It's only local to China in the sense that they choose what {packages/CCP poison} go in there.

For all values of u, Linux is local to u in the sense that u chooses what packages go in there.

Ha ha, Linux already has..

# top -o gpu
top: unrecognized field name 'gpu'

.. d'oh! Never mind.

And is asking an AI to proofread something that much different than MS Word flagging spelling and grammar errors?

Great question, and I think there isn't any significant difference.

But LLMs use a type of math that apparently has the evil bit set. Meanwhile, up here in VirtueVille, the computer code we write to create things, uses different math that doesn't have the evil bit set. Conclusion: we can look down on those AI companies and their stupid users. ;-)

Now, I wouldn't expect an AI service to be better at copyediting than an average human, but I haven't kept up so I'm ignorant of their performance. (All my "AI" experience is that I used Claude twice to, it turned out, fix some typos in some RoR config initializers.) But whether an AI service is as good at copyediting as a human or not, it's that user's choice and, as long as they sign off on the work and really take responsibility for it (declare the words to be their opinion), then I've got no problem with them doing that, whether it uses LLMs or Markov chains or human slaves or Searle's Chinese Room.

The words will be judged by their quality, not provenance.

That said...

Here's a similar story where I'd probably come to the opposite conclusion(!), despite the clown in it pretty much doing what I preach above, though he was using the AI service to do a lot more than copyedit. Behold this 3M expert witness cross-examination (134 page PDF) in the Watson Grinding case. It's slow going at first, but eventually you'll be laughing and cringing like in a good episode of Curb Your Enthusiasm.

Fuck. I don't know.

> What you're not considering is that the manufacturers are building fab capacity like crazy right now.

Which they'll shutter or repurpose if the demand plummets. You think they're going keep those factories churning out memory chips if prices drop significantly?

At the end of the day they have to make a profit. Making a bunch of memory chips with a profit of $-1 per chip is not making a profit.

It just means Gemini will start being rude and will refuse to try to help anyone with a question.

No. The hardware being built for AI is basically using hardware that more closely resembles graphic cards minus the DisplayPort output than anything you can pick apart to insert in your PC. Even the memory modules are completely un-DIMM like.

At best, supply will return to pre-AI levels, but that'll just mean that memory reduces in cost to somewhere between today's absurd prices and Trump's post-tariff prices. The chances of it being reduced to Biden levels is pretty low.

Memory is going to be expensive for the foreseeable future. And all because of what's essentially a con job, an attempt to take over every substantial business in America by making their own employees unable to understand what it is they do, that nobody's prepared to do anything about.

Facial recognition is really neat technology, and we are all slightly better off thanks to its development. It's also imperfect, and nearly-guaranteed to be a net negative if you fully trust it, without human verification.

If you want to come out ahead by using cool technology, you're going to have to think about how to use it. There just isn't any shortcut around thinking.

If you have something which is only 99.9% reliable, that can be useful but it's going to be wrong sometimes. How are you preventing that wrongness from breaking your system? Insulting (much less turning-away) regular, recurring, paying customers is usually considered a really bad move, generally resulting in lower profits.

Free money falls out of the sky. Should we take it? Today we'll be interviewing Republicans, and learning that the answer is No because asceticism is a fundamental ideal that we should all strive toward. Burn a sacrifice instead!

If it's an AI data center (which is generally what people are concerned about the massive growth of) responsible for 911, then you're going to have a bad day whether powered or not.

The obvious solution here is to run the data centers in reverse. Basically feed them long winded manipulative slop, and have them convert it into the original art. This will mean the centers will then start generating power instead of using it up.

It's so obvious I can't believe nobody's thought of it before. We could probably solve global warming if we did this.

https://www.404media.co/we-tra...

Ask the average system administrator who maintains any small website whether the AI companies are still scraping the web, and you might find yourself surprised. Pretty much anything with high quality content that isn't a major well funded site like Reddit or Wikipedia is getting hammered to the point that websites are either constantly going down, or resorting to reCAPTCHA and other similar "Legit or AI?" technologies to limit their use. Every BBS/etc I've been using lately has that anime nurse with a magnifying glass interstitial (see openwrt.org for example) to try to limit the amount of hammering they're getting from OpenGPT, Anthropic, Google, etc.

This is also impacting the ability of those same websites to be found via search engines. It's literally a disaster for the open web.

Anthropic adding watermarking isn't going to do much except add even more self-serving justifications to the AI bros to hammer websites constantly, as now they can distinguish between their own crap and stuff that might possibly be human written.

If all that's changed is one word, there's not nearly enough word choices by Claude to allow them to put in the watermark. The watermark is based upon word choices.

So that specific incidence wouldn't actually get watermarked, not unless Claude decides to do more than correct one typo.

OTOH, it's entirely possible the response "one typo, you wrote CIa instead of CIA, fixed" would be worded in such a way that it would include the watermark. Which would make sense and fit with what Anthropic are claiming. It would also explain why the text explaining that you didn't write CIA correctly consists of four paragraphs plus a bulleted list and starts with an entirely unnecessary complement on your writing skills. Well, that, and the necessity for LLM companies to post manipulative language to try to make you feel good about using their products.

> Certainly no indication they were more popular than Twitter

Twitter died a long time ago. Elon Musk bought it and replaced it with X which is a far right website that "inherited" the user base Twitter had. The reason Threads is increasing is because nobody wants to be bombarded with messages about how they, or their friends, or just people who have done nothing wrong, are subhuman vermin who need to be exterminated.

X's user numbers are inflated. Bluesky is a fraud and BS's users are catching on to that. Threads is, for better or worse, the devil you know without having fascism pushed into your feed. In 2021, there's no way it would have beat out Twitter, but with X being what it is, it's both larger than it would have been then (I suspect, actually, it'd have closed by now), and X is, in reality whatever Musk claims, much much smaller than Twitter is.

And those of us who actually don't want corporate media have our alternative and are enjoying it very much. Once upon a time I would have mentioned it and been furious at all the supposed tech experts here who freak out about "having to choose a server", but, you know what? I don't want those guys in my socials anyway. Maybe one day you'll all "get it", but I doubt it.

> Having to produce physical media is a barrier to competition

No it isn't. The only barrier to competition is the whole digital signing thing which is true of both physical media and online delivery.

Physical media is dirt cheap and always has been. You can print off 1,000 BDs for $1,000 or less. Virtually every software developer capable of buying the equipment needed to produce a game is capable of spending that. What they can't do is actually get hold of the proprietary Sony keys and digital distribution stuff, and that's not going to change.