Hrm, I would like to admit that while I talk about the AI bubble bursting, the Chinese stuff is a very different question. That could well prosper, which is something I find worrisome even though at least some of the Chinese stuff is open source.

View this comment - View article - View full comments

Man. Those kids are so sweet and wholesome playing fun games together. Not like the ones that used to chase me around the schoolyard hoping to beat me up. Envy.

View this comment - View article - View full comments

Some of the same artists now throwing spite at AI and anyone using it or being even mildly in favor of it, said the exact same thing about Photoshop and why it will kill all art.

I don't think they did, though? There were some people who were upset about art done on software, but they were different people with different concerns. Many of the people upset about AI are generally quite pro-technology.

I think it's also worth noting that in the past, some of the concerns about new technologies have been justified--just because a new technology won, does not mean that its overall impact was positive. It just means somebody made money from it. Success and beneficial impact are two different topics.

You can go back, for instance, to the water mill. Water mills were often imposed by feudal lords. Water mills ground flour more efficiently than hand mills, they harnessed the power of water to replace human muscle power, but that's not the main reason feudal lords liked them. Feudal lords liked them because they were a central point peasants had to bring their grain, where the lord could tax it. In order to force peasants to use this central point, feudal lords actively outlawed hand mills to stop peasants from grinding their own grain at home and not paying tax. Clearly the peasants would have preferred to use the less-efficient hand mills and not pay the lord, or there would have been no need to ban them. By the time the feudal period ended, hand mills were largely obsolete; the water mill had won the technological battle, which was really a political battle. But that wasn't actually a good thing for the peasants.

Generative AI is also a very political technology, and if it succeeds (which does not look like happening, at least in the current iteration) any benefits will flow overwhelmingly to uber-wealthy oligarchs. Some other benefits will go to people who, like me, can't do art, or who, like some other people, can't write their own essays. But costs will go to a lot more people than benefit and will be much more serious. A lot of the objections to LLM AI are essentially political in nature, and I think anyone should think two or three times before choosing to be on the oligarch side of those politics. Most other AI criticisms are about an observed dumbing down of what's on the internet from AI "slop", which does seem to be a very real phenomenon.

Photoshop was not a very political technology; it did not really shift the category of people who were doing art or how much money they made from it; to the extent that it made artists or graphic designers more productive, that mainly led to a "speed-up" in what was expected of artist production, and thus an inflation in how much art assets could be produced for any given project. Objections to it were mainly about artistic sensibilities, about the "feel" as it were of doing art, about the general idea of heightening and computer-orienting the technology around creating art. It was the kind of stuff we call "Luddism" even though it quite specifically isn't--Luddism has come to be associated with the idea of romantic, instinctive dislike of technology, but actual Luddism was about politics and the very real fact that industry was turning independent craftspeople who made decent livings into impoverished factory workers who worked 12 hour days in incredibly bad conditions. I do have some sympathy, to a point, for general appeals for less technology in our lives, but it's a quite different argument.

Just because people A didn't like technology A, and people B didn't like technology B that had applications in the same field, doesn't mean people A had the same concerns as people B, let alone that they were the same people.

View this comment - View article - View full comments

All that still doesn't necessarily make a game on Deck usable with older eyes. The basic thing about the Deck is that it makes up for the smallish screen by being closer to you than a big screen would be. But, lots of older people can't focus from close up. I myself don't have this problem big time, but I have noticed that when I'm looking at a fine print label, and I bring it closer to try to read it, I lose focus just before it gets close enough. That didn't happen when I was younger, I could focus from quite close in. Many older people have this problem worse than I do, plus not great sight in other ways; a Deck just isn't going to be that usable for them--not because anybody's done anything wrong, just because that form factor isn't going to work for them.

elmapul, I do think maybe it's getting to be time for a Steam Deck II. Maybe a couple of recent/upcoming chips and stuff make it worth while? That and Valve are going to need to strongarm the anti-cheat situation somehow.

View this comment - View article - View full comments

It's true that it is, in theory, possible to build good products with AI assistance. And a few people will do it. And those few people will be absolutely upset that, just because 95% of what's built with AI assistance is crap, everyone will insist that things built with AI be identified as such, and most will just use that to completely ignore AI stuff. It won't be very fair to that 5%, but that is not "the consumer"s problem.

So no, I don't think we are going to see lots of "good enough" AI projects finding their way to our wallets, because they will be tarred with the brush of all the lazy grifters making the "not good enough" AI projects, which will be the vast majority, and so nobody will even bother to find out that those particular ones are actually OK.

As to whether (generative, large language model) AI is here to stay, and whether (generative, large language model) AI is going to get better . . . It's here to stay, sort of, but it will get FAR less common for a while after the bubble bursts. All the companies making it . . . in the West, at least . . . are losing money on it at high speed. Bubble's gonna burst. And, it doesn't seem to be getting much better; in some ways it's getting worse. There isn't much larger to go in terms of the training; to get something really better they're going to need some different basis for AI, a new kind of model. That could happen, but it would be a whole different thing, at some unknown time in the future--it's vanishingly unlikely to come along in time to save the current crop of AI-related companies. And they wouldn't have the rights anyway.

View this comment - View article - View full comments

I have a couple of games that with luck would really benefit from this. Definitely interested.

View this comment - View article - View full comments

I suppose in theory, especially what with many games being so huge compared to the OS itself, you could distribute a game with its own custom Linux kernel that incorporated the kernel-level anticheat and have it run on that . . . .

View this comment - View article - View full comments

without clear direction

But there is a clear direction: Up the screen! :grin:

View this comment - View article - View full comments

There's always jerks in any community. In corporations they become the boss and nobody can say no to them. One of the big benefits of Free/Open Source software is that you can tell the jerks to go (insert appropriate expletives here) and do good things without them.

View this comment - View article - View full comments

So up to now, Steam Deck support wasn't really . . . Satisfactory.

View this comment - View article - View full comments