Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror

Comment The integrated coprocessor du jour (Score 1) 46

I remember when floating point was the luxuriously optional silicon. I try to be welcoming to new things even if I don't know how/if I'll use them, because I think they don't really inflate the cost of the processors much. (Am I right? I don't actually know.)

Long-term, I think there's widespread consensus that integrated floating point was a good idea. Even less controversial, integrated MMUs are a critically necessary part of our modern world. (It's hard to imagine that separate chips like the 68851 used to exist.) The vector stuff? Some code uses it. The cryptographic instructions? Oh hell yes! Maybe I'll get reamed for this, but I think the processor industry has a pretty good track record of making silicon that we eventually truly do light up.

This time, it's a little harder. The applications for LLMs seem so niche. Part of me thinks they're doing this several years too soon. But that said:

0. Neural networks have more applications than LLMs. However worthless you think LLMs are: if your computer is good at LLMs, what else might it be good at?

1. I strongly disagree with everyone who says the hypothetical applications for this should run "in the cloud" instead of on the user's own hardware. All my experience tells me that's definitely wrong. IF this "AI" stuff really isn't a bubble (I think it probably is), then getting coprocessors widely deployed for it out there, is a very good thing. "AI" is no different from non-"AI" logic, in that whatever you're doing, from the user's point of view it should be as local as possible, and with as few external dependencies as possible. You don't need to teach me that lesson again for the 100th time, dammit. Maybe a lot of laypeople will get stuck with OpenAI's (or whoever's) services, but we will want to run it on our machines.

2. Maybe the reason there are so few existing applications that use neural nets, is that the cheap hardware to make it practical isn't out there yet! Get the silicon out there and then developers will find uses for it. Back when I was stealing my employer's electricity (and coffee) at night, I ray-traced on a network of 80386s and 80486s, and the 486's floating point performance made me a lot more excited to work on my ray-tracer. Had it ran slower, I would have moved on to the next amateur time-waster sooner.

But I can see why consumers wouldn't care a bit, right now. By the time you have real use for this hardware, I think you'll have already retired the new machine that you're buying today. But wasn't that sort of the case with vector and crypto instructions too? Different people will check it out at a different pace. It's always been like that.

Comment Re:Time for legally mandated clarity (Score 3, Interesting) 116

The government should clearly define what this is.

One of the things that makes these such Interesting Times is that while your statement would normally make perfect sense, these days the government has a financial conflict of interest in this situation, so I think they should be completely disqualified from defining anything!

Trump is doing this for the purpose of personally profiting from it, and might possibly might even be one of the bettors. He (and everyone who works for him, since they necessarily lack independence) should not have any say at all in this bet's adjudication, other than as a biased advocate for whatever position best maximizes their own interests.

If you and I have a conflict, then neither you nor I are fit to be a neutral judge in that conflict.

Comment Re:neighbor's cow (Score 2) 53

Here's the problem: if you stop relying on my cow, then the rest of my family here in my house, might start thinking they are allowed to get their milk elsewhere, too, thereby avoiding all the mind-altering drugs that I have been secretly putting in the milk to control everyone. Stop poisoning my family's minds with this subversive "use a different cow" talk!

If UK and EU citizens don't have to use our data-hungry and ad-barking servers, then US citizens might get the same idea! Surely you can see why that's totally unacceptable.

Comment Probably slightly (Score 0) 149

As an ignorant outsider shooting off his mouth on the internet, I would speculate that hybrids are likely helping the adoption of purely electric cars in a minor way, by adding scale to the production of various electric components.

A hybrid EV (at least the type popularized by Toyota) is just an EV with a local gasoline engine+alternator bolted on, right? So there are still electric motors in each. And I ass/u/me all the stuff that would be on an ICE's serpentine belt are electrically-powered on a hybrid, so the AC compressor in this hybrid could theoretically be the same exact AC compressor as in that pure-EV, etc. Thus some of the two markets' parts can scale as one, making them both slightly cheaper.

Throw in plug-in hybrids, and then we also get the fact that these plug-in hybrid drivers are creating some demand for charging networks, which of course increases the utility of pure EVs as well.

Comment We're going to lose the word "algorithm" (Score 1) 37

algorithmic feeds

We need to find, capture, try, execute, and then piss on the grave of whoever decided that the word "algorithm" was the best word for what they didn't like about Facebook. Their hasty decision, combined the word's apparent mainstream sexiness (who knew?!) is going to result in the word's loss.

Comment Pretty good 4/1 article (Score 1) 144

The summary makes the article sound a lot dumber than it really is. But it is dumb, and the core dumbness from which the rest of the article derives is here:

I thought about setting up a self-hosted media server to stream everything to my phone. But ultimately, I got lazy

He knew real the solution to all his problems, but like he says, "I got lazy," and so he went to his comical Plan B. Despite weird statements like this..

Many folks are sick of streaming in general. They’re sick of giant corporations, algorithmic playlists, and an internet infested with AI slop.

..he actually doesn't appear to have any problem with streaming at all. It's just that he had been using a proprietary streaming service called Spotify, which does things very differently than a user-oriented approach (e.g. self-hosted subsonic API server) would do. Navidrome isn't a giant corporation, the algorithm of its playlist is "play what the user told me to play" and whatever AI Slop you play depends completely on what AI Slop you decided to add to your collection.

But by conflating streaming tech with proprietary streaming services, he gets to make up a lot of non-existent problems with streaming and sneak by the core premise of his entire article: "I got lazy."

So he decided against the obvious, and instead, went with a less convenient alternative. I particularly liked the part where he called cassettes "compact and super-portable" by comparing them to vinyl, instead of the actual media competition: flash memory, remote spinning-rust, etc.

Slashdot Top Deals

Nothing is rich but the inexhaustible wealth of nature. She shows us only surfaces, but she is a million fathoms deep. -- Ralph Waldo Emerson

Working...