Well, theft has never been the best foundation for a business, has it?

History would suggest otherwise.

What you see as "spot the error" type training, another person sees as absolute fact that they internalize and use to make decisions that impact the world. The internet gave rise to the golden age of conspiracy theories, which is having a major impact on the worsening political climate, and it's because the average user isn't able to differentiate information from disinformation. AI chatbots giving people the answer they're looking for rather than the truth is only going to compound the issue.

Ashley is a senior policy reporter for Ars Technica, dedicated to tracking social impacts of emerging policies and new technologies. She is a Chicago-based journalist with 20 years of experience.

And yet, despite 20 years of experience, the only side Ashley presents is the technologists' side.

I hope LLMs and generative AI crash and burn.

I'm thinking, honestly, what if that's the planned purpose of this bubble.

I'm explaining - those "AI"'s involve assembling large datasets and making them available, poisoning the Web, and creating demand for for a specific kind of hardware.

When it bursts, not everything bursts.

Suddenly there will be plenty of no longer required hardware usable for normal ML applications like face recognition, voice recognition, text analysis to identify its author, combat drones with target selection, all kinds of stuff. It will be dirt cheap, compared to its current price, as it was with Sun hardware after the dotcom crash.

There still will be those datasets, that can be analyzed for plenty of purposes. Legal or not, they are already processed into usable and convenient state.

There will be the Web covered with a great wall of China tall layer of AI slop.

There will likely be a bankrupt nation which will have a lot of things failing due to that.

And there will still be all the centralized services. Suppose on that day you go search something in Google, and there's only the Google summary present, no results list (or maybe even a results list, whatever, but suddenly weighed differently), saying that you've been owned by domestic enemies yadda-yadda and the patriotic corporations are implementing a popular state of emergency or something like that. You go to Facebook, and when you write something there, your messages are premoderated by an AI so that you'd not be able to god forbid say something wrong. An LLM might not be able to support a decent enough conversation, but to edit out things you say, or PGP keys you send, in real time without anything appearing strange - easily. Or to change some real person's style of speech to yours.

Suppose all of not-degoogled Android installations start doing things like that, Amazon's logistics suddenly start working to support a putsch, Facebook and WhatsApp do what I described or just fail, Apple makes a presentation of a new, magnificent, ingenious, miraculous, patriotic change to a better system of government, maybe even with Johnny Ive as the speaker, and possibly does the same unnoticeable censorship, Microsoft pushes one malicious update 3 months earlier with a backdoor to all Windows installations doing the same, and commits its datacenters to the common effort, and let's just say it's possible that a similar thing is done by some Linux developer believing in an idea and some of the major distributions - don't need it doing much, just to provide a backdoor usable remotely.

I don't list Twitter because honestly it doesn't seem to work well enough or have coverage good enough.

So - this seems a pretty possible apocalypse scenario which does lead to a sudden installation of a dictatorial regime with all the necessary surveillance, planning, censorship and enforcement already being functioning systems.

So - of course apocalypse scenarios were a normal thing in movies for many years and many times, but it's funny how the more plausible such become, the less often they are described in art.

Fucking good!! Let the AI industry BURN!

And this is exactly the reason why I think the IA will be forced to close down while AI companies that trained their models on it will not only stay but be praised for preserving information in an ironic twist. Because one side does participate in capitalism and the other doesn’t. They will claim AI is transformative enough even when it isn’t because the overly rich invested too much money into the grift.

Not everything is public domain, thief scum.

That would be glorious. If the future of your company depends on the LLM keeping track of hundreds of details and drawing the right conclusions, it’s game over during the first day.

Good!!! Let the AI industry fucking burn!!!

Do they even teach the constitution anymore?

I agree that this has to become better in the future, but the technology is pretty young, and i am pretty sure that fixing this stuff has a high priority in those companies - it's bad PR for them. But the people are already gorging themselves on faulty info per social media - i don't see that chatbots are making this really worse than it already is.