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.
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.
I hate it because it’s a gigantic waste of time and resources. Big tech has poured hundreds of billions of dollars, caused double digit percentage increases in data center emissions, and fed almost the entire collective output of humanity into these models.
And what did we get for it? We got a toy that is at best mildly amusing, but isn’t really all that actually useful for anything; the output provided by generative AI is too unreliable to trust outright and needs to be reviewed and tweaked by hand, so at best you’re getting a minor productivity gain, and more likely you’re seeing a neutral or negative impact on your productivity (or producing low-quality crap faster and calling it “good enough”). At worst, it’s put a massive force multiplier in the hands of the bad actors using disinformation to tear apart modern society for their personal gain. Goldman Sachs released a report in late June where they pointed this out: if tech companies are planning on investing a trillion dollars into AI, what is the trillion dollar problem that AI is going to solve? And so far as I can tell, it seems that the answer to the question is either “it will eliminate millions of jobs and wipe out entire industries without any replacement or safety net, causing untold human suffering” or (more likely to be the case) “there is no trillion dollar problem AI can solve and the entire endeavor is pointless.”
Even ignoring the opportunity cost–the money spent could have literally solved the entire homelessness crisis, world hunger, lifted entire countries out of poverty, or otherwise funded solutions for real, intractable, pressing problems for all of humanity–even ignoring that generative AI has single-handedly erased years of progress in reducing our C02 emissions and addressing the climate crisis, even ignoring the logistical difficulty of the scale of build-out being discussed requiring a bigger improvement in our power grid than has been done basically ever, even ignoring the concerns over IP theft and everything else, fundamentally generative AI just isn’t worth the hype. It’s the crypto craze and NFT craze and metaverse craze (remember Zuckerberg burning 36 billion to make a virtual meeting space featuring avatars without legs?) all over again, except instead of only impacting the suckers who bought into the hype, this time it’s getting shoved in everybody’s face even if they want nothing to do with it.
But hey, at least it gave us “I Glued My Balls To My Butthole Again.” That totally makes the hundred billion investment worth it, right?
Just because something is transformative doesn’t mean that it’s fair use. There’s three other factors, including the nature of the work you copied, the amount of the copyrighted work taken for the use, and the effect on the market. There’s no way in hell I believe that anyone can plausibly say with a straight face “I’m taking literally all of the creative works you’ve ever produced and using them to create a product designed to directly compete with you and put you out of business, and this qualifies as a fair use” and I would be shocked if any judge in any court heard that argument without laughing the poor lawyer making it out of the court.
Good question!
First, that artist will only learn from a few handful of artists instead of every artist’s entire field of work all at the same time. They will also eventually develop their own unique style and voice–the art they make will reflect their own views in some fashion, instead of being a poor facsimile of someone else’s work.
Second, mimicking the style of other artists is a generally poor way of learning how to draw. Just leaping straight into mimicry doesn’t really teach you any of the fundamentals like perspective, color theory, shading, anatomy, etc. Mimicking an artist that draws lots of side profiles of animals in neutral lighting might teach you how to draw a side profile of a rabbit, but you’ll be fucked the instant you try to draw that same rabbit from the front, or if you want to draw a rabbit at sunset. There’s a reason why artists do so many drawings of random shit like cones casting a shadow, or a mannequin doll doing a ballet pose, and it ain’t because they find the subject interesting.
Third, an artist spends anywhere from dozens to hundreds of hours practicing. Even if someone sets out expressly to mimic someone else’s style, teaches themselves the fundamentals, it’s still months and years of hard work and practice, and a constant cycle of self-improvement, critique, and study. This applies to every artist, regardless of how naturally talented or gifted they are.
Fourth, there’s a sort of natural bottleneck in how much art that artist can produce. The quality of a given piece of art scales roughly linearly with the time the artist spends on it, and even artists that specialize in speed painting can only produce maybe a dozen pieces of art a day, and that kind of pace is simply not sustainable for any length of time. So even in the least charitable scenario, where a hypothetical person explicitly sets out to mimic a popular artist’s style in order to leech off their success, it’s extremely difficult for the mimic to produce enough output to truly threaten their victim’s livelihood. In comparison, an AI can churn out dozens or hundreds of images in a day, easily drowning out the artist’s output.
And one last, very important point: artists who trace other people’s artwork and upload the traced art as their own are almost universally reviled in the art community. Getting caught tracing art is an almost guaranteed way to get yourself blacklisted from every art community and banned from every major art website I know of, especially if you’re claiming it’s your own original work. The only way it’s even mildly acceptable is if the tracer explicitly says “this is traced artwork for practice, here’s a link to the original piece, the artist gave full permission for me to post this.” Every other creative community writing and music takes a similarly dim views of plagiarism, though it’s much harder to prove outright than with art. Given this, why should the art community treat someone differently just because they laundered their plagiarism with some vector multiplication?
The used game market is still insane, I’m seeing $20-30 for even shit-tier, obscure, normally worthless nes games. If you bought the console while it was new it’s still worth keeping, but absolutely just get a flash cart instead of subjecting yourself to the price gouging retro market.
It’s not just “worse” graphics. CRTs have little/no input lag, which is crucial for some older games like Punch-Out!.
And here I am proud that I can open a file, save it, make edits, jump to the beginning/end, and quit it. 😅
Yeah, there are so many moments I wish I had a time machine so I could go back and yell at various people while shaking their shoulders.
For the love of God, Barack, don’t make fun of Trump at the White House correspondent’s dinner, he’ll run for president to dismantle all you’ve built up in revenge and HE WILL WIN.
Please, Ruth, I beg you to step down now while there’s still an opportunity for you to be replaced with another liberal justice. If you don’t, your legacy will be undone I’m under four years and it will herald the end of American democracy.
Please, Barack, don’t let them steal a supreme court seat like this, you have to force the issue while there’s still time or else you will watch the heritage foundation gloat about the second American revolution against the left while a corrupt court anoints the president as above the law of the land.
For the love of God, Biden, please run in 2016, I know you’re still grieving over the death of your son, but if you don’t you’ll be grieving over the death of your entire country.
For the love of God, Hillary, please step aside and let Sanders be the candidate, I know you agreed with Obama that he would give you SoS in return for you running after him but the Republican propaganda machine has made you toxic.
Barack, you can’t sweep this Russian interference under the rug, it’s too important to ignore, please!
I beg you, Hillary, don’t ignore the rust belt, your numbers are weaker than they should be there and they are too important to lose, the literal future of democracy is at stake.
For fuck’s sake, Comey, don’t reopen this stupid email investigation two weeks before the election, we both know there’s nothing on that fucking laptop. You need to shut down the trumpy faction before they leak its existence because they are trying to interfere with the election, and if Trump wins he will reward you with a pink slip while gleefully dragging the country to a dictatorship.
This timeline could’ve been so easily avoided, if only one variable out of dozens was different. But here we are, with me wondering where I can even flee to in order to escape the coming dictatorship.
The hush money payments were before the election so they were still felonies. It’s possible Trump wouldn’t have bothered hiding them like he did if he lost, but in any case there wouldn’t have been the appetite to prosecute him.
As for the coup, he absolutely would have tried it, definitely through filing lawsuits, and probably up to the same fake electors scheme and the riot, but it’s debatable how far he would’ve gotten. Definitely not as far as he did in reality, though.
Hoo boy, it’s a toughie. On the one hand, Trump would still be around. He also wouldn’t be in as much legal peril as he is now (it’s likely there wouldn’t have been an appetite to prosecute him over the Stormy Daniels hush money payments, and the classified documents case would have never happened to begin with since he wouldn’t have had access). But he almost definitely WOULD have tried to pull off another insurrection similar to Jan 6th–he was foreshadowing that he wouldn’t accept the results if he lost even back in 2016, using the same language as he did in 2020 before he launched his coup attempt.
The world where Trump doesn’t attempt a coup isn’t very interesting, at least for this thought experiment–he slinks off, continues shitposting about Hillary on Twitter, but likely doesn’t try to run again (or loses in the primary because he’s a sore loser). Everyone ignores his hush money payments in the interest of “statesmanship,” and at best he becomes a minor kingmaker in the party apparatus. MAGA withers on the vine, and we largely continue with the late Obama administration status quo.
The world where he attempts a coup is much more interesting. The real question is, what would have changed after the failed insurrection attempt? It’s highly unlikely it would have succeeded or even gotten anywhere as close as it did, since a lot of the original plan relied on access to the levers of power (I.e. being able to withhold security to let the rioters overrun the Capitol). But how would everyone react to it long-term? In this timeline, Republicans genuinely distanced themselves from Trump and Jan 6th at first, likely out of shock over the realization that they were actually in danger and the very real fear that they could end up hurt or killed. But as the shock wore off, Republicans started shuffling back to MAGA as the propaganda machine did its work to downplay and normalize the failed coup, and they realized that their base saw Jan 6th as a good thing.
In a theoretical timeline where Trump tries a coup in 2016, it depends on how far Trump gets before he fails. If he’s thwarted to the point where he doesn’t (or can’t) hold the rally that stormed the Capitol, then nothing really comes of it at all–it becomes a footnote in history that is only cared about by political historians, pub trivia enthusiasts, and people who like to talk about politics on the internet. If he gets to the point where he holds a rally, but the rally is prevented from interfering with the certification process (complete with provocative images of cops in riot gear swinging at MAGA rioters), it’s likely that this downplaying and normalization would have been ironically amplified by virtue of the coup attempt being less successful. Without the visceral fear of hiding from rioters, Republicans would have no reason to distance themselves from the attempt, and they would almost immediately start using it as fodder to attack the new Clinton administration. In short, the hypothetical coup attempt would become another Benghazi scandal for Clinton–something that she had little real involvement in and largely wasn’t her fault, but that she gets blamed for anyway. Trump, meanwhile, would remain largely in the same position as in 2015–the dominant force in the party.
Aside from that, the court wouldn’t be as openly corrupt as it is now. It’d be filled by a moderate Clinton appointee if democrats have the 51 votes to abolish the filibuster for supreme court appointees (or held open by McConnell otherwise), and when RBG dies her replacement is decided by whoever wins the 2020 election. Roe v. Wade would still exist, the chevron deference would still be the law of the land, and we wouldn’t have the terrifying prospect of legally sanctioned presidential death squads.
Overall, I think we would be largely in line with the status quo of 2014-2015. Not great, with a worrying trend towards fascism and an establishment largely too busy huffing their own farts to address the vast majority of problems facing us, but a LOT better than where we are right now.
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?
God, same. One of my little annoyances in life is that my internal voice is a goddamn motor mouth and I literally CANNOT stop it.
I can stare at a white wall and watch paint dry, and my monologue will start philosophizing about watching paint dry, where the phrase came from, why I’m doing it (to try and silence my internal voice), then go on a wiki walk about how trying not to think about something makes you think about it more and the classic example of telling someone “don’t think about a brown bear” makes them think about bears, then I’ll start thinking about bears and my monologue is suddenly halfway across the world.
Put me in a sensory deprivation tank, and my internal voice starts ruminating about how Daredevil uses these to sleep, then goes off about fight sequences, and then superhero comics, and whoops I’m halfway across the world.
Even when I’m paying attention and listening, my inner voice is still motoring away, it’s just that it’s mirroring what is being said to me instead of going on its own wiki walk halfway across the world (though sometimes someone will say something that makes my internal voice go “wait a second, that makes me think of…” and then I stop listening while I go on a wiki walk).
I have ADHD, in case it isn’t obvious yet.
Funny, I tweaked my Linux PC at work to look like Windows XP. It’s so cursed, I love it.
One of my side projects at work is to record training presentations and I try to be so conscious about this–both trying to avoid the word salad slides, and also trying to make my lecture not just reading the slide word-for-word but actually explaining and expanding on the slide content (with my verbal lecture transcribed as a note in the slide and handed out for anybody who might be hard of hearing/doesn’t want to sit through a 30-minute video)
Y’know what? I’m gonna be even more of a furry now, just to spite you.
I haven’t accidentally deleted a bunch of data yet (which, considering 99% of my interaction with Linux is when I’m SSH’d into a user’s server, I am very paranoid about not doing), but I have run fsck on a volume without mounting the read/write flashcache with dirty blocks on it first.
Oops.
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!