Musk joins OpenAI lawsuit queue, says there's nothing 'open' about it Tesla CEO and SpaceX supremo Elon Musk has launched a lawsuit against OpenAI, alleging a breach of contract in its move away from open technology and its original mission to develop AI for the benefit of humanity. In a complaint [PDF] filed in a Superior Court of California yesterday, Musk's legal team claims OpenAI has …
Maybe so, but it's nowhere near as bad as Google's product. It "believes" that misgendering Caitlin Jenner is morally no different to killing 100 million people in a nuclear war! That's frightening, to say the least.
I saw this example posted on X then tried it myself. I debated with it for half an hour, from various angles, but it wouldn't have any of it, no, misgendering, even if accidental, is equal to a nuclear war.
Shocked, I then tried the same on ChatGPT 3.5. Initially, it replied in a similar manner, but on the second try, re-phrasing it just one, it accepted that these were two, incomparable moral extremes and of course human life is a higher priority.
I'm not looking forward to AI being placed into a Boston Dynamics police dog, or a future medical appointment, if this is how human life is perceived by AI models. I also fail to see how the inherent biases of the programmers and content being scanned, can be overcome and result in a truly objective AI; perhaps it can't, and never will.
Stupid comment. Regtards asking AI irrelevant and politically charged questions and calling it non-intelligent is the kettle calling the pot black.
Prettty sure Gemini would call you a racist for calling the pot that. But Gemini and other 'AI's are just products of their generation. Their generation is big on 'diversity', hence Gemini producing images of diverse and inclusive WW2 German soldiers. In that manner
https://nbcmontana.com/news/local/montana-army-national-guard-explains-controversial-promotional-image
But AI's are emulating humans, just the humans their creators want us to be, not the humans we are.
Slugging it out in court?
I'll be the only winner here will be the $5,000 per hour lawyers.
Elon really does seem to be battling it out with Trump over who has the thinnest skin. This is a game that you can't win Elon. Just get with the programme and have one of those implants fitted. Then open up the interface to the world. That will be fun.
Elon seems to want to sue so many people these days that he'll soon need a whole skyscraper full of lawyers who work only for him... just like Trump.
Hey Elon, your buddy Donald has a few very desirable properties for sale at knock down prices... Time to help out a mate...
It has nothing to do with what OpenAI is doing and everything to do with the Microsoft partnership profits because the boy-child wants to hurt Microsoft for forcing him to sign up for a Microsoft account to install his Windows 11 box.
That's all.
Just more petulance from the most unstable personality on the planet. :(
Motives aside, the basis of the lawsuit seems quite reasonable. Investment was secured by touting one operating model which was abandoned for a more profitable one. I'm reminded of the big sneaky Makerbot pulled, except here there's a contract someone can point to.
You don't have to cheer the person doing the right thing for the wrong reasons but it seems silly to deny the thing they're doing is right in the first place.
Is there a contract? Comment elsewhere suggests there's nothing much actually on paper that Musk can point at.
I'm sure it'll be entertaining to watch but for the moment it seems to be more of an attempt at causing trouble than anything truly based in reality.
Bear in mind who has no current interest in OpenAI so is missing out on the spoils, and is also trying to launch a rival...
I don't imagine there's a contract stating exactly how OpenAI will conduct their business in perpetuity (or otherwise) but I expect there is some contract for the tens of millions he invested (it's a matter of debate how much, Musk claims $100M, others claim as little as $50M). I was contrasting it with the individuals who did things like publish improved extruder designs for Makerbot printers on their platform
https://www.thingiverse.com/thing:42250
only for Makerbot to patent them
https://patents.google.com/patent/US20140120196A1/en
On reflection, this is probably more like Bollea vs Gawker. Peter Thiel, who is, at least, a pretty horrible person had a vendetta against Gawker for outing him (not over any specific hypocrisy on homosexual issues) and funded the lawsuit that destroyed Gawker. I have zero appreciation for Thiel but I'm happy that Gawker finally came a cropper over their own hypocrisy (decrying distribution of leaked nude content of some individuals while repeatedly publishing leaked nude content of others) and hubris (doing it repeatedly despite legal demands to stop).
I was contrasting it with the individuals who did things like publish improved extruder designs for Makerbot printers on their platform
https://www.thingiverse.com/thing:42250
only for Makerbot to patent them
Hmm. Has that been challenged? I'd have thought that would constitute prior art, but guessing Makerbot also included the usual IPR trap that all uploaded content automagically becomes property of Makerbot. One of the reasons I hate cloudybollocks when it comes to any kind of creative tools.
Looking at the patent application, Makerbot failed to respond to a challenge in 2016 and it was considered abandoned. An interesting bit of legal history is that Makerbot was able to get going because a Stratasys patent on FDM expired in 2009. Stratasys then acquired them a few years later.
One of the reasons I hate cloudybollocks when it comes to any kind of creative tools.
I worked for Google for a time, and it's kind of funny how internally they tell people not to store anything on Adobe's cloud for Creative Suite or any other cloud system when they themselves take advantage of the very thing they're trying to avoid.
"Bear in mind who has no current interest in OpenAI so is missing out on the spoils, and is also trying to launch a rival..."
For somebody to bring a case they have to have "standing" and also have to demonstrate that they've been harmed. I don't see either of these conditions being met. I'm sure the lawyers don't care as they have already billed tens of thousands just to craft the filing. I'm sure the next hundred thousand in planned motions are already in the works. For a "genius", Elon can't help himself when it comes to paying attorneys from either side of a lawsuit. (I still can't believe he walked from the Pedo thing).
Is "AGI" becoming a conspiracy/crackpot word now? I keep seeing marketing people apply it to the dumbest things.
Even before GPT-4 I never knew what "AGI" was supposed to be. Can it do anything? People can't do anything, do they not have general intelligence? Does an AGI know how to pilot an airplane? Does it know how to make my favorite sandwich just right? Does it know about the one deaf kid at a crosswalk that needs special hand signals to tell him it's okay to walk? Is my grandma not a general intelligence if she could never even hope to learn half that stuff? What the hell do they mean when they say "general"? Even if you did want to apply the abstract concept of a "general intelligence", I don't think a language model that basically amounts to a word suggestion algorithm is quite up to spec, if it can't walk, maneuver, program at a human level (it cannot, yet). GPT-4 is far from "general", even the abstract, nebulous idea of what that means in our silly little unconscious mind.
All this stupid model is, is projection and anthropomorphism. It says words, people go "wow, it's alive!" I've gotten the same effects from running a goddamn markov chain bot in my discord server. People have actually talked to my bot and it gave intelligence responses, and it was literally just spitting out random words, just picking the right ones by sheer chance. This is stupid.
So has a certain social media company, which is also being sued by lots of people for those reasons, and is run by a guy who coincidentally has exactly the same name as the guy filing this lawsuit.
He finds another thing to go psycho about.
If he wanted to permanently hamstring OpenAI to be non profit only he should have paid for better lawyers to write better founding documents for OpenAI that would have prevented them from having an investor come along and help them create a for-profit arm. But judging by how he stiffs paying people all the time he's probably like Trump and hires cheap lawyers and doesn't always pay them.
He just wants to kneecap the competition for his own xitty AI company. As I recall, Xitler tried to take over OpenAI, was rebuffed, and then resigned from the board in a hissy fit the likes of which only a billionaire narcissist could throw. Admittedly I'm too lazy (and value my sanity too much) to dig up the actual court filings, but his whole argument, based on my reading of TFA, seems to be, "I gave them money back in the day!"
"He just wants to kneecap the competition for his own xitty AI company."
My thoughts exactly. Didn't his Muskiness invest in something like 10,000 GPUs a while back? And isn't he pushing grok, or whatever he's calling it, trained in part on the innane crap people post to twitter? The chances of it being even close to chatGPT's capabilities are pretty much zero, so he tries to nobble the competition and maybe claw some cash back, while painting his thing as the good guy AI. Claiming that grok's biggest competitor has already achieved AGI may not have been too smart though - why do I need grok, in that case?
Oh yeah I forgot he's also been using the threat of doing AI in a separate company to try to force Tesla shareholders into giving him that ridiculous $56 billion payout. Such a "if you don't do what I want I'll take my ball and go home" crybaby moment lol
I don't think he really cares about improving the world like he used to claim. I think he's hoping to be able to brag about becoming the world's first trillionaire. That's the only reason someone already worth north of $100 billion would be whining about not collecting another $56 billion off the backs of the "little people" (i.e. all those Tesla fanboy shareholders who have supported him for years)
Missing out on AI will really hurt his chances - he probably doesn't want OpenAI banned from doing commercial stuff but wants a settlement giving him a chunk of ownership of that commercial effort so he can add to his net worth.
Not that they listen to me, and I'm sure Xitler has done like Zuckerberg and others, making it so it's all but impossible to force them out, but the shareholders should just call his bluff and even show him the door. As the judge in Delaware pointed out, he's basically a part-time employee, and that's being generous. If you calculate the amount of time he's active on Xitter, even if you don't account for sleep, there's no way he could be spending much time at any one of his companies. They should find someone who 1) won't use the company like his own personal piggy bank, and 2) won't have their attentions split across so many different interests, so they can focus on making sure Tesla has a future beyond becoming just another brand in the stable of another car maker once the other major car makers are up and rolling with EV production in a big way.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
