It was listed at $25 million, later cut to $20 million. Ok, I understand paying a bit of a premium for an iconic property, but... "Even with the price cut, the eight-figure home is about $18 million higher than the median for its surrounding neighborhood".
That's not a premium price, that's 10x over market. Plus, it's on the National Register of Historic Places, meaning whoever buys it is committing to maintaining it according to the listing.
How about the Stahl family donate it to the National Park Service, along with an endowment to fund upkeep? Let a historic property be historic, not a shiny bauble for a billionaire status symbol.
The evidence is a combination of two things: Tristan insinuating they did (note, not claiming but suggesting or implying), and their own statement saying "While unlikely, we cannot rule out that de-identified data derived from their usage of our products helped improve our models ."
The latter is being used by everyone as a sign of guilt, but it's clear that they are making a legally-safe statement since promising a forensic level guarantee that nothing Tristan has every typed into ChatGPT has ever made its way into any training data is a massive claim.
It's also a ridiculous expectation since tons of researchers use ChatGPT or Codex and many have "Use this data to improve" setting on.
And a third thing: OpenAI's thuggish behavior. Threatening a researcher with reputational ruin and demanding one of their competitors be denied proper credit is unacceptable, and we know OpenAI did both (unless Buckmaster is simply lying. which I find unlikely).
>many have "Use this data to improve" setting on.
Quite irrelevant. We are talking about scientific misconduct, not IP law. It would be misconduct for OpenAI to publish without proper reference to prior art even if that prior art had been explicitly committed to the public domain 1,000 years ago.
> And a third thing: OpenAI's thuggish behavior. Threatening a researcher with reputational ruin and demanding one of their competitors be denied proper credit is unacceptable, and we know OpenAI did both (unless Buckmaster is simply lying. which I find unlikely).
yeah, its not "we know" but "Buckmaster said this".
People lie often, not sure why we would presume unconditional innocence of one side in this case.
>It is in that context that I said “it would be simpler if Levent was not an Anthropic employee” because I felt it would be inappropriate for an Anthropic employee to author OpenAI’s work.
>I was confused why one would turn an incredible source for celebration (of their achievements!) into such bickering, which is when I said that I did not understand why one would risk their career [over unfounded accusations].
Maybe Buckmaster exaggerated the threatening tone and maybe Levent doesn't actually deserve authorship. But he wasn't lying.
Come on this is all just pure copium from people who swallowed this hook line and sinker.
The worst thing anyone mustered about this is it was a breach of the cordial way mathematicians treat each other.
OpenAI solved Navier Stokes and that’s all anyone will remember about this no matter how mad people are about it and no matter how much they’re coping that they “stole” it.
Newsflash: team Anthropic was using AI too full of people’s “stolen” content.
Without seeing the full correspondence it’s hard to evaluate for sure, but parent linked to a tweet from the OpenAI employee at the center of the controversy, that’s a primary source you can read and evaluate yourself
Personally I don’t find the tweet a satisfactory explanation of their behavior, it seems like a lot of deflection without directly responding to the specific claims of front-running, and the screen cap of the correspondence doesn’t include all the relevant context. If there was really no bad behavior, why not post the whole thing?
The screen cap of their full correspondence in the Twitter thread, which was the subject of the preceding sentence. The Twitter post has an obviously incomplete fragment of the conversation that doesn’t resolve what the author presents it as resolving, which is the dispute over how the discussion of authorship of Alpöge actually went down
I read Tristan's allegation from https://cims.nyu.edu/~tristanb/statement.pdf, and he didn't make it clear if threats to him were made over email, texts, calls or during personal meetings, so it is not clear if the whole thing was in writing.
He mentioned that they had meetings:
I asked to speak the following week. On Friday, September 4th, I was asked
whether I could meet that day; I again said the following week. At 12:45 on
Sunday, September 6th, I was asked whether I could meet “at any point today.”
Sebastien Bubeck joined. The three of us spoke twice that afternoon. Levent
was not on the calls.
>and their own statement saying "While unlikely, we cannot rule out that de-identified data derived from their usage of our products helped improve our models ."
Legally safe? Forensic level guarantee? Since when do they make those kinds of statements? Talk about moving the goalposts. Those are clear weasel-words from OpenAI, I'm shocked by the people who don't realize this.
Bold of you to think undocumented workers making subsistence wages under a brutal system have access to annual physicals, much less the sort of continuously documented medical history needed to spot and address abnormalities before they become life-threatening emergencies.
It's actually an interesting question there. Because I've been wondering if I buy something in taking ownership of it, whereas if I steal something I am still taking ownership of it without compensation. However if I'm not buying something but merely licensing it the wouldn't that mean that technically I'm not stealing it I'm only violating the terms of the license of it, which is now a civil not criminal matter.
Firstly, you're not "stealing" (committing the crime of theft) by making digital copies of a work. You are infringing copyright. Even a single copy infringes copyright. To not infringe copyright, you need a license for your copying from the copyright holder.
You can still manage to infringe that license even if you went through a normal storefront and paid the copyright holder the amount they asked for. You can do that by doing things with your licensed copy that the copyright holder wrote into your license that you mustn't do. If they sold physical items, they would not be able to apply such adverse terms to your copy.
If you void your own license in that way, you no longer have that license, and so any copying/modification/distribution you're doing is back to being copyright infringement.
Finally, there is such a thing as criminal copyright infringement. That's when you make a significant profit from copyright infringement (for example, getting people to pay you to watch a pay-per-view event that you're illegally redistributing, or even just having adverts that make you money on a website offering unlicensed Nintendo ROMs for emulators)
Yeah, exactly, it's not stealing, it's just ignoring the terms of license. The same thing in spirit as e.g. different corps shitting on GPL. And should be enforced with the same eagerness (lol).
Arguably, piracy has never been theft as "theft" means to deprive an entity of their item/copy and downloading does not destroy the original. The usual counter-argument is that piracy deprives them of a sale, but that is not necessarily true and it doesn't really fit the meaning of "theft".
I mean does it? DRM is cracked within a week or two of launch now. companies are actively removing drm from their games a few months after it comes out because of the cost of licensing it.
AI has accelerated this if anything, and especially the server side stuff. Look at the destiny 2 private servers right now recreating lost content.
Buying in this case is "buying", and is a redefined term inside the legal contract which you choose to agree to when Sony accepts your money for a license to access their published video games, and so, it does not mean buying in the mundane sense of physical goods.
I would suggest this workaround: If you don't accept society's rules about related norms, such as you having a different definition of theft froom society's, you must proactively declare -- in your mind at the minimum -- and accept without two-way agreement on behalf of everyone else whom you steal from, that theft means what you choose to define it as at any moment. Do that, then that logic is solid.
> If you don't accept society's rules about related norms, such as you having a different definition of theft froom society's
Society's "norms" around buying is that you own the thing you buy, which is why we have different words for buy, lease, rent, etc. Society's norms around theft is that you have taken something from someone else and they no longer have access to it.
Sony is the one with the different definition. They're the ones that should proactively declare it. And no, codifying it in legalese and burying it in a EULA isn't declaring it as society would expect.
To add: when Blockbuster existed, we all understood that renting a movie there was a temporary thing and buying the VHS at Best Buy was permanent. That is the expectation one has, not whatever a giant multinational decides it is.
Yes, if your transaction contract redefines very common well understood words like “buy” it honestly should be nullified in the eyes of the law. It’s an inherently deceptive practice.
Want to create new definitions for new words or phrases? Totally fine, but then Sony would need to replace the “buy” button with “blease” or “blent” or whatever they want to call this fake purchasing shit that all software has been playing at for entirely too long.
I disagree that it's "inherently deceptive". That claim deserve justification IMO.
It's an obtuse-feeling abstraction.
> ".. or whatever they want to call this fake purchasing shit that all software has .."
Yep - I agree with the sentiment.
Reality is, we have a set of systems in society capable of over-complicating (from a HUMAN sense) life, which reasonably can be seen to make life unpleasant for the less-enfranchised populations - those without a personal lawyer to review every action they take in life...e.g you and me.
I get it. And the point I failed to make is they did redefine it via a normal process of contract.
A bit of pedantry:
> Sony is the one with the different definition
Yes, agreed.
And to build on that, Sony is using a normal legal vehicle, a contract to make that different definition. And customers of Sony are entering into contract .. which is very normal thing for customers to do when engaging in business relationships with businesses.
> "That is the expectation one has, not whatever a giant multinational decides it is"
Yep.
To reinforce my yep, (tiredly - sorry but I am) - The yoke of bureaucracy and this complex economy we live in also rests upon us, the citizens. It is a burden - and an increasing risk as AI is unleashed into more and more of our systems, digital AND legal / policy / social. (AI has "infinite" memory so it can use circumlocution, taking advantage of overly complex and distracting
Looping back:
I could have said "it suuuuucks, f' the man, man!" as I do hold that view. However we are here, with legal contracts capable of imposing these broad and uncomfortable relationship constraints (eg between customer & business).
SO the question, in my mind, is HOW are we going to understand WHAT we want society to look like, as presently and into the future an INHUMANLY complex set of rules is capable of limiting our expressive existence, leading reasonably to general dissatisfaction in life?
Well, in the UK under the Consumer Rights Act 2015, with goods, you're entitled to refund, a repair, or a replacement. We're not good at class actions in the UK, so it's down to the individual, but if a one time payment for a game that was physical media, isn't a 'good', what is it?
I'll post the google/ai summary, but this mostly it:
Your Main Rights by Timeframe
Within 30 Days: You have the right to reject the item and get a full refund.
After 30 Days and up to 6 Months: You must give the retailer one opportunity to repair or replace the item. If that repair or replacement fails, you are generally entitled to a full refund. The law assumes the fault was there when you bought it unless the retailer proves otherwise.
After 6 Months: You can still request a repair or replacement, but the burden is on you to prove the fault was present when you received the item. You have up to six years (five years in Scotland) to bring a claim for faulty goods in court.
For part of the above sentence ... "it" is existent, not nonexistent.
And about the ".. I have .." part of the comment you wrote, @kstrauser, by using their service a person must agree to their contract.
So, if you, per-se, don't use their service, then you haven't entered into the (existent) contract with Sony!
Here is the segment:
> 10.1. All intellectual property rights subsisting in the Content, including all software, data, services, and other content subsisting in or used in connection with our Services, the Online ID and access to content and hardware used in connection with our Services belong to SIE, its affiliates, and its licensors. Use of the terms "own," "ownership", "purchase," "sale," "sold," "sell," "rent" or "buy" in this Agreement or in connection with the Content does not mean or imply any transfer of ownership of any content, data or software or any intellectual property rights from SIE, its affiliates, or its licensors to any user or third party.
Startups define different classes of stock. The class A shareholders are the founders and investors. Everyone else gets class B shares. The A class shares don't get diluted, and they are inherently worth more anyway.
VCs get preferential shares, not common. Preferential shares have economic rights to protect the investors, but more importantly they usually have extra control rights like veto abilities, board seats, IPO control, or ability to sack the founder (which may even cut out the founder's voting rights by sunsetting their class A common into class B common shares).
Employees get a third tier of stock (e.g. options that convert to non-voting class B common shares).
After IPO the preferential sheets becomes common shares. The dual A class may be removed or have sunset clauses because large public investors prefer one plain common share class.
Not a VC - so take above as written by a student. Founders in zero sense have the same voting control as VCs.
Edit: VCs play the same game over and over again, against different innocent founders. VCs know how to stack everything in their favour - especially using social cues and "norms" that benefit them. My favourite article on this is: https://siliconhillslawyer.com/2019/02/18/relationships-and-...
I got the details about classes of shares and preferential stuff technically wrong, it’s true. I still got the gist of the arrangement correct: VCs and founders get the pizza, everyone else gets the crusts
Pref shares do get diluted, they are however senior to common stock so they get money FIRST if there’s not enough to go around. There is some cap on this and sometimes it’s pretty high. Huge pref overhangs are, indeed, a problem.
Pref shares with a 1x preference are still worth like 10x common stock in early stage companies and it’s common for employed to get fucked by this.
Founders don’t get preferred shares (I think it’s really, really rare). There is founder pref stock, which is somewhat different. It’s common for founders to cash out some shares along the way, though.
Founders that take a pay cut from a high paying job should demand preferential shares to the value they are giving up.
If they were in job where they were saving $50k a year, then after becoming a founder they should be getting $50k worth of preferential shares per year because they are investing that much in the business.
Not that I've actually ever heard of founders getting preferential shares to match their dollars invested.
This is not standard. Normally founders and employees get common stock and investors get preferred stock. Founders may get more stock issued in a round, and VCs/founders can pretty much rework the cap table to their liking if they really want to. The difference in return between founders and employees is down to percentages. Founders get 25-75% where employees get 0.01-1%, maybe a bit more if they're lucky.
> Founders get 25-75% where employees get 0.01-1%, maybe a bit more if they're lucky.
So that's a scam by the founders to the employees, in my book. It's fine, it's just that I am not sure young professionals joining a startup know that.
Said differently, if you join a startup, you should not work too much without compensation, and you should not care about making it super valuable, because you don't benefit from it. If you have a super good idea or realise you have expertise that would make the startup valuable, you should leave and become a founder yourself.
I agree it's a scam. I just wanted to point out the way the scam works, by ownership percentage and new shares issued during new rounds (and cap table shenanigans), not generally through a different class of stock issued to founders vs employees. That's much more rare.
So fundamentally, "everyone else" is scammed. Unless the class A scammers get so, so rich that everyone else gets rich as well. In which case it's still a scam, but the "everyone else" are happy anyway.
Why is that a scam? Nobody ever promises you any specific valuation or fraction of the company. When I joined a company relatively late but well before IPO, some funny number of shares at 12 cents or whatever each did not even enter my calculation any more than "oh and they also give me a free lottery ticket". In my case depending on when one sold after IPO they would have been in the range centered around about compensating for the salary differential I think, but nobody promises you they'd ever be worth more than Monopoly money
> Nobody ever promises you any specific valuation or fraction of the company
Would you mind asking before saying what I have been promised?
Also it feels like you have never been in a startup. The whole language of growth everywhere, the "billion-dollar startup", the "becoming a unicorn", this is all suggesting that "you're part of it and it matters to you if it becomes a unicorn". But it doesn't, really. Because you get diluted.
I was given a very large number of very low value shares. The company was about 60 people iirc and I think the hr guy said well we won't IPO soon but when we do, these shares might be worth a lot! Companies often aim for shares to be worth 10, 20, 50 dollars at IPO! Something like that. But it's obviously just vague pep talk. They also sometimes say everyone is there to make the world a better place...
It would be a scam if they promised you 0.2% of the company but then it was diluted to 0.1%. and nobody prevents you from asking i think. Otherwise it's no more a scam than a lottery ticket commercial showing the guy who won a Ferrari.
It is a scam to me because they imply "if the founders get rich, we all get rich" because "we're on the same boat". And it's not the same boat at all: the founders may get rich, the employees most likely not.
It’s strange, every startup offer either is obviously a horrible scam or comes from a place of fairness and is sold like a total scam.
I guess being honest brings about too many opportunities for people who don’t understand the finances to make (or be perceived to make ) promises they can’t keep. So you might as well just get into a race to present the most ridiculous stuff possible.
> But it doesn't, really. Because you get diluted.
At this day and age, if you don't understand dilution before you join, it's entirely on you.
This isn't a new concept - it was the case decades ago. Even when I left school over 15 years ago, the standard advice when trying to get a job with a startup was "Get a good salary and value the equity at zero."
And class A vs class B isn't even a rich vs everyone else thing. I have class A shares in an LLC, where even the (richer) founders are class B. The operating agreement is that we class A folks are "guaranteed" a fixed rate of return on our investment, and the class B folks don't get anything unless we get at least that rate of return. This is very normal in that industry.
> At this day and age, if you don't understand dilution before you join, it's entirely on you.
I don't know what to tell you. Young graduates get an offer to work at a startup, nobody tells them how it works. They are just excited, as I was. And they don't think about "what happens if the startup is successful" because they do know it probably won't be.
And when the startup is successful (happened to me) is when they realise that they got scammed. But all they can do is see their founders become rich and tell everyone why THEY deserve it because it was THEIR idea and THEY are the best.
> the standard advice when trying to get a job with a startup was "Get a good salary and value the equity at zero."
That does not say AT ALL that the founder gets rich when you get nothing. It says "be careful, most startups fail, so make sure you get a salary". Usually that salary is subpar.
“The long term sustainability of the commons depends on an AI ecosystem that produces shared benefits”
Do any of the big players in the AI business even have this on their radar? Other than vague “we’ll eventually get AGI and it will cure cancer and fix climate change and everything”?
That's not a premium price, that's 10x over market. Plus, it's on the National Register of Historic Places, meaning whoever buys it is committing to maintaining it according to the listing.
How about the Stahl family donate it to the National Park Service, along with an endowment to fund upkeep? Let a historic property be historic, not a shiny bauble for a billionaire status symbol.
reply