OpenAI Sued for Fraud Allegations
webapps.sftc.org
webapps.sftc.org
Just see the full list of defendants:
SAM ALTMAN, an individual; REID HOFFMAN, an individual; GREG BROCKMAN, an individual; ILYA SUTSKEVER, an individual; WOJCIECH ZAREMBA, an individual; JOHN SCHULMAN, an individual; TASHA MCCAULEY, an individual; BRIAN CHESKY, an individual; ADAM D’ANGELO, an individual; ANDREJ KARPATHY, an individual; PAMELA VAGATA, an individual; TREVOR BLACKWELL, an individual; BRAD LIGHTCAP, an individual; CHRIS CLARK, an individual; JESSICA LIVINGSTON, an individual; WILL HURD, an individual; HELEN TONER, an individual; TOMER KAFTAN, an individual; MICROSOFT CORPORATION, a corporation; MATTHEW BROWN COMPANIES, a company; BEDROCK CAPITAL PARTNERS I, L.P., a limited partnership; SEQUOIA CAPITAL U.S. VENTURE 2010 PARTNERS FUND, L.P., a limited partnership; ANDREESSON HOROWITZ ANGEL FUND I, LLC, a limited liability company; TIGER GLOBAL MANAGEMENT, LLC, a limited liability company; KHOSLA VENTURES, LLC, a limited liability company; HOFFMAN KOFMAN FOUNDATION; OPEN RESEARCH LAB, INC., a nonprofit corporation; CASETEXT, INC., a stock corporation; INSTACART, a corporation; ZAPIER, INC., a corporation; and DOES 1 through 10, inclusive,
1. https://www.cnn.com/2023/01/31/politics/student-loan-supreme...
That doesn't make sense and is completely backwards. The allegedly damaged party is society, with the major shareholders as defendants.
As a pro se litigant this probably won't go far. But the directionality of the suit seems correct.
In some ways it's a more powerful result if plaintiff prevails though, as it dramatically increases the amount of legal stakeholders in the activity of OpenAI... “You filed as a non-profit, now if you start acting like a for-profit, anyone who is systematically excluded from your business dealings can sue you.”
They can select most at risk populations, etc.
The case is for material misrepesentation to benefit OpenAI shareholders at the expense of the identified intended beneficiaries of (and donors to) the OpenAI nonprofit. Why would the beneficiaries of the fraud challenge it?
> or the FTC, not some random dude from Texas.
The law often creates multiple different parties with a legitimate and enforceable interewt in the same requirement; while the plaintiff here doesn’t have the strongest standing claim in the world, it is not diminished for thr fact that some of the allegations create much stronger claims for, say, the IRS and state tax authorities or donors to the nonprofit (particularly the accusations of the nonprofit being fraudulently organized and operated for self-dealing for the commercial benefit of certain of its organizers.)
Even if OpenAI were committing tax fraud in this way, they wouldn’t be required to give the plaintiff free stuff. That’s a bit like complaining the Met Museum’s gift shop won’t give me a free t-shirt. If the museum were separately breaking the law, they still aren’t obligated to give me a free shirt.
There are several different arguments in the complaint, but the most dangerous to OpenAI seems to be the accusation of principals using the nonprofit for commercial self-dealing which, notwithstanding that this is a pro se filing with a bunch of tangential conspiracy theories in its initial narrative seems both adequately pled and quite likely to be true from the information publicly available.
The big problem for the plaintiff is the theory of standing the complaint relies on, which while reasonably well-pled seems to be something the courts are likely to be leary of as it would grant open season on broad-claimed-benefit nonprofits.
They are calling fraud because OAI was mixing nonprofit and capped profit to avoid taxes and the likes. They also made claims about unfair representations coming from ChatGPT and the executive composition of OAI, which apparently looks like insider trading and information manipulation. They alleged that OAI participated in election interference during 2020. They also said OAI was recklessly making advances in disruptive tech faster than legislators can make laws for them.
Basically, a prime example of a vulture coming in to sue anything that looks like it can give them some money. This is part of the reason why every company need a team of lawyers to deal with this perversion of justice.
The way I see it, the biggest losers are the average consumers. If you got skin in the game and you know how to automate the OpenAI API at scale - you're winning big. Google can't keep up with what is real or what isn't if you actually put in the work to structure the output in a useful manner. This can be done very efficiently and at scale if you know what you are doing.
So, not only is OpenAI winning but so are people who know how to build content websites and know how to get them to rank. The repetitive content cycle has always been a problem on the web (which is more the reason why places like HN are so great to find actual first-person experiences), but it is about to get amplified a hundredfold.
What OpenAI did is give everyone access to the Internet's content and let everyone mash it together in a way that fits their agenda. Everything that came before ChatGPT (and GPT-4) is pretty much rendered "the past". I'd definitely be pissed. That content was always there, of course. But before all this, you actually had to manually put in the work and the hours to try and piece together multiple key points from 10 different sources to create one unique perspective. OpenAI took a massive dookie on all that and just said fuck it. Let the genie out of the box and let us take a massive cut of the profits in the process.
I have zero doubt in my mind that OpenAI knew exactly what they were doing. They knew what kind of lawsuits or other complaints were going to come their way. And they still went ahead and released it anyway. And they're about to go even more crazy with Plugins. I'm not expecting this to happen per se, but I do have a gut feeling that some kind of brakes are going to be put on all this because whether you agree or not, this whole thing is out of control and not being controlled properly at all.
It’s supposed to be a bad thing that OpenAI is making the world more efficient?
I am talking about creative work. The kind of work where people had to go and first identify the issue themselves (not to mention come up with a unique solution in the first place), and then tell others how to work with it. In fact, there are too many variables for me to truly express how I feel or think on the matter.
All I am saying is that OpenAI knew that they were playing a dangerous game, and their only excuse is that enough "powerful" people will back their project for it not be stomped into the ground by regulations.
Do you not find it pathetic that OpenAI talks about security, precautions, responsible AI and so forth, and yet at the same time rake in millions of dollars in revenue? Do you think their goal was to 'advance humanity' or something like that? I'd be very doubtful of that.
I try not to guess at motives and goals of people I don't know because I don't think they matter. It doesn't sound like your opinions would change if Altman submitted to a brain scan where you could verify his purity (or not), and honestly mine wouldn't either. And no, I dobn't see pathos in for-profit entities talking about safety. Carmakers, construction companies, and theme parks all do the same thing.
I see AI as a productivity multiplier, much as you described factory automation. And I see a lot of information workers suddenly echoing the same concerns we all ignored from factory workers, because oh wow now it could affect us.
I’m sure film photographers complained about digital photography when it was first introduced. And painters about film photography. Etc etc.
1. Humans were generating massive amounts of false information that Google couldn't keep up with, years before Openai appeared.
2. The people who stand to lose the most are those whose jobs are replaceable: information workers. I cringe hearing "but this is different than the industrial revolution and government needs to protect me".
3. If Openai "stole" publicly posted content, then so does Google, and so does every human who reads anything online.
4. You may be right about brakes. Revolutions that displaced factory workers, taxi drivers, typists -- those were all well and good. But threaten the moneyed class and suddenly we're seeing so-called libertarians (not you, I mean like Musk) calling for regulation.
Why should things have human rights and privileges then ? I really don't know the solution to the emerging issues we face here, but I read the same "like a human" argument over and over and I can't really help but notice that LLM's are NOT human. We deny a lot of human rights to primate animals but should immediately accept the human-likeness of LLM's ?
"If Openai "stole" publicly posted content, then so does Google"
This also makes a lot of assumptions which I think we shouldn't make just yet. That's because if we equate LLM's with a search engine then the argument of plagiarism makes a lot more sense. At least google leads to the original content.
Likely they adopted the non profit status as a recruiting tool and they should drop it and join the ranks of for profit companies.
* Despite OpenAI’s mission to ‘benefit all of humanity in a broad and distributed manner,’ early stakeholders that are part of the YC network or bought in are now getting commercially invaluable early access to these systems.
* These GPTs have been trained on large datasets that belong to the commons; text, images, and video freely available across the internet. This free data is typically only used for research purposes under Fair Use law. No compensation has been offered to anyone who created this data. Much of it is copyrighted.
* The blitzscaling of OpenAI is problematic and illegal in a number of ways (Fraud, Breach of Trust and Fiduciary Duties, Legal Disruption, Copyright infringement, etc, etc, etc)
* Demonstrates bias against Trump and not Biden.
I've personally seen a few blind spots / biases in non-political topics as well, but had previously written them off as merely incomplete training data. Lately, though, I've been pondering how many accidental biases have crept in and to what extent that will shape how people work when they're assuming they are getting an unbiased answer.
Suddenly conservatives are worried about the legitimacy of nonprofits, and uninterested in free speech for corporations. Shocking.
I'm not sure what the relationship between the alleged victim of fraud and the alleged perpetrators is, but I doubt "being a member of humanity" is good enough to imply a contractual obligation.
The allegation of copyright infringement due to AI stealing and remixing/republishing online content may hold water if the argument is constructed well, but I don't think this lawsuit is going to be the one answering that important open question.
The litigant filing pro se also doesn't give me much confidence that this lawsuit will succeed. This seems like a rather spurious lawsuit to me. I don't know why you would possibly throw such resources against the court, but I'd expect someone suing a billion dollar company to at least get a(nother) lawyer.
Honestly, the claim that both not wanting to generate something nice about Donal Trump and that YC seems to be democrat-aligned are somehow "Banned Political Activities" makes me think this person needs help. If one politician's values match more with an organization's goal or algorithmic output that's not "political campaigning", that's just a symptom of thinking alike. If reading all the internet has to say about Trump means the algorithm concludes there's nothing good to say, it's probably time to reevaluate your political compass rather than sue the tool for disagreeing with you.
Edit: I would mirror it on one of my sites, but all my sites are pay-per-gb for bandwidth.
Because it has 280 pages of newspaper articles and white papers about OpenAI, ChatGPT, and other startups attached to it as exhibits, many of which are only tangentially related to the case.
PDF embedding is funny.
* "Altman and the other parties to this suit are increasing the risk of global human extinction or actual world domination by a small set of individuals for a chance to personally gain extended lifespans. It is the reasonable explanation for taking such massive risks with this technology and flaunting the law so obviously."
* "OpenAI and at least one of its partners most likely filled social media like Twitter, Facebook, and Reddit with politically charged commentary designed to push votes towards the Democrat party." (And no, the filing doesn't provide any substantial evidence for this assertion.)
* "Y Combinator is, according to ChatGPT, the most notable Tech Accelerator in the world. A screenshot of ChatGPT-3.5 stating this is included as Exhibit L."
* "Open-source and closed-source, not-for-profit and for-profit, are binary choices, or Booleans. Booleans are a form of data with only two possible values, which are typically opposites. When defendants drastically changed the Boolean values that structure [OpenAI]... the founding mission went from ‘true’ to ‘false.’"
Sure, but the kooky claims are mostly tangential to the identified causes of action except the fourth, and they aren’t the sole basis for that one, so they have marginal bearing on the overall suit.
This legal complaint alleges that defendants operating a non-profit entity for the benefit of humanity have committed massive fraud on donors, beneficiaries, and the public. The complaint raises concerns about OpenAI's operation, including its dual structure as a non-profit and a for-profit entity, potential insider dealings, and the exclusion of the general public from its benefits. It claims that OpenAI has used deceptive advertising, unfair competition, and fraud to develop its valuable resource for personal gain.
The complaint highlights OpenAI's mission of benefiting humanity and points out that a narrow group of stakeholders have received commercially invaluable early access to its technology. It also argues that OpenAI's for-profit operations might infringe on copyright and fair use laws, as the technology is built on large datasets, much of which is copyrighted. It accuses OpenAI of breaching trust and fiduciary duties, disrupting legal frameworks, and potentially engaging in willful and wanton negligence by increasing existential risks related to AI.
Finally, the complaint alleges that OpenAI might have engaged in banned political activities, specifically suggesting that the technology may have been used to influence the 2020 US presidential election in favor of the Democratic party.
I've never found it particularly useful for most articles which are easy enough to read/skim (the first and last 2 paragraphs will usually tell you what you need), but long complicated legal documents are a whole other matter. This is great.
My impression was that hallucination happened when it simply didn't have facts in the first place, had conflicting facts, etc.
I thought summarization was generally fairly reliable, but I'd be happy to know if this is not the case.
The failure foolishly and misleadingly called “hallucination” is only one manifestation of an attribution error. If your summarizer leaves out something very important because it doesn’t understand it the result will be quite misleading.
For your average web text which these days is 90% filler and not important anyway, this is no big deal. This particular lawsuit appears the same. But for anything important, I wouldn’t trust it.
On the other hand, if you're in a field where there's an adversarial use of text and the uncomprehended 20% might be used to nullify, contradict or make loopholes in the main body, then relying on ChatGPT is similar to using Tesla Full Self-Driving in a construction zone, near firetrucks, during a snowstorm.
The only challenge is chunking the larger bills and synthesizing the larger summary without losing out on possible nuances. Something like California's SB423, for example, is over twice the 8K token limit and that's not even a large bill.
Unfortunately, things like the US Code or Code of Federal Regulations are in the range of 100s of millions of tokens.
6.4MB
Gzipping the 52 MB PDF just shrinks it to 51 MB. You got it to 6.4 and kept it PDF.
gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 -dPDFSETTINGS=/screen -dNOPAUSE -dQUIET -dBATCH -sOutputFile=output.pdf input.pdf