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.
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?
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…?
The speculation I heard in the Ars Technica article is that the board was unhappy with how quickly he was pushing to commercialize OpenAI, and they were wary about all the AI side hustles he was starting, including an AI chip company to compete with nvidia.
Yeah, I feel like having someone who can cast Magic Missile is almost mandatory for that fight, simply because the illusions have 1 HP, are very spread out, and Magic Missile can target multiple enemies and is guaranteed to hit. It’s perfect for killing almost all of the illusions in a single turn.
I think it’s sorta okay that the enemies don’t get too much stronger, especially since (at least for casters) a lot of the added power comes in the form of gaining access to stupidly OP spells like Hypnotic Pattern. I don’t think it would be very fun if enemies started using tactics that amounted to “Hahaha, I rolled higher initiative so now you don’t get to play for the next three rounds while I can do whatever I want.”
Theoretically it can happen. In practical terms, 99% of those cases are out of three things:
A charade to get an angry customer to go away (pretending to fire an employee)
The last straw in a series of incidents that add up to justify firing the employee (i.e. the employee has repeatedly made a mistake with no improvement over a long period of time)
Misconduct egregious enough to warrant firing them on the spot (for example, the employee punches a customer, or shows up to a job site blackout drunk)
The remaining 1% of cases are truly shitty managers that are a nightmare to work for.
Not OP, but I think the point they’re making is that LTT screwed up the video, and that the drama sparked from LTT’s screwup gave Billet a lot of publicity they wouldn’t have had otherwise.
Personally, I’d trade the publicity for my only working prototype and $2,000 GPU back and a video that didn’t shit on me, but if you believe any publicity is good publicity…
This is admittedly REALLY pedantic, but there were some non-game cartridges released for the NES and SNES, such as Taboo: The Sixth Sense (a tarot card reading program), Miracle Piano (a program for teaching how to play the piano), Mario Paint (a basic music composition and drawing program), and a modem add-on for the Famicom that supported banking, stock trading, and horse race betting.
You know it’s coming. Why would a streaming company want a consumer buying one month, binging a single show they’re interested in, then immediately cancelling the subscription after, when you could guarantee a 6- or 12-month revenue stream for them?
It’s very user friendly in terms of tooltips, and if you don’t make deliberately bad choices during level up (e.g. taking a feat that gives you a cantrip from the Wizard class… that scales off your INT score… while playing a Barbarian with 8 intelligence that can’t cast spells while raging) it’s fairly difficult to make an unplayably bad character.
There’s a few cases where some general knowledge of D&D is helpful, such as knowing to never take True Strike because it’s literally worse than just attacking twice and having some knowledge of good builds is useful, since it helps guide what you take when you level up. That said, there’s also entire categories of actions in BG3 that don’t really have an equivalent rule in TTRPG 5e, such as weapon proficiency attacks, so online cookie cutter builds don’t capture the full extent of what you can do.
It’s free*, insanely easy to set up, you don’t have to worry about port forwarding or ddos or hosting fees, has powerful moderation tools, and there’s a plethora of easy to deploy bots that help manage permissions and automate routine tasks. Literally, if it had a proper web-accessible forum similar to phpBB, it would be perfect.
I just checked and literally every post in the past three days was submitted by just two accounts, with one user accounting for easily 95% of the posts.
Jesus, reddit really is dying.
I worked graveyard shifts at a gas station for a year or two. My general experience beyond what other people have said–good commute, fucking with your social life, taking its toll on your body, all that–is that working graveyard shifts is lonely. I cannot understate how lonely it got; there were stretches of multiple hours where there were no customers at all, and it was just me and the long list of nightly chores I had to do (mopping floors, prepping food for breakfast rush, restocking shelves, etc., etc.). Not having any human contact at all fucks with your head something fierce, especially when you mix in sleep deprivation and your body rebelling against the normal sleep rhythm into the mix.
My advise is that if you’re going to be working night shift all alone, get into podcasts. Having a radio that I could use to listen to NPR was the main thing that kept me sane, because I could at least have a human voice to listen to and keep my mind somewhat engaged.
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!