deleted by creator

It’s amazing how many people forgot about the classical “get a rise out of everyone with shitty arguments” troll, or forgot that the way to deal with them was to ignore and ban on sight. Fuck, I was practically in diapers when Usenet and BBSes were a thing and I still remember “don’t feed the troll.”

Yeah, suuuuure you weren’t.

Note that the proof also generalizes to any form of creating an AI by training it on a dataset, not just LLMs. But sure, we’ll absolutely develop an entirely new approach to cognitive science in a few years, we’re definitely not boiling the planet and funneling enough money to end world poverty several times over into a scientific dead end!

You literally were LMAO

Other than that, we will keep incrementally improving our technology and it’s only a matter of time untill we get there. May take 5 years, 50 or 500 but it seems pretty inevitable to me.

Literally a direct quote. In what world is this not talking about LLMs?

Did you read the article, or the actual research paper? They present a mathematical proof that any hypothetical method of training an AI that produces an algorithm that performs better than random chance could also be used to solve a known intractible problem, which is impossible with all known current methods. This means that any algorithm we can produce that works by training an AI would run in exponential time or worse.

The paper authors point out that this also has severe implications for current AI, too–since the current AI-by-learning method that underpins all LLMs is fundamentally NP-hard and can’t run in polynomial time, “the sample-and-time requirements grow non-polynomially (e.g. exponentially or worse) in n.” They present a thought experiment of an AI that handles a 15-minute conversation, assuming 60 words are spoken per minute (keep in mind the average is roughly 160). The resources this AI would require to process this would be 60*15 = 900. The authors then conclude:

“Now the AI needs to learn to respond appropriately to conversations of this size (and not just to short prompts). Since resource requirements for AI-by-Learning grow exponentially or worse, let us take a simple exponential function O(2n ) as our proxy of the order of magnitude of resources needed as a function of n. 2^900 ∼ 10^270 is already unimaginably larger than the number of atoms in the universe (∼10^81 ). Imagine us sampling this super-astronomical space of possible situations using so-called ‘Big Data’. Even if we grant that billions of trillions (10 21 ) of relevant data samples could be generated (or scraped) and stored, then this is still but a miniscule proportion of the order of magnitude of samples needed to solve the learning problem for even moderate size n.”

That’s why LLMs are a dead end.

Or they’ll do shit like put Harris on full blast for not providing “detailed policies,” and then moving the goalposts to “but how do you pay for it” when she does, and nitpicking every word of every sentence she says. Meanwhile, Trump will cancel interviews, go up on stage at rally, spew a word salad response, and the NYT will bend over backwards to reword the salad to make him look better, while casting his decision to dodge a second debate as “smart” and avoiding any form of scrutiny as “efficient use of campaign funds.” At best, they’ll halfheartedly throw in a fact check like “his plan to fix inflation by levying tariffs will increase inflation” but they don’t dare portray him as the senile, hate-filled lunatic he is because they’re terrified of angering their right wing audience (who are already shifting away from legacy media anyway to reinforce their bubble). They also do this because virtually all forms of legacy media have been coopted by the billionaire sociopaths that would very much like a second Trump term to give them another tax cut and the “freedom” to pollute our world and grind the heel of their boot into the face of the working class so that they can race to become the first trillionaire.

When IT folks say devs don’t know about hardware, they’re usually talking about the forest-level overview in my experience. Stuff like how the software being developed integrates into an existing environment and how to optimize code to fit within the bounds of reality–it may be practical to dump a database directly into memory when it’s a 500 MB testing dataset on your local workstation, but it’s insane to do that with a 500+ GB database in production environment. Similarly, a program may run fine when it’s using a NVMe SSD, but lots of environments even today still depend on arrays of traditional electromechanical hard drives because they offer the most capacity per dollar, and aren’t as prone to suddenly tombstoning when it dies like flash media. Suddenly, once the program is in production, it turns out that same program’s making a bunch of random I/O calls that could be optimized into a more sequential request or batched together into a single transaction, and now it runs like dogshit and drags down every other VM, container, or service sharing that array with it. That’s not accounting for the real dumb shit I’ve read about, like “dev hard coded their local IP address and it breaks in production because of NAT” or “program crashes because it doesn’t account for network latency.”

Game dev is unique because you’re explicitly targeting a single known platform (for consoles) or targeting for an extremely wide range of performance specs (for PC), and hitting an acceptable level of performance pre-release is (somewhat) mandatory, so this kind of mindfulness is drilled into devs much more heavily than business software dev is, especially in-house dev. Business development is almost entirely focused on “does it run without failing catastrophically” and almost everything else–performance, security, cleanliness, resource optimization–is given bare lip service at best.

There was a period of blissful ignorance circa 2006 where the only thing I knew about him was that he was rich and starred in the apprentice. Then a black man became president and the world hasn’t been free of his stench ever since.

And now we’re in full mask-off accelerationist theory “it’s okay to let Trump win as long as Democrats are punished” bullshit. You’re unhappy with Democrats, so you’re okay with letting throwing literally everyone on the left in the US under the bus, along with the entire country of Ukraine, and throwing even more bombs at Gaza.

What an entitled, smug, self-righteous, holier-than-thou position, utterly divorced from real life consequences. Thanks for admitting that you’re a thoroughly unserious poster, though!

A McDonald’s burger from the 80s when they still used tallow for their fries and higher quality ingredients? No contest, it was better than the TrumpDonalds that got served in 2017-2020

Shit, there’s roughly even odds that the records of her employment no longer exist because the franchise she worked at shut down, or destroyed them because they were 40 years old.

I used to work at a Sprint call center in the early '10s, but if you asked me to prove it I literally couldn’t, because the staffing agency I was hired through, the call center, and Sprint itself no longer exist. Doesn’t change that I worked there.

Jokes and (valid) worries about how many men are still supporting this dumpster fire aside… A poll like this has got to be setting off the fire alarms at Trump campaign HQ and I am giddy as hell to see it. The last time Democrats came this close to winning the overall male vote was 2008. If this margin holds out we could be looking at an absolute blowout (or at least as close as one gets in today’s climate). Shame the Senate map means we won’t get a 60-seat Senate, though…

The only way I could buy it is if it came out that it was faked, but their plan involved keeping the gunman alive so he could spout fake-leftist buzzwords and they either didn’t realize that the secret service shoots to kill, or forgot to clue them into the plan ahead of time.

I personally did read it that way, but I will concede that perhaps I was being uncharitable.

Regardless, I have seen people explicitly questioning whether it was faked elsewhere, and it makes me cringe every time. Talking about this serves literally zero purpose–it makes the left look crazy, any alternative explanations that make Trump look bad fall apart under the barest scrutiny, and it just serves to keep the assassination attempt in peoples’ minds. There are literally hundreds of other things to complain about Trump over, talking about this doesn’t help.

Okay, but what’s the alternative? Trump faked the whole thing in some sort of false flag? He planted a fake gunman to get killed by the secret service, and put two of his close supporters in the hospital in critical condition, for a bump in the polls, when he was already confident that he could beat Biden? Is that really a more plausible explanation than “someone decided to kill Trump over the Epstein files, missed, and was killed”? I absolutely hate the guy, buy I just don’t buy it. I can accept “he got hit by a shard of glass instead of a bullet” or “he got grazed elsewhere and it just looks like he was hit in the ear” but claiming the whole thing was faked is just a bridge too far.

We’re supposed to be above this type of shaky conspiracy theory level thinking.

But hey, now they can claim that not picking Shapiro was antisemitic! 🙄

Never mind the tiki torches and chants of “Jews will not replace us” in the distance…

I was one of those that thought biden shouldn’t drop out, because I was worried about the risk of infighting breaking out over who would replace him, distracting everyone and driving away voters, and I was also concerned about throwing away the incumbent advantage. I still feel the risks of that happening were real and valid, but I’m immensely pleased that those worries didn’t come to pass and everyone immediately unified behind Harris.

And the second and third quotes, that were you?

Take the goddamn L, man. You made a statement in ignorance, you were wrong, and you were given evidence showing you were wrong. Accept it, learn from your mistake, and be better in the future.

Basically, X11/Xorg doesn’t isolate programs from one another. This is horrible for security since malicious software can read every window, as well as all the input from mice and keyboards, just by querying the X server, but it’s also handy for screen reading software, streaming, etc. Meanwhile, Wayland isolates programs in their own sandbox, which prevents, say, a malicious browser tab from reading all of your keyboard inputs and logging your root password, but also breaks those things we like to use. To make matters worse, it looks like everyone’s answer for this and similar dilemmas wasn’t “let’s fix Wayland” but “let’s develop an extension to fix Wayland” and we wound up with that one fucking xkcd standards comic that I won’t bother linking because everyone has seen it a zillion times.

ETA: Basically, my (layman’s) understanding is that fixing this and making screen readers work in Wayland is hard because the core Wayland developers seem to have little appetite for fixing this themselves. Meanwhile, there’s 3-4 implementations of Wayland that do things differently, so fixing it via extensions means either writing multiple backends in your program to do the same damn thing (aka a giant pain in the ass) or getting everyone to agree on the same standard implementation (good fucking luck).

Why make a better UI when it’ll probably introduce a slew of new bugs?