uriel238

Curiously, when I was drafted into jury duty here in California, we were shown a video that said something like The people are society's last line of defense against bad law.

I took it personally, figuring it was my duty to not enforce any law that was unjust or protect any institution that was cruel. But I didn't get selected.

Though this happened with the grand jury, which only needs a majority vote (sometimes a supermajority). It means a lot of people chose to nullify or didn't trust the prosecution to give a fair trial.

The Department of Justice has been doing stuff like that a lot since the Trump administration, to the point that they lost presumption of regularity (the assumption in advance that state actors are being honest and following the rules in good faith). There were so many prosecutions that were clearly directed by the White House and were vindictive persecutions of political enemies that the judges just stopped believing anyone from DoJ.

It didn't help that all the honest prosecutors in DoJ resigned when they were ordered to do something unlawful, resulting in massive brain drain. All the rehires are unskilled if not incompetent, and sometimes they issue legal statements that sound like a Truth Social post.

The anti-piracy efforts of the RIAA and MPAA (and the publishing houses and...) are still pretty robust, but that didn't stop any of the big AI companies from using gigatons of copyrighted material as datasets to train their LLMs. It's why when you ask them to generate an image featuring Winnie The Pooh, they all know what you're talking about.

But the big companies absolutely did not get permission to do this. Nor did anyone give permission to allow for the recent jailbreaks by AI task systems to hack into other companies. If that were any singular human (not of the owner class) they'd face charges under the CFAA. But since it's a massive company with an army of blue-haired lawyers, nothing was done about it, and the companies actually bragged about their AIs escaping containment.

Extinction by AI takeover is far more interesting than extinction by global drought.

Once again Hallowe'en is getting back to its roots.

PS: I wrote a thing in 2024. Sorry, I don't know how to do intra-lemmy links.

It depends on the judge. Click-wrapped ToS, EULAs and privacy policies are a known dark pattern:
They're longer than Shakespearean plays.
They're difficult to read even at a college level.
They're often presented in a tiny window with difficult controls.
They're presented suddenly and mid process, often as part of an already lengthy string of steps. Often the user is eager to just move on.

Some judges have recognized this and do not see click-wrapped contracts as binding, especially when the person clicking them can be a child, or a cat, or a computer glitch.

But then a lot of judges were appointed through a willful organized effort to put pro-business judges on the bench, and will rule in the favor of companies over individuals regardless of the circumstances.

Not at all! China isn't trying to do the same buildout. In fact, they're turning towards the AI-as-software model, in which AI is run locally and sold as software rather than being sold as compute from large supercomputers.

It turns out that most of the things common folk and common businesses do with AI are managed pretty well with open source AI software run locally with a dozen terabytes of training data.

And if the US companies sold tokens at their actual rate (not at a loss) then it would be cheaper for even small businesses to put together a high-powered PC and set it up to do AI tasks. So for everything that isn't the stuff I listed, the AI-as-service model is already obsolete.

I betcha a nickle that the fines paid to the canteen would be more than offset by the increase in profits due to worker productivity.

It's like the time theft myth, the belief that workers will intrinsically slack unless you micromanage them. In fact the opposite is true, when workforces are allowed to chat with loved ones and surf the internet their productivity improves.

Curiously, when a company mandates that 15-minute breaks are taken and lunch is eaten during lunch break (even when workers are obsessive and want to work through breaks) then their productivity improves even more.

But the upper management of companies don't actually want to manage in a way that data shows maximizes profits.

It was everyone's golden era. And then the owner class pressured the political class to memory hole the period so we forgot all the lessons we learned.

It's why we need to bring unions back to forcibly make sure those lessons are applied.

Can confirm. How Money Works talks about it (Here on YouTube) regarding the recent mass layoffs by big tech, first, by pressure through RTO mandates, then blaming layoffs on AI, only they got rid of too many people and their businesses couldn't run.

So why did they do it? Vindictiveness. Because of the great resignation during which workers had some negotiating power, and the C-suite resented that and wanted to show everyone who's boss.

I've been following the trend since the early 2010s when we realized that crunching development teams is useless for meeting deadlines, but dev companies crunch anyway, traumatizing the team.

I'm pretty sure billionaires aren't especially bright, rather started with generational wealth and then got richer through luck and ruthlessness, to the point that they only failed upwards. (There are a few exceptions who understand the social contract, but they really are exceptions.)

And they surround themselves with sycophants so that their psyches are completely comfortable. They stop actually considering data or news that might hurt their feelings, which means they make bad decisions that cost their companies money.

This is how the kids know they are not in an institution of learning, but a prison for the crime of being juvenile.

What we're also not going to hear is how often devices are misplaced, confiscated without return or just dissappear like socks in a dryer.

Right now AI is looking a lot like a bubble. Wealthy people keep investing in it, but all the AI companies are losing money way faster than their profits. They're selling compute at a vast loss, and if they charged full price, no-one would buy it.

With tin-foil hat firmly in place, this is what the owner class hopes to get out of AI:

1. Automate everything. As Corey Doctorow surmises, they want to automate as much as possible, until the working class just isn't needed anymore, so they can build their Ayn Rand private city utopias (Peter Thiel has plans) and let the rest of us starve somewhere, displaced by the quarries, data centers and automated factories.

2. Create armies of swarming killer drones that are commanded by AI, as Randall Munroe predicted, not only to annex any territory that they can't buy, but to garrison all the working class they've confined to reservations to die of thirst and hunger. And...

3. In the meantime, process all the mass surveillance information in order to predict who is a dissident or a troublemaker or is capable of organizing resistance, so that they can be dispatched early. Also to micromanage their dwindling workforce and make sure women can't go to an abortion provider.