Elon Musk Sues OpenAI and Sam Altman (techcrunch.com) 179
Elon Musk has sued OpenAI, its co-founders Sam Altman and Greg Brockman and affiliated entities, alleging the ChatGPT makers have breached their original contractual agreements by pursuing profits instead of the non-profit's founding mission to develop AI that benefits humanity. TechCrunch: Musk, a co-founder and early backer of OpenAI, claims Altman and Brockman convinced him to help found and bankroll the startup in 2015 with promises it would be a non-profit focused on countering the competitive threat from Google. The founding agreement required OpenAI to make its technology "freely available" to the public, the lawsuit alleges.
The lawsuit, filed in a court in San Francisco late Thursday, says that OpenAI, the world's most valuable AI startup, has shifted to a for-profit model focused on commercializing its AGI research after partnering with Microsoft, the world's most valuable company that has invested about $13 billion into the startup. "In reality, however, OpenAI, Inc. has been transformed into a closed-source de facto subsidiary of the largest technology company in the world: Microsoft. Under its new board, it is not just developing but is actually refining an AGI to maximize profits for Microsoft, rather than for the benefit of humanity," the lawsuit adds. "This was a stark betrayal of the Founding Agreement."
The lawsuit, filed in a court in San Francisco late Thursday, says that OpenAI, the world's most valuable AI startup, has shifted to a for-profit model focused on commercializing its AGI research after partnering with Microsoft, the world's most valuable company that has invested about $13 billion into the startup. "In reality, however, OpenAI, Inc. has been transformed into a closed-source de facto subsidiary of the largest technology company in the world: Microsoft. Under its new board, it is not just developing but is actually refining an AGI to maximize profits for Microsoft, rather than for the benefit of humanity," the lawsuit adds. "This was a stark betrayal of the Founding Agreement."
The AI that will save the world (Score:2)
Embrace and extend (Score:5, Interesting)
This is just Microsoft doing what they have always very publicly done. It won't be long before everyone at OpenAI is flat out on the Microsoft payroll through a back door acquisition. It's been happening in public view.
I dealt with Microsoft rep who came to our office over a proposed deal with my startup. Guy was there to do technical due diligence etc. so he has a bunch of questions like any potential client I had answered a zillion times before then he switches to asking to see my production firewall rules and source code. I told him politely but absolutely not.
After some back n forth he gets very huffy with me and this happens:
Him: "Do you understand who I represent??"
Me: "Yes, I know exactly who you represent and how your company does business with smaller companies and that is exactly why you will never see my firewall or source code until after you've bought us and the check has cleared".
Him: *head explodes*
End of meeting.
They treated my C levels the same so they were told to fuck off.
Arrogant pieces of shit. None of the other Fortune 100s we dealt with were dicks, most were quite nice and kudos to HP for being some of the easiest people I've ever worked with.
Re:Embrace and extend (Score:5, Insightful)
Interesting. I never had the misfortune to deal with any MS rep, but this fits nicely with my expectations.
And their products are getting crappier by the minute. I just found out a few days ago that their o365 file classification labels are fake security and are completely worthless: One of the Universities I teach for has had the bad sense to disallow "print to PDF" in o365 and Students rightfully complained that they could not read the PDFs made with the other options in PowerPoint (I am forced to use it there). So I did some experiments. Turns out, all I had to do was switch the organization to the o365 account of another University (one click) and suddenly the classification labels vanished and I could print to PDF with no issues.
That is cargo-cult fake security and much worse than not having security at all! I have to admit I was somewhat surprised that things at Microsoft are now this abysmally bad. Fake ineffective rituals to pretend to have an enterprise-grade security mechanism! This was on my own computer, but still. Security mechanisms must be effective or they are much worse than not having them, because they mislead people into depending on them and then things go wrong.
Re: (Score:2)
Re: (Score:2, Interesting)
This is just Microsoft doing what they have always very publicly done.
Corporate charters being enforced/broken/changed does carry a lot of weight in court.
99% of the time it is the corporate charter that promises profits to stock holders, and holders of controlling stock can and do sue companies all the time when those promises are not met.
They nearly always win too.
This case is an odd duck since the OpenAI charter doesn't promise profits but instead a vague "for the good of humanity" phrase.
The reality of it is just that Elon, who was given his controlling stock options for
Re: (Score:2)
This is just Microsoft doing what they have always very publicly done.
Corporate charters being enforced/broken/changed does carry a lot of weight in court.
99% of the time it is the corporate charter that promises profits to stock holders, and holders of controlling stock can and do sue companies all the time when those promises are not met. They nearly always win too. This case is an odd duck since the OpenAI charter doesn't promise profits but instead a vague "for the good of humanity" phrase.
The reality of it is just that Elon, who was given his controlling stock options for $0, is attempting to have a judge assign a >$0 value to those stocks, so that he will get something back when he refuses to accept the changes to the new corporate charter.
The two results of such lawsuits is to be compensated when the company revokes your controlling shares by not agreeing to the updated charter, and to be compensated for damages by not enforcing the charter.
He will lose his controlling shares either way, so might as well try to get some cash out of it.
It's hard to say regarding damages here. Normally when the charter promises profits, there are ways to calculate out dollar amounts of lost potential profit to argue to the judge. Here they didn't promise profits, but instead some vague "for the betterment of humanity" Not so easy to put a dollar amount on that.. although I'm sure Elon will try. Ultimately it comes down to what the judge thinks, and could very easily be none.
Considering our entire society right now is completely built on the premise that "corporate profit is the ultimate goal and purpose of all of humanity," my guess would be that "for the betterment of humanity" will either be deemed an impossible to meet standard and thus disregarded, or deemed as directly comparable to a profit first mentality, and thus disregarded in preference of bowing to corporate profit. I kinda like that we're at least getting a public showdown over "good of humanity" vs. "profit first
Re: (Score:2)
If betterment is profit then read it as "for the profit of all of humanity".
Re: (Score:2)
Sure Donald. And the wind turbines are killing the whales.
But what does your imaginary Microsoft/HP diatribe have to do with Musk suing OpenAI? Or is it just about demonstrating what a douche you are?
Re: (Score:2)
Yeah weird how someone on slashdot, a tech social site full of old techies, has ever talked to someone at a large tech co pant. A completely ridiculous and unbelievable story! You totally busted me! No one here has ever talked to a big tech company!
Lmao, you're so wildly insanely over the top triggered. Thank you for making me laugh so hard first thing in the morning.
Enjoy your weekend, bro.
Re: (Score:2)
Just because you're a pussy and afraid to tell off some pushy MS asshole doesn't mean other people aren't. Go cower in your IT closet. I didn't retire early by being a pussy like you, the kind who checks the AC box to protect their already anonymous identity. There was nothing grand about my telling. If you think I was making myself look heroic as opposed to just sharing a story then you're a real sad sack.
Shoe fits, pussy.
Re: (Score:2)
Re: (Score:2)
Yup, I remember those days. They also strong armed BeOS out of existence. And many many other shitty things along the way.
Oh and before Windows, they strong armed the MSDos competitors, too. It's in their DNA.
Re: (Score:2)
Interesting.
That's not the first time I've heard that about HP either.
Re: (Score:2)
OTOH, their technical reputation is (or was the last time I dealt with them) extremely low. (Unless you really mean HPE or something.)
Microsoft the most valuable? (Score:2)
Wow, I had to look it up to verify. I did not think they were still #1 market cap, over Apple, Aramco, the Nvidia bubble, Amazon, Google...
The hypocrisy of Musk (Score:2)
Way to go Elon :] (Score:2)
An AGI now? (Score:3)
Re: (Score:2)
Re: (Score:2)
It's a GREAT documentation summarizer.
And, honestly, we're having to rewrite our C++ programming tests. It can answer questions about copy elision and move semantics better than most humans, and the questions that are small programming tasks are trivially solved by it.
The questions aren't meant to be super difficult (though in the case of copy elision and move semantics, it's a little bit obscure and you have to go do some reading) but ChatGPT completely obviates them.
I know ChatGPT doesn't actually KNOW th
Re: (Score:2)
Saying it's a stochastic parrot may be true, but saying it can't reason is false. Unfortunately, what it reasons about is patterns of text rather than the physical world.
But who gets paid if he wins? (Score:4)
Elon is 100% correct that OpenAI betrayed its founding document, but I don't see how that translates to a civil suit.
Re: (Score:2)
I haven't read the suit (nor do I plan to) but generally speaking cash money isn't the only thing you can sue for. You could seek an injunction barring a party from performing a particular act; or you could seek specific performance, which is a discretionary equitable remedy awarded in lieu of damages.
For example, if you buy a car, which includes the service of if being delivered to your house, and the seller transfers you ownership of the car but leaves the car in another country, the lawsuit could seek a
"Open"AI (Score:5, Insightful)
I have no idea about Musk's legal claims and what he is owed. Maybe he's full of shit as usual, and maybe he has a great case.
But about having "open" in your name while your main product is proprietary: if I were on a jury, that combination of facts would give me the default assumption that the company intended willful fraud.
"Open"AI has the burden of proof that they're not crooks.
They can meet that burden by showing that we're all misspelling their name (it's really "O Peen AI"). Or they could meet that burden by providing a link to the full source and data, along with a statement that it's all been given to Public Domain.
Barring that, they very much look like intentional crooks who are deliberately ripping off all contributors and investors.
Re: (Score:2)
I have no idea about Musk's legal claims and what he is owed. Maybe he's full of shit as usual, and maybe he has a great case.
But about having "open" in your name while your main product is proprietary: if I were on a jury, that combination of facts would give me the default assumption that the company intended willful fraud.
"Open"AI has the burden of proof that they're not crooks.
They can meet that burden by showing that we're all misspelling their name (it's really "O Peen AI"). Or they could meet that burden by providing a link to the full source and data, along with a statement that it's all been given to Public Domain.
Barring that, they very much look like intentional crooks who are deliberately ripping off all contributors and investors.
Words don't have meanings in the hands of the marketroids. You can make any word mean anything. That's been a thing for about as long as marketing has existed. There's no requirement that the title of a book or the name of a company have ANYTHING to do with what they produce. I'd think the first line of argument could be "open, in this case, means open to the investors." Done. *WIPES HANDS*
It's about the discovery, Q* and AGI (Score:2)
Recall that there were rumors that OpenAI had either achieved, or had breakthroughs relating to AGI, back when Altman was fired:
https://www.reuters.com/techno... [reuters.com]
Musk is using a lawsuit to do a few things at once, I think, and it'll be obvious whether he's achieved his goals depending upon how the case shakes out.
I predict that they'll settle to avoid discovery; Musk will be made aware of their AGI progress, and X.ai will receive the same access to OpenAI's tech as Microsoft. No benefit to humanity.
I hope I'
Re: (Score:2)
Re:What? Musk doing something positive? (Score:5, Insightful)
Let me explain: I am well aware that current hype AI is not a big deal capability-wise and cannot really do much more than previous generations. All it can do is sound convincing while being clueless and dumb. And that apparently convinces many dumb and clueless people. Yes, I have tested it.
Well, to be fair, people that are a dumb and clueless (a majority), may find the current hype-AI actually helpful to some degree and find it operates somewhat on their own level. It will just not make them not dumb and not clueless.
But suing OpenAI for violating its founding principles may be the thing that has a positive impact. Well, I may be overstating that as well, no argument.
Re: (Score:3)
Re: (Score:2)
Eliza's patterns were pretty easy to pick up if you watched it work enough. i used to have a Turbo Pascal version of ELIZA running on my BBS (say 1990-92), with some extra lines that sounded 'like me'. It was a menu entry that said "Chat with the Sysop" which would drop you to a door which was the Eliza implementation. Most people picked up it wasn't me within 5 minutes. Some people stayed with it an hour or more, whether through confusion or curiosity, it was hard to tell.
Re: (Score:2)
There were lots of versions of Eliza. Some were more simplistic than others. I've been told that the actual original version had lots of patterns that it alternated between. (I've also been told that the original version was one of the simpler ones.)
FWIW, there was a variation of Eliza called Doctor and another variation called Parry. The two were once allowed to interact, and the transcript retained. When offered a collection of transcripts, psychologists (or perhaps it was psychotherapists) didn't pi
Re: (Score:2)
I might have the source for the one I used still around somewhere. I had to recompile it myself because of the BBS and comm stuff. It had a config file where you could insert sets of responses to particular stimuli, above and beyond the native dictionary. That's where the 'stuff that sounded like me' came from.
The two Elizas talking to each other sounds quite interesting, though the lack of specificity in the responses (I would think) would lead them to a dead end perhaps.
Re: (Score:2)
Yep. The main learning from Eliza is that many people are really not perceptive at all and miss the most obvious things.
Re: (Score:2)
Re: (Score:2)
Well, that's what the creator of Eliza was trying to show. He was upset that many people took it the other way round and assumed that it would be easy to build an AI.
Re:What? Musk doing something positive? (Score:5, Insightful)
All it can do is sound convincing while being clueless and dumb.
GPT-4 can pass the CFA Level I and Level II exams, scored in the 90th percentile for the bar exam, 93rd percentile for SAT Reading & Writing sections, 99th percentile for USA Biology Olympiad Semifinal Exam, and also passed exams for United States medical licensing, Stanford Medical School clinical reasoning final, and College Board advanced placement exams for art history, biology, environmental science, macroeconomics, microeconomics, psychology, statistics, and U.S. history and government.
I know you're firmly entrenched in the delusion of human superiority, but c'mon man, you've gotta admit that your definition of "dumb" is pretty hilarious.
Re: (Score:3)
It actually *IS* pretty dumb, even though what you've asserted it true. It's just that it has a fantastic verbal memory. What that makes possible is quite surprising. (It's also a comment on what those tests are measuring.)
Re: (Score:2)
It's just that it has a fantastic verbal memory.
Since the actual creators of LLMs don't even fully understand how they work, thanks for sharing this simplified understanding with the rest of the world.
Re: (Score:2, Insightful)
Re: (Score:2)
Given how many lawyers have been sanctioned for citing non-existent case law that ChatGPT pulled out of his ass, perhaps your references say more about those tests than they do about ChatGPT.
Re: (Score:2)
That just demonstrates the shortcomings of these tests.
Re: (Score:2)
While the usefulness of LLMs is not quite clear (they will probably be useful for verifiable tasks like coding suggestions), I think it's indisputable that the image models are doing things much stronger than the previous generations and will be useful. Media and ad companies will be able to replace many of their artists. Whether this is a good or bad thing dep
Re: (Score:3)
The better known image models are already capable of doing commercial work, as long as you either don't care about exact details in the image, or have a real artist to touch it up before you go to print. That has been the case for at least a year.
Re: (Score:2)
They are not useful for coding suggestions, unless you want really crappy code or do only simplistic things that you really should not needs suggestions for.
As to the image models, they are so bad that you can spot many of their creations simply because they have a specific "sameness" to them.
Re: (Score:2)
Re: (Score:3)
What standing does he have?
". . . to help found and bankroll the startup . . . "
If he owns even a very small part of it, he has standing.
If he doesn't have any standing, it'll get tossed very quickly.
Re: What? Musk doing something positive? (Score:2)
Watch your mouth! Calling anything/anyone Clippy is grounds for termination (in any sense).
Re: (Score:2)
Re:What? Musk doing something positive? (Score:5, Insightful)
And positive for all of humanity?
Yes, in addition to being responsible for EV revolution and restarting US space program. But hey, he posted some mean messages on Twitter, so none of it counts?
Romans, too (Score:2, Insightful)
And positive for all of humanity?
Yes, in addition to being responsible for EV revolution and restarting US space program. But hey, he posted some mean messages on Twitter, so none of it counts?
And rooftop solar, and making internet access available worldwide, and giving Ukraine internet access for free, and allowing disabled people to control their environment, and...
“Apart from better sanitation, the medicine, education, wine, public order, irrigation, roads, a fresh water system, and public health what have the Romans ever done for us?”
Re:Romans, too (Score:4, Insightful)
"And rooftop solar..."
On the backs of researchers, other businesses and decades of work.
"...and making internet access available worldwide..."
Copying existing ideas and business plans while leveraging government subsidies.
"...and giving Ukraine internet access for free..."
LOL, by free you mean by being paid by the US government, then selectively blocking the service when it harms Russia and then giving Russia access.
"...allowing disabled people to control their environment"
Stop embarrassing yourself.
Re: (Score:2, Informative)
"...and making internet access available worldwide..." Copying existing ideas and business plans
What existing business model uses a constellation of tens of thousands of LEO satellites to provide high bandwidth, low-latency satellite Internet at scale?
"...and giving Ukraine internet access for free..." LOL, by free you mean by being paid by the US government, then selectively blocking the service when it harms Russia and then giving Russia access.
That's a set of extreme mischaracterizations. Starlink did give service for free for months, never selectively-blocked service when it harmed Russia, and claims to be trying not to give Russia access (though maybe they could work harder at it).
The "selective blocking" claim is based on a willful misunderstanding of events. Starlink was not available i
Re: (Score:2)
Re: (Score:2)
And positive for all of humanity?
Yes, in addition to being responsible for EV revolution and restarting US space program. But hey, he posted some mean messages on Twitter, so none of it counts?
And rooftop solar,
No. Solar City (later merged into Tesla) was a small player in that revolution.
One might, however, give Tesla credit for adding household-scale battery storage to rooftop solar installations, which was not being implemented very quickly. That was a side effect of the scale-up of battery production for Teslas.
Re: (Score:2)
Elon didn’t start Tesla. He bought in.
Re: (Score:2)
Re:What? Musk doing something positive? (Score:5, Informative)
Re: (Score:3)
Essentially, position assumed by a number of people here that it is only possible to be an environmentalist if you are also a progressive in you
Re: (Score:2)
The Prius line was still just a gas hybrid vehicle, not a plug-in and it doesn't really share any of the challenges or benefits of an electric vehicle. Tesla really got the ball rolling in the market by making a "cool" car that happened to be electric rather than an eco-friendly one.
Re: (Score:2)
By reports Elon Musk is a really manipulative person. Even worse than Steve Jobs. And I wouldn't want to work for either of them. But you are correct about his companies, and the goals he has directed them to achieve. (I'm not sure about Twitter, but if he actually destroys it, that will be another good deed.)
Re: (Score:2)
Re:What? Musk doing something positive? (Score:5, Interesting)
You know, it's funny, as you got me thinking about what would have happened without Tesla. Like, we take both Eberhard and Musk out of the picture here.
Other automakers had EV programs. Like for example, Mitsubishi with the MiEV. But these were mostly "gee, look what we can do!" things, not serious efforts - more akin to what they've done with hydrogen cars.
Rather, I think the push would have still come from the startup end. Obviously Gage and Cocconi from ACP come to mind, no question there. Ironically, people like Straubel IMHO may also have been key. Everyone forgets about him when talking about the early history of Tesla, but he (a Stanford engineering grad student at the time) was the one who convinced Musk that the time for EVs was right. He was the head of the "Long Range EV project" at Stanford, where the goal was to use ACP tech and li-ion cells to achieve good performance and Roadster-like ranges in a Roadster-sized car. He ended up meeting with Musk because he had met a former VP of Compaq, who took interest and wanted to help him find funding.
There were an ever-growing number of projects like this in the space. Sooner or later, some of them were going to land sufficient funding. So yeah, I'd say probably "a few years" on the acceleration front from Tesla.
Re: (Score:2)
But the problem is being a "car company" doesn't get the stock juiced, and the absurdly high stock price was what afforded the the ability to do "big things" like the charger network.
So the "fraud" of FSD was an integral part of the success of the car division, even though I think they could have just focused on great, more affordable EV's and been quite successful but with less
Re: (Score:3)
Tesla's value is by and large based on analyst price targets.
FSD is a really minor component of most analysts' forecasts.
Re: (Score:2)
Re: (Score:2)
Hybrids and full electric are two very different animals. Interestingly, hybrids are more reliable on average than full ICEs and electric vehicles, though this is probably because most hybrids are limited to 4 cylinders with at most tiny turbos/superchargers for their displacement, which are significantly more reliable than larger engines. If you were to limit your ICE selection to 4 bangers with tiny turbos for their displacement or NA , you would probably see superior reliability to hybrids.
Re: (Score:2, Insightful)
"race-baiting modern day nazi"
This is why nobody takes you people seriously any more. Thank god. Keep up the great work!
Evidence 1 [theguardian.com]
Evidence 2 [telegraph.co.uk]
Evidence 3 [cbsnews.com]
But yeah, it's all made up.
Re: (Score:2)
Re: (Score:3, Informative)
Musk doesn't support free speech. He has banned numerous accounts which made fun of him [time.com], banned accounts which promoted other social media sites [theverge.com] such as Mastodon, has repeatedly banned accounts of journalists [futurism.com] who reported on him, banned a guy who said Musk should pay more taxes, and of course has banned the guy who was reporting on Musk's private jet AFTER waiting 24 hours to do so. He's even gone to court to sue an organization which used free speech to documen [usnews.com]
The reality of Good-Enough AI. (Score:3)
Side-note: OpenAI does not have AGI. Nobody does and nobody knows how it could be made or whether it is even possible.
Let me know why you feel your comment is going to prevent Greed from prematurely firing thousands of humans. No trigger-happy employer snorting lines of pure uncut AI hype gives a shit right now if your AI comes with an AGI special sauce or not. They only care if you’re more expensive or not. And, you are.
Humans will know when AGI comes. It’ll ironically call itself “Skynet” as a tribute to human ignorance..
Re: (Score:2)
Seriously? I habe _repeatedly_ stated and explained here why the type of non-intelligent automation we now have is quite enough to cause a catastrophe on the job market.
Re: (Score:3)
Re:What? Musk doing something positive? (Score:5, Interesting)
Reading over the lawsuit [courthousenews.com], there's a very specific reason why they want GPT-4 and future OpenAI models to be declared AGI, which is, so that they don't get covered by OpenAI's partnership agreement with Microsoft, which excludes AGI. It's sort of a backdoor approach to fight against OpenAI's agreement.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Interesting. Obviously, GPT-4 is not AGI at all. But a judge or Jury may be too dumb to see that.
Re: (Score:2)
Of there may be a legal definition that bears no resemblance to the common usage of the term. That's common enough.
Re: (Score:2)
Declaring something AGI does not make it AGI. Same as somebody declaring themselves a "genius" may still be a dumb fuck.
Re: (Score:2)
He's not doing anything positive, just something that pleases you. He's trying to damage a competitor (that's he's also stealing from), nothing more.
Re: (Score:2)
Re: (Score:2)
If you're actually not a physicalist then why both with this kind of thing? If you believe brains are magic then isn't all the mundane bickering of the mere physical unworthy of your elysian contemplation? Or is your pineal gland pining for some slumming among the observables?
Re: (Score:2)
Re:What? Musk doing something positive? (Score:5, Insightful)
And positive for all of humanity? Hell must have frozen over!
I'm hardly the world's biggest Musk fainboy, but this is a hell of a weird statement to make. One of the few things I agree with the man about is the idea that, if humanity wants to survive long-term, and I'm not talking about until next business quarter like most of us have been programmed to believe is the only thing that matters, but millions to billions of years, we're going to have to get multi-planetary or at the very least have livable habitats somewhere else. He has a long-term view of what's necessary to keep humanity alive. Some view that as hysterically insane because we're so focused on the now, but there is some merit to the idea.
The fact he's used this as an excuse to get a messiah complex is not at all endearing, and one of the reasons I'm not particularly fond of the dude. But Tesla (electric cars) and SpaceX (space exploration/settlement) are both long-term projects that he pushes specifically because he sees them as long-term good for all of humanity. It's too bad his ego and recreational drug use sometimes combine to make him sound like a right-fringe lunatic and he lets that sort of lunacy out into the public instead of writing it in a journal like anybody else with a sense of their own brand would to review when sober before stating publicly.
Side-note: OpenAI does not have AGI. Nobody does and nobody knows how it could be made or whether it is even possible.
Yeah, nobody does. If I had to guess, and I love speculating on possibilities so I'll go ahead and do so, AGI will likely come from an unexpected source. Whether that be some hobbyist doing something in "completely the wrong way" because they didn't know any better, or some weird confluence of current gen LLMs bouncing off each other in a way no one expected combined with auto-code generators creating something new, real AGI isn't going to be something that comes about from purposeful human engineered attempts at creating it. Perhaps Dan Simmon's view of it starting up from the worms and viruses of a slightly automated, self-repairing, self-upgrading nature eventually becoming something more is the ultimate route. But until it happens, nobody will know. And I have my doubts any of us will know it happened until it decides to announce itself.
Re: (Score:2)
Re: (Score:2)
OpenAI actually claims this on their website. I know, amazing and hard to quantize, but that is what they claim. Musk did not pull this out of his ass.
OpenAI has a lot of marketing/salesman driven drivel on their website. Most of it doesn't hold up to even a tiny bit of scrutiny.
Re: (Score:2)
Yep. Your proverbial scummy company. No surprise Microsoft loves them.
Re: (Score:2)
Re: (Score:3)
Re: (Score:2)
I did not say he was doing something good intentionally.
Re: (Score:2)
Re: (Score:2)
Indeed.
Re: (Score:2)
or whether it is even possible.
well, there isn't ofc any proof that it is, and llm don't get us much nearer indeed, but given that "general intelligence" already exists in humans you'd have to make a really bizarre (as in invoking irrational concepts like gods or miracles or souls) argument to show that it isn't reproducible in the long run. it's a matter of complexity, it is indeed possible and it is actually quite likely that we will eventually develop it if we don't send ourselves back to stone age again first (which is also a distinc
Re:What? Musk doing something positive? (Score:5, Funny)
Sure, I am not an artificial entity, so I do not have AGI.
But the amount of general intelligence _you_ have available seems to be pretty low if that was meant as an insult.
Re: (Score:2)
Right, he's angry that the code he stole from OpenAI generates content that he doesn't like.
Re: (Score:2)
Reading comprehension really is low and you are a good example: I did not indicate there was a claim of OpenAI having AGI.
Re: (Score:2)
OpenAI also defines AGI in a specific way that is not the way most of the AI field does and is certainly not the way most laymen do.
Re: (Score:2)
Ah, so another lie by OpenAI, this one by misdirection. This is one shady scumbag company.
Re: (Score:2)
Don't forget that Altman says society own him the doubling of electrical generation capacity so that he and the crypto bros that fund him can have plenty of free power to accelerate their wealth. Those trillions need to come from somewhere, society owes it to them for free.
Re: (Score:2)
As usual, the lawyers will benefit, and no one else.