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.
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 gave up on it for now when the questline involving the NPC learning to write broke, and then I started crashing to desktop (without any logs anywhere, either in the Buffout directory or even in Windows’ Event Viewer) every time I left the Swan or fast traveled directly to it, even though traveling to another point literally fifty feet south worked just fine. And since there’s no logs describing the crash, I have no idea how to fix it.
I could probably fix it by uninstalling and re-downloading it again, but I have a goddamn data cap that my roommate already blows through every month with the fucking massive updates Fallout 76 has taken to pushing out, I have zero desire to download 60 GB of data (30 GB base game + 30 GB FOLON) every fucking time I sneeze wrong and make the game start crashing again. =|
It’s pretty okay. If you like the gameplay loop of scavenging parts to maintain and upgrade your car, and don’t mind the roguelite elements, it’s pretty fun, and it does a good job of creating tension–there’s been multiple occasions where I wanted to loot more but I was out of time and likely to die if I stayed much longer.
The world building is immaculate, but IMO unfortunately the plot doesn’t really pay off, and the ending isn’t… super satisfying. It does enough to drive you along (no pun intended). The best part of the game is easily the soundtrack, and the best song in the soundtrack is easily The Freeze.
Are you talking about the chibi models vs. more realistic models? I think that was an artifact of an FF trope left over from the NES era where the world sprites were limited to one tile due to NES hardware limitations while the battle sprites were more detailed 1x2 tiles, and this was kept all the way up to FF6 where they finally used the same sprite for world and battles.
I have no clue why they went back to using different/less detailed models for world exploration in FF7 (if I had to guess they were unfamiliar with the PSX hardware and the chibi models used fewer polygons), but that go a long way to explain why the FMVs sometimes used different models–IIRC, the FMVs with chibi models played directly from the field, and the ones with more detailed models had some kind of scene transition into them, or otherwise were used for major plot beats. It’s good they abandoned this entirely with FF8 onwards, though.
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?
For what it’s worth, I played the NES release of DQ1, and then a translation of the japan-only SNES release of DQ2 recently (I actually beat DQ2 last week) and I found DQ2 to be a much better game than DQ1 overall. DQ1 was… interesting, but it was very much a game that did not respect the player’s time in the least, to the point of expecting the player to fight literally hundreds of battles in order to grind up enough money and experience to afford the gear. The most charitable thing I can say about it is that the battle system was so rudimentary and so grindy that the gameplay felt more like it was focused on resource management–there was a tension in deciding whether you could afford to take another fight, or if you needed to return to town and spend money sleeping at an inn to heal (setting your grind back at least 1-2 fights with how piddly gold and XP drops were), optimizing efficiency in spending your MP to heal vs. the risk of dying to the next monster, etc.
DQ2 meanwhile was a much more robust and much less grindy game–the simple addition of multiple party members and multiple enemies in a single battle meant that your gold and XP gains were multiplied over the first game. While it still demanded grinding, it was much more reasonable about it, and it felt much more like a “modern” JRPG like you’re used to seeing.
Really? Fusion stops holding your hand around the time you get to sector 4, but it’s strictly linear until you’re literally about to go fight the final boss–the game explicitly blocks you from backtracking to areas that aren’t immediately plot-relevant.
God, yes, I tried to get into the game twice and both times I bounced off right around the part where you go from Hell on Earth to a fucking high fantasy castle on some random planet. I’ll just replay Doom 2016 if I want to shoot some demons.
This bird definitely listens exclusively to metal bands from Nordic regions featuring an opera singer dressed like an ice queen backed by instruments that sound like they just got dragged through a tar pit
Yeah, IIRC a knight’s suit of armor and weapons alone were worth more than most people in medieval times would ever earn in their entire lifetime. Knights traveling on horseback were the modern day equivalent of a celebrity rolling around town in a Ferrari
The problem is that there’s no incentive for employees to stay beyond a few years. Why spend months or years training someone if they leave after the second year?
But then you have to question why employees aren’t loyal any longer, and that’s because pensions and benefits have eroded, and your pay doesn’t keep up as you stay longer at a company. Why stay at a company for 20, 30, or 40 years when you can come out way ahead financially by hopping jobs every 2-4 years?
The guy apparently rescaled the models to make them similar in size, which, idk if I’d call that altering
I still think it’s kind of a BS claim though
It makes sense to judge how closely LLMs mimic human learning when people are using it as a defense to AI companies scraping copyrighted content, and making the claim that banning AI scraping is as nonsensical as banning human learning.
But when it’s pointed out that LLMs don’t learn very similarly to humans, and require scraping far more material than a human does, suddenly AIs shouldn’t be judged by human standards? I don’t know if it’s intentional on your part, but that’s a pretty classic example of a motte-and-bailey fallacy. You can’t have it both ways.
Who even knows? For whatever reason the board decided to keep quiet, didn’t elaborate on its reasoning, let Altman and his allies control the narrative, and rolled over when the employees inevitably revolted. All we have is speculation and unnamed “sources close to the matter,” which you may or may not find credible.
Even if the actual reasoning was absolutely justified–and knowing how much of a techbro Altman is (especially with his insanely creepy project to combine cryptocurrency with retina scans), I absolutely believe the speculation that the board felt Altman wasn’t trustworthy–they didn’t bother to actually tell anyone that reasoning, and clearly felt they could just weather the firestorm up until they realized it was too late and they’d already shot themselves in the foot.
…So your metric of “too much AI safety” is that it won’t let you fuck the fish…?
This is gonna turn into the gamer version of “this is extremely dangerous to our democracy” isn’t it