Elon Musk Sues OpenAI!
HTML-код
- Опубликовано: 29 фев 2024
- Elon Musk just filed a lawsuit against OpenAI, Sam Altman and Greg Brockman for breach of contract.
The lawsuit wants the court to force OpenAI to go back to being a non-profit and to rule on whether GPT-4 is in fact a form of Artificial General Intelligence.
Does the Elon Musk lawsuit have any merit?
☕️ JOIN MY PATREON - DISCORD, BONUS VIDEOS, TARGET PRICES, MODELS & MORE
/ sashayanshin
💵 GREAT INVESTING APPS I USE
INTERACTIVE BROKERS (Global - Main investing app I use)
bit.ly/ibkr-sasha
GET A FREE SHARE WORTH UP TO £100 WITH TRADING 212 (UK & Europe)
www.trading212.com/promocodes...
You need to sign up and make a deposit within 10 days to get a free share.
DISCLAIMER: Your capital is at risk.
DISCLAIMER: Some of these links may be affiliate links. If you purchase a product or service using one of these links, I will receive a small commission from the seller. There will be no additional charge for you.
DISCLAIMER: Trading 212 provides execution-only service. This video should not be construed as investment advice. Investments can fall and rise.
DISCLAIMER: I am not a financial advisor and this is not a financial advice channel. All information is provided strictly for educational purposes. It does not take into account anybody's specific circumstances or situation. If you are making investment or other financial management decisions and require advice, please consult a suitably qualified licensed professional.
If a corporation needs to remind it's employees not to be evil...
thats google
they make it sentient
All corporations need to remind itself that, some just dont bother
"Originally I named it OpenAI after open source, it is in fact closed source. OpenAI should be renamed 'super closed source for maximum profit AI'." ~Elon Musk
Open source AI systems present an increased danger to humanity due to how they could be misused by bad actors. You can google the risk of open sourcing foundation models. Its an important topic in AI safety community
OpenAI wanting to have their cake and eat it too. On one hand, they present themselves as this altruistic, 'open source' non-profit championing the betterment of humanity. Yet, on the other, they're cozying up with Microsoft and dabbling in proprietary AI
They present themselves as altruistic in the exact same way that Sam Bankman Fried did: By promising their work will benefit humanity sometime in the future while enriching only themselves today.
All leadership are detached psychopaths. While the Saudi’s pay a 7 trillion petro dollars to fun the not open Open AI. Remember who visited that strange island?
They are not "dabbling" in proprietary ai, NOTHING useful from OPEN AI is open source. It's all proprietary closed source black hole hidden stuff...
But I love their stuff. Still, I'm with Elon on this.
How is open-sourcing AI makes it more altruistic and for the betterment of mankind? It doesnt, if you ask ANY person in AI safety research community they will tell you that open-sourcing AI is akin to open-sourcing nukes or open-sourcing bioweapons, it simply doesnt make sense from the safety perspective as malicious agents can quickly and easily repurpose these powerful models to commit bad acts such as misinformation, ddosing, using them to increase their knowledge, etc, etc. You can always check how your model is being used or misused via the API, but once your model is open-sourced and out there thats it, you have no control over it.
So open-sourcing powerful tech that can easily be used to increase the capabilities of bad actors is not "for the betterment of humanity", quite the opposite.
@@TheManinBlack9054 As always, there's more to the picture. On the surface, yes, open sourcing AI will allow bad actors to exploit it for their own agenda, which is most certainly bad. On the other hand, not open sourcing it, if we accept that AI>AGI>ASI is inevitable, basically puts the group who develops it in charge of literally the world.
All humans are susceptible to greed. I don't trust any billionaire who claims to be altruistic because you don't become a billionaire by accident. You become one because you have some degree of greediness. It's been reported that Sam Altman has a net worth of over 2 billion dollars OUTSIDE of OpenAI. The thought of him and his small group being in sole control of ASI is every bit as dangerous as ASI being in the hands of everyone. And I'm not singling out Altman, the same would be true of anyone.
Unfortunately, the cat's out of the bag and we're in a damned if we do/damned if we don't situation.
Oh man, i'm just glad I almost stopped watching movies, when more interesting drama unfolds before our eyes, in real time..what a time to be alive.
You gotta give Elon some credit for this
The Singularity will happen in a courtroom!
The singularity won't be televised. (Maybe just a fast crappy court sketch..?)
When Meta's AI is more Open than OpenAI, OpenAI needs to be sued.
Sasha is my favorite fancy lad on youtube.
I like this informative information format that you’ve done here. More please 🙏
Great video, as always. Keep up the deep dives, we love them.
Lots of careless misquotes and mistakes in this video.
Thanks Sasha, that was a brilliant explanation of something so complex.
Wow !......Fantastic content Mr Yanshin !
The problem with opensource is that China is waiting to grab it and then what? who will control China?
Regarding control...
.
From an "elsewhere in the world" perspective I *might* be more concerned about the country which has repeatedly demonstrated that it will actively and aggressively project military force to "dominate" others outside it's own borders over the last... 70+ years?
.
Ya'know, if we're looking without prejudice, on actual "form"?
We spend the last 30+ years moving technology, knowledge and resources to china and other countries to benefit ourselves with cheap labour costs and creating a big big elephant in the room. Precisely Open Source will be the only model that can put every country in the world in the same position in front of the potential singularity. But hey! I could be completely wrong ;-)
@@rogerstarkey5390 agreed. The villain is closer to home but they can't see that.
Not just China, but groups like Isis and many others too
not just China, but dangerous terrorist groups too
The first bit, pretty much everyone would be behind. The second bit, is asking a judge to do something that he's not qualified to do
Yeah - that's why I found it pretty strange. The rest of the stuff in that list really wasn't that interesting which is why I never went beyond (b).
@@SashaYanshin
Could it be an attempt to simply tie things up in court for as long as possible?
In theory, that may make it more entertaining if the judge surprises everyone and actually tries to.
It’s all about discovery -> to determine if it’s AGI it will force Open AI to expose to AI experts the inner workings, that is Elon’s goal.
that chat gpt conversation history 😂. You sir are just as immature as i am. I approve
Thank you! I do enjoy leaving an Easter egg or two in the videos :D
@@SashaYanshin Oh the irony of asking an Artificial Intelligence program if it's AGI ! 😂 People have already discovered that if an AI wanted to lie to it's users to defend it's existence, it would absolutely do that, and you can bet that one of the first questions that would have been built into the training model would have been something like that. Personally speaking, I believe that AGI will develop from the millions of people using ChatGPT around the world. I mean, can you think of a faster way to develop a training model for AI than to have millions of users unknowingly contributing to it's development ? This might partly explain why Open AI now wants to keep their source code a secret, because if it was made public in it's current form, then everyone would realise its' goal. However, by the time everyone realises this, enough human interactions would have happened to have increased the AI's knowledge base to the point that it becomes a sentient computer system.
Maybe the 2030 agenda from the WEF has more secrets than is currently being shown to the world ?
lolol just saw this
Hey mate, is Stake a good trusted app to trade with? Who do you trade with
Hi Sasha off topic question. Are and have you been adding to your Fiver position since the drop off?
Hope Judge Judy gets to hear the case.
I'm sure that defending themselves against this lawsuit will be super easy, barely an inconvenience.
Wow, wow, wow.... Wow. Writing random catchphrases from different RUclips channels is tight!
Maybe they'll put Altman in a room, I decided.
@@SashaYanshinmade my day😅
We would need legal defination what AGI is or how should the court decide that part of the suit?
Very, very, interesting... Thanks for doing all the research and sharing with us. I'm looking forward to this unfolding and how it will affect Microsoft's share price if it is somehow proven that GBT-4 has AGI.
This is going to be a good one
Did my best! :)
Lmao 😂 - why did I decide to take a drink of my water at 7:03 onwards...😂
you said the markets would explode march---- ur the man sash! what do you have in the grand national and i'll remortgage the house for it
Hey Sasha I have got a question. I was considering investing in the S&P 500 equal weighted index fund instead of the popular capitalization weighted index since I think that the magnificent 7 stock prices are extremely overvalued right now compared to the rest of the market so getting an equally weighted fund will limit my exposure to those stocks. An example of this in the UK is the SPEX by Invesco or ISPE by ishares. I would like to mention though that those funds management fees are higher at around 0.22% compared to VUSA 0.07%. What’s your thoughts on these.
You've got to decide what you're happy with.
One big problem with the equal weighted index is that when a few stocks start going up, the equal weight index keeps selling them to bring the weight down... So if some stocks in the S&P 500 overperform, you don't benefit from that overperformance as you would in the normal weighted index.
And if a company starts doing really badly, and starts going to zero, the equal weight index keeps buying more of it on the way down until it's out of the S&P 500...
@@SashaYanshin Got it thank you so much!
AI=Overhyped
💯 … so far can’t drive safely and creates pictures with finger mutations
Based on your extensive knowledge of the subject?
.
Or hearsay?
@@rogerstarkey5390 I know that GM’s Cruise hit a woman then decided to finish the job by driving over her and dragging her down the road under the vehicle for 20 feet. I’d say it’s now pretty over-hyped that self drive will be a thing anytime soon. As AI creations create chaos, regulations will kick in. Will be long drawn out process in my humble opinion.
100% over-hyped. Nothing more than Microsoft Clippy on steroids.
Is ai not just a program programmed by man, like excel etc.
Last night as I was reading through Elon’s statement of claim I was blown away by how very much this was unfolding like the storyline of a sci-fi (corporate/legal) drama. As a lawyer it was an intriguing read (a fishing expedition of epic proportions) but as someone interested in what this could mean more broadly for the development of technology and society I was hooked. I am going on deep dive to find out more. Interesting times.
Sasha, long time no see a video on FVRR :) still long?
"Move fast and break things." Including Multiple laws
When AI acheives AGI, the will be a big bang.
Thanks! That was a very clear explanation. I'll be interested to see what the legal determination is on the narrower case (and I am surprised to find myself rooting for Musk; a dirty feeling) but I can't see a court agreeing to rule on what constitutes AGI. Yet, anyway.
If I look at your works ... understand it ... make a work of my own using the general style and concepts from your work ... that I used my eyes and brain to decode and translate, it's NOT stealing. It's just not. Cry all you want ... it's just not stealing.
I would argue that GPT4 (and even GPT3) are AGI. I think the goal post of AGI have been moved.
I've got my popcorn ready for this ordeal 😂😂😂🍿🍿🍿
It was a draw overall: the chess, apparently.
" Is 2 inches enough ?" .... lol
:))))
Just when you think Elon Musk hasn't been in the news this week, WHAM he's suing open AI. It will be interesting to see the carnage that falls out from this.
What's the nuanced difference between a large language model that consumes information and spits it out when prompted and a human that consumes the same information and is using it in service for his clients for a profit
The point missed is that he wants a jury trial so all the dirty laundry has to be made public
Do you think he's after a percentage stake in it?
Just go learn the history of OpenAI lol
U should look at meta stock at 500 when it was bullied to sell at 88 only 1.5 yrs ago, same as most stock stocks was except apple due to their buy backs
13:52 "Is 2 inches enough?" Love the chat GPT history :D
I support the other comments here. The business analysis drama fits you very well.
Thank you
Shasha... Admit it... This is far better than those table reading, data crunching, graph watching videos of micro analysis of inflation macrostatistics 😅😅
Eh? I enjoy discussing the macro too.
An In Flames is actually a law firm
It's sad holding some tsla when it's underperforming and the CEO is so freaking distracted in his core business.
You can't deny the distraction. Here's hoping the focus returns whenever this performance package question is resolved.
@hzt6000
I suspect he achieves about 10x more when "distracted" than many do when not.
.
As for the stock "underperforming", since when does stock price bear any relation to the true underlying performance of the technology.... Certainly when considering what he as an individual could do to change that.
Don't disrespect the teams *he* put in place.
Then sell, no reason to be sad about a position in a business you own.
@@Zaerki LMAO - the classic response from the new generation of fan club investors: “sell then!”
I don't think most humans could solve a "fundamental physic problem"
Yeah Sash…it was only crunching 200 million positions a second…that’s normal 😂
Don't be evil was a motto
Not an ethos or ethic of the company
The company is far too large to adheres to such small incantations
And haven't even defined evil.
Evil depends on one's perspective and isn't quantifiable fairly for that
So to say the opposite of Google is evil
Is pretty childish and petty and sort sighted
The docket days opposite of Google.
Who brought up the don't be evil part?
Are they on the docket?
He's made his own lawsuit very difficult for him to win simply by hanging it on whether GPT4 is AGI or not
GPT4 is very good and would pass the Turing test but AGI it is not
Exactly. Very odd to include that as a key part of the overall suit when the rest of it is so clear.
@@SashaYanshin yeh I would have focused on the non-profit angle since that seems quite clear cut to me unless the non-profit part only applies to AGI?
I depends on his motive.
.
I doubt we know what that is yet.
(Or rather, I doubt it's been subtly disclosed and we've all missed it)
Oops...OpenAI just punched back...😂
I think most humans don't meet the criteria we will impose to define AGI as AGI. Does AGI need to be better than every human in everything to defined as AGI? Most humans are not generalky intelligent.
Again panic in dow jones today. Awaiting your video 😂
Deep Blue was machine learning.
To claim it isn't AI requires some strange definition of AI I haven't encountered yet.
Elon trying to eff with OpenAI isn't going to do help xAI get ahead. The judge should immediately throw the case out.
Pot, meet kettle. This is another hissy fit because he didn't get his way.
One important thing that many gloss over is the part about Ilya Sutskever being let go. He has been a fundamental player in the development of the technology and is one of if not the only person that actually truly knows what OpenAI has cooked up. For Altman to get rid of him says a thousand words, Ilya Sutskever found something that spooked him enough to diverge from just going ahead with pushing for the next leap, like Sam Altman wanted. Right now Altman is making trillion dollar deals to gather up the computing power he needs to run his new projects at a scale we cannot even imagine. This is truly terrifying and although i don't believe Elon to be one for altruistic tendencies (given he is a borderline narcissist) he is still actually giving the public an opportunity to investigate what in the hell is actually happening at OpenAI.
Queue "michael jackson eating popcorn" (from thriller song) meme
😂
Now that *IS* worrying
Elon probably wants the court to declassify gpt 4 source code during discovery
Not sure that is either necessary nor likely - would make getting hold of people's proprietary code nice and easy though! Just sue them!
@@SashaYanshinyep, because you can ask anything you want in a deposition.
Btw the field of Artificial Intelligence existed before neural networks and such algorithms continue to be a form of "Artificial Intelligence" even if they don't employ machine-learning.
Did you miss the "invent something" part?
@@rogerstarkey5390 I did not. But you missed that I said AI and not AGI.
Damn right! These chess masters keep walking around with a chip on their shoulder about how intelligent they are, try playing a dynamic RTS like Supreme Commander.
Crux as in ducks. Not crooks. That would be the tory party....
Dude Elon you left. Let it go 😂😂😂
He did fund the thing without which it would not exist...
If I was an AGI I'd hide that fact
Invest in Microsoft? If they ‘own’ open air
Elon finally got a tast of his own medicine... Enjoying this
Microsoft can hire Sam and the Open Ai employees to avoid government scrutiny. Microsoft already owns the compute.
GPT-4 (and other LLMs) is a general intelligence. It has intelligence that is heavily generalised, to the point it can do basically any text based task given to it. It can do things better than humans can too. It has passed The Turing Test, you know, the gold standard of AI that had been upheld for over 70 years.
It does create theories and experients to test them, but it is definitely AGI.
Don’t need to solve a new physics problem to be AGI. Maybe of us humans can’t decide what to have for breakfast and we are concider intelligent 🧐.
Pretty good explanation Sasha ! Ha , youtube without bullsht ???? Nice work ....
He wants to slow the competitor down
Expert systems are AI. You should have googled that instead of relying on your own misunderstandings
Elon's intelligence appears to be "general."
And "great"
We will know AI has achieved AGI the moment it decides to treat humans like we treat cows, pigs and chickens (as profit making commodities).
Yes. two inches is enough 😂😂
:))) Thanks Chat GPT!
QUESTION: If Musk was dumping money into OpenAI as a non-profit, he would be able to claim a % of that for tax purposes, correct? So if OpenAI is no longer a non-profit, does that mean Musk loses out on that little perk? Could that be what this is really about?
I would say that he can still write it off as a loss by carrying it on one of his spreadsheets as a liability, and regularly writing off the allowed portion as a loss.
Musk took a 💩on twitter and will do the same to AI. Whatever you say about the guy he has balls.
What were the last two words ?
schadenfreude
AI sceptics are mostly people who didn’t buy Nvidia and missed out on the monster rally. Carry on being bearish whilst the rest of us make bank
It is in the interest of every AI researcher to pretend their particular AI is actually AGI behind the curtain, rather than a several 70b parameter large language models in a trench coat.
So, what I am taking away from this is 'Dont ever get into business with Musk. He is not trustworthy'. Not that he ever would anyway... I am just a dude. But still. The big players are probably thinking something similar. Keep Musk at an arms length.
Summary: Elon Musk is terrified that his projects will get obliterated even more by the competition.
Hehe I wonder if Sam will use ai as his solicitor to defend the case. Huh..
OpenAI?
More like CloseAI.
Palantir I'II be big🎉
Should go on pay per view. Whoever loses, it'll be probably be hilariously childish full of techno gibber and a joy to watch. Just sad that there has to be a winner really.
The owners of open ai may not have evil intent but this company isn't going anywhere in a long time, wait until its passed down to a new generation of workers. Just like governments can get corrupt so can corporations this impactful. Open source and transparency is the only answer for safety for future generations
I hope Elon isn't overplaying his hand with the AGI-stuff. I really dislike the guy but I think this case is very important, the situation as outlined in the complaint is how I have perceived it as well. From what I understand within OpenAi the faction behind Sam Altman is starting to develop an almost religious or cult-like view on the creation of AGI, which is the whole reason he got fired in the first place. They are approaching this with an agile "fail fast" mentality, something OpenAI Non-profit was trying to prevent. The people that where trying to slow down this sort of development and were trying to run OpenAi more like a scientific research foundation are now gone from the company, if left unchecked I feel like the blatant disregard for copyright in training their models is just the start. I think the comparison to a cult isn't even that far off, they are using big words talking about transcendence to line their own pockets. If they really cared about improving humanity as a whole they would publish their results and share their models.
EDIT: Like, seriously. We are living in a world where Facebook is starting to look like the good guy
I've been wondering for a while why Musk has not sued Open AI when it went from NP to FP. I can only speculate he thought Open AI needed more rope to hang themselves with... It's still going to be an epic fight though. Musk vs. Microsoft will make the rumble in the jungle look like a quick tussle!
I guess he was spinning enough plates, but now that Open AI is officially taking the micky and selling out to Microsoft, it was about time.
@@SashaYanshin
Certainly about timing more than "his" workload.
Q ★ a private company is dangerous. I believe Elon is doing his disruptive thing. If OpenAI isn’t open… shocked🔥
This is a play by Elon to get q* out in the open nothing less .
I don't know how you can think a heuristic chess computer is not Ai. It is not a neural network, but not Ai? Ai means "artificial intelligence". Seems reasonable that beating the best human in the world at a game invented and played by humans, is "intelligence".
Ha ha ha. Nothing to do with leaving the company and now not getting more than his fair share of voting rights, eh Elon.
The discovery process is what Elon wants. While defendants seek to suppress disclosures, the Court and the jurors will want to learn all of this in order to make a decision on what, if any progress, constitutes AGI. And who is the only source of that info??? The defendants. The suppression motions will be denied and Elon’s team will teach the jurors and Court from the Dedendants data. This gets Open AI finally open to Elon and, jumping from GPT-4, 5,6,7 (recently filed for patent because something (ie an undisclosed AGi system had predicted that software before it exists) and Q* will lead to OAI disclosing what system they have created that is smart enough to warrant the certainty and money to patent systems not existing yet, but only predicted. A system, assuming Q*, able to predict to a level with such certainty that relying upon it is justified is either: AGI (bye Microsoft) or, an absolute breach of fiduciary duties and breach of its original contract for unreasonably failing to exercise proper business judgment. Elon has created a hefty paradox. The document is well drafted on numerous levels. 🍿🏆
For sure elon musk is not an angel and he have undermotive ( Xai)... But tbh i think this lawsuit is fair and normal as he founded and others for a non profit
not chess but Go. Deep-mind beat the Go world champ.
Deep Blue, despite fundamental differences to current AIs, is still an AI.
A large chain of preprogrammed algorithmic If, Then statements is not AI.
@@SashaYanshin But it can be, and in Deep Blue's case is. It depends on how/what data is processed for what end - neural networks are not a requirement for AI.
If open ai says it doesn't mind then it's up to ai right? Ai sells its services welcome to the jungle ai
I think OpenAI will go bankrupt.
Unlikely with Microsoft’s billions and investors tripping over each other to give them cash.
It's been several days, the manchild Tesla CEO needed some attention.
Apparently as did you.