn't assume they aren't or won't be, and they certainly are being scanned for CP.

Yeah, that's the question, but since they're already scanning for CSAM (which I'm not against, BTW) there's ample precedent for scanning your data for unlawful content.

You should never, ever trust any data storage provider to keep your data secret, or to not snoop through it. That's not a thing. They are snooping through data all the time, if yours doesn't get looked at, it's only by luck.

Every chance to get another story about how AI is so powerful it must be regulated serves at least two purposes. First, it is an advertisement for the power of the technology. Second, it is a plea to allow the entrenched AI companies to write regulation to permit them to control the industry. These are not extremely complicated ideas, though they do require that you not cede responsibility for thinking to software which is incapable of doing it because the tokens don't mean anything to it. It has the same ability to comprehend the ramifications of killing a process and killing a person, none. It can produce you a report about the difference and still not have anything like understanding. And the less you use yours, the more unable you're going to become to do so.

The immediate problem is that the 1% can't sustain the 1% for more than a generation or two tops before everything breaks down and goes to shit and they need the people they exterminated.

The longer-term problem would be inbreeding. Last I read you need a population of around 10,000 minimum to have enough people to breed with without having to reinforce dangerous recessives, and that's with care and planning. And the means of long distance travel will break down faster than most other stuff, and they won't be able to fix that either.

There's really no model in which the wealthy exterminate the masses which doesn't lead to the end of the species, with modern technology. Maybe if you had nanotech general assemblers.

The same person has the same mentality and will do both things when given the opportunity. They don't care what you want. They will do it to you because they want to and they can.

Correct, not without a mandate. But when the government above says now it's time to do the thing, the thing happens. And that's why this is a problem that must be solved at a higher level. For too long we have allowed uneven standards in policing and its administration, from state to state, county to county, city to city. We must institute strong laws which rein in police to protect citizens, and this must include a national registry which police departments are required to use during the hiring process so that they are at least informed about which officers abused their position. Nobody should ever have the excuse that they didn't know, what fucking year is it?

Police unions will fight this to the very end, so we should fight those unions over it until they bend or fail. I'm generally pro-union but we shouldn't pretend that they are all good, or that everything they do is positive, or that we don't have a right as a populace to oppose them when they are doing things which harm us.

Perhaps you should be realizing by now that homework is no longer a viable way to test how well students actually know the subject matter, and the only viable way to do so is in-class tests/exercises - frequent enough and worth enough of the grade to force them to study between lessons.

OR, alternatively, you could just keep doing what you're doing, and instead, in effect grade people on:

Comparably poor grades:
* People who actually do the work by hand
* "AI N00bs" who let stuff like that slip through

Comparably good grades:
* AI-experienced people who know enough to skim over the work, do multiple passes, or issue prompts that prevent such tells from getting into the work

Wherein, in effect, your net reward/punishment is inverted vs. what you should want it to be.

Closing one's eyes and ears to the fact that the environment students operate in has changed isn't an acceptable option. Education must adapt. And watermarks, BTW, will only catch "the n00bs"; more experienced / up to date students will keep track of what watermarks and what doesn't, and use AIs that don't watermark, or de-watermarking services.

Austerity has never worked because of velocity of money, you are literally preventing the creation of wealth when employing it. It's the same when applied to Greece by the EU. They've been happy to suck wealth out of Greece, require that it meet standards they set while some of their own members don't meet their own standards (including Germany which has been the strongest voice for austerity) while also doing nothing to help them meet targets, like increasing imports. Come back to the USA and compare to blue states subsidizing social programs in red states while also having people leave those states for social reasons (i.e. if you're not heteronormative and white then you're othered) and come to our states where we have to support them, too. Then they tell us we shouldn't spend money on things like HSR, well guess what motherfuckers, we need a way to move all these people around that doesn't create more problems than it solves.

It's all greed-driven, it's all victim-blaming, it's all the rich protecting their privilege while we dick around with culture wars instead of getting out the guillotines.

I'll repeat: watermarks are anti-marketing. They discourage use.

I don't think that makes the document any less your own than having a human copy editor conduct a similar review which has been done since forever.

Which is what I've been saying as well. If all you're doing is asking for suggestions to something you have already written, using AI is no different than using a human editor. It's your work. It's been written. You're just looking for advice on how to improve it.

This is completely different than people who tell AI what they want and it does all the work.

SynthID in text genuinely is invisible. The probability shifts are small, and all the pathways are valid pathways to express the same thought.

And no, you cannot "collide" text watermarks. At best, it'll only have the latter one's watermark. At worst, both.

It'll still get tagged by SynthID.

You have to use a non-watermarking LLM. Saying "Don't do anything that will watermark it" doesn't help.

All of this crap is in a database, it's a relatively simple matter of writing and running a report. I used to do that for a living, it's not as hard as people imagine it to be.

And ruining education.

If they would stop acting like psychopaths then we would stop assuming they are psychopaths.

Watermarks are "marketing"?

What user wants watermarks?