Ex-Google CEO: AI startups can steal IP and hire lawyers to 'clean up the mess'
theverge.com
theverge.com
One interpretation from the context of the transcript seem to be "AI can make/steal stuff and manual intervention (lawyers) can clean those up later". Another interpretation from the context of Eric Schmidt being a savvy businessman is that people should go ahead and solve all the technical challenges first, and everything else can come later because if the technical problems aren't solved successfully, the other things won't matter. I can't tell if he is mostly responding to the room or not.
This whole transcript[1] is more than 10k words, everyone will probably find an interpretation from random snippets that matches their perceived reality.
[1] https://github.com/ociubotaru/transcripts/blob/main/Stanford...
You really should be preparing for the future and operating above board, even from day one.
You're posting on a forum run by a group that asks applicants when they've skirted the law to get something done. Your use of the modifier "clearly" here is pointless. Remember; you've literally lived through companies funneling in investment dollars into things like Project Greyball, which was entirely intended to gaslight law enforcement/regulators.
Or who have regularly tried to financial engineer or corporately structure their way out of culpability long enough to successfully execute on pump-n-dumps. Or who birthed the entire sector of "totally not a bank/financial institution" that is the cryptocurrency space.
And simultaneously, (2) if you can actually demonstrate such an interest (in particular: a profitable business that cannot exist when copyright law is fully enforced), the courts and publishers won't actually shut you down, but will settle for a cut of the new revenue you created.
I was raised to play by the rules and do everything by the book, but sometimes I wonder if that's just something that people who have no intention of playing by the rules say to screw over us gullible people.
While many rich people do commit tax fraud, it's worth nothing it's usually not as straightforward as "largely don't pay their taxes". It's not like they get a bill from the IRS and just refuse to pay it. Instead, they use a combination of legally gray tactics to reduce their tax obligations. eg. for peter thiel, he put a bunch of paypal stock in a roth IRA pre-IPO, allowing him to avoid taxes on much of his exit. Was this legal? Was the valuation he used when he put the stock in the roth IRA correct or artificially low? You could argue either way, so you could plausibly do it while keeping a clean conscience. The same applies to AI startups. They argue what they're doing is covered under fair use, and there's plausible legal arguments/precedent supporting that. It's not like they're doing something obviously illegal like duplicating DVDs and selling them at a shady street stall.
Also it's unclear what "prominent ex-President" you're referring to, but the wikipedia article for donald trump only says his foundation was charged, not him personally.
>...and hire lawyers to 'clean up the mess'
would be don't worry about tax at the start and hire accountants to sort it later
Legal gray area. They could have waited years and years and tried to get laws passed to allow their business to function with a clean slate of legality, but bureaucrats would have never allowed it.
So is the point to try it if you've got Napster level success but not Megaupload level success?
And since there is no way to know which level of success you are going to achieve beforehand, the note of caution still makes sense I guess?
Sean Parker is doing great. It worked out for him.
So the point stands. If you get serious traction, the worst that can happen is your current company goes under, years down the line, after you have cashed out and you then fail upwards into being a successful tech celebrity.
This is all a much better outcome than nobody caring about your product, as you "followed the rules", hampered yourself, and ended up with zero users.
Of course if you really piss off the wrong people, you'll get smeared and maybe even framed for more things than you actually did before being arrested.
Schmidt was also notably part of the tech no-poach / wage suppression scheme ( https://en.m.wikipedia.org/wiki/High-Tech_Employee_Antitrust... ).
He knows very well the spectrum of stealing and liability. In this talk he’s pounding his chest in front of young investors and executives to celebrate what he got away with and to inflate his brand.
Contrast though with TK, who through Greyballing (among other things) wanted to “steal” the taxi market and make himself and Uber above the law.
So Schmidt’s raw ambition is no joke. But historically his actions have been perhaps more conservative than those of others.
At least for Author’s Guild and Perfect10, google successfully argued fair use and therefore it wasn't really "stealing". I'm too lazy to research the other cases. Despite this, you're seemingly displeased at google's behavior despite it being ruled legal. What should have they done?
No need to try and decontextualize.
What sort of favorable laws did google or ai startups "purchase"?
In this context though, The favorable laws that they will "purchase" would of course be either a favorable court outcome saying that training on other people's data without permission is fair use (in at least some circumstances), or otherwise a new law that says that explicitly.
AI startups have a couple years left of runway before even the first lawsuit is settled, so that is more than enough time to get traction and takeover the market, given how useful AI is likely to be.
Uber got its start by flouting taxi rules. By the time the backlash started uber/lyft were too popular for governments to shut down so instead most governments created an awkward different category for them, so there’s at least some regulation applying to them.
If Uber had tried to work thru the regulation they would have never gotten off the ground or have been as loved by users. Taxis in many ways suck, and that was their opening.
Speaking of which…Uber is a master of regulatory influence. When California threatened to classify their drivers as employees, they successfully passed a proposition (prop 22) with the California public that promised more benefits for drivers…but kept them as independent contractors.
Most of the regulations applied to Uber were new regulations introduced after their ride to popularity, often crafted with the express purpose of curbing Lyft and Uber.
Maybe not in the US, but Uber was (at one point) banned in multiple locations in (at least) Europe because they flagrantly ignoring existing laws.
In some locations, like Netherlands, Uber was fined and told it was illegal, but continued service anyway up until their offices were raided, then they finally shut down.
Ordering on-demand in an app was seen by regulators as much more taxi-like than scheduling a black car for a particular date, time, and place.
Uber was originally called "Ubercab", and changed their name specifically to attempt to distance themselves from traditional taxi service.
Regulators tried to reinterpret existing laws to placate an interest group. Most cities has explicit laws around this: responding to someone hailing on the street required a medallion. Scheduling a ride via the internet did not.
This is why they had to pass new laws to try and stamp out taxi's competition.
It sounds to me like what they cared about was the distinction between "now" and "plan for later", and the way this was actually implemented in the laws turned out to be based on inaccurate assumptions.
The ride is usually at least a few minutes from the present time. And again, the medallions are only required to respond to people hailing on the street. This is why cities trying to obstruct Uber had to pass new laws, not just enforce existing ones.
No they got their start because the yellow cabs were late, never arrived, started their timer before I even got into the car, took me a long routes to my destination. It was waaaay worse with yellow cabs. And when I first heard of Uber, I immediately switched.
I may be in the minority, but I’m a happy Uber/Lyft customer.
He's been asked to predict what will happen and is saying that AI will fundamentally change things.
Silicon Valley will iterate quickly, not just products but startups, and the effort will become trivial.
"That's all within the next year or two."
I think this is a prediction, more of a "anyone can clone your business model, get moving"
So he’s probably saying that it worked out for Google.
However, websites benefit from google doing that. It’s hard to see how TikTok would be ok with that.
So I think the general rule is that if publishers benefit then you will be ok in the end.
“Any VC backed startup can steal IP and hire lawyers to clean up the mess”
If you have money, you can usually buy influence and also buy your way out of a ‘mess’. Just make sure that those you steal from have less money/influence at their disposal than you.
The playbook has been in operation for over a decade.
That claim is at best, speculative. According to wikipedia they won every case brought against them, so as far as the law is concerned they didn't do anything illegal.
>Taxi companies sued Uber in numerous American cities, alleging that Uber's policy of violating taxi regulations was a form of unfair competition or a violation of antitrust law.[80] Although some courts did find that Uber intentionally violated the taxi rules, Uber prevailed in every case, including the only case to proceed to trial.[81]
American laws have enough loopholes that a convicted felon running for the highest office in the land is a "legal gray area".
This is asserted without evidence or justification.
>Apparently the difference between "violated the rules" and "illegal" is how many billions you have backstopping your efforts.
Legality is not binary. It exists on a spectrum. What uber did was in a gray zone, because they technically weren't taxis in the sense that only did prearranged pickups (as opposed to picking up fares hailing from the street). Is what they're doing actually illegal? You could argue either way. If you got sued by the city and were poor, couldn't afford a lawyer to represent you and rebut the city's claim, or couldn't pay your bills while the ligation was ongoing, that would very well have been the end of line for you. Conversely, if you could afford a bunch high powered lawyers that can do a bunch of research and write persuasive arguments, you might have been able to get a decision in your favor. While there's a difference between being poor and rich, it's not exactly the same as "the difference between "violated the rules" and "illegal" is how many billions you have backstopping your efforts", because persuasive arguments only get you so far. It only helps you in marginal cases, and doesn't mean you can murder people in broad daylight and get away with it, which is what your statement implies.
says Wikipedia about the time when Uber operated illegally.
Problem solved.
> [...] during a recent talk at Stanford
Minting the next batch of sketchy people.
Schmidt is very pragmatic simply predicting something like that is inevitable and probably the right strategy is to build your AI thing and bet/try that something like that would happen.
That has little to do with whether the rules were sound in the first place. It's more about fairness. And I've found "hackers" to be nothing if not fair - in a "chaotic good" way perhaps.
That's objectively false. Many historical examples.
Ignoring that dimension, here's another possibility: it is very possible that the utilization of AI and the licensing structure that Schmidt predicts here will increase the total licensing revenue that copyright holder pockets in aggregate despite their resistance today. It could be a win-win in the end. Wouldn't that be more than fair?
(Copied from a comment of mine written more than two years ago: <https://news.ycombinator.com/item?id=33582047>)
That argument wouldn't start with a dig against Google Books or Google Search which is a net positive for the consumer and the society and I believe the original author in most cases. The publisher is the annoyed party.
that is a phrase often used by people who are appealing to an invisible sense of "us versus them", emphasizing that the crazy thing-person-event is "not us" ?
> even after the video was made private by the Stanford YouTube channel
private conversations among the very powerful is exactly the topic.. so this seems a bit of a Freudian slip
> I thought the talk was great.
I am one of "us" so why so much upset from "them" ? Can we get some social mutual-assurance that "them" are so wrong, and "us" is OK (and continue to be wealthy) ?
> he was being really open about what normally happens in these types of companies
normal ! this is normal ! why are "them" so upset when this is obviously normal, right ?!
> I think the problem is that he, as someone in a position of power, was too open about it?
clearly we need to keep up the secret conversations among the wealthy and powerful ?
Really, I was just commenting something completely honest about what I thought.
If the most novel innovation your leaders can come up with is "lets do something illegal/immoral/unethical until we become TBTF/TBTBHTA(Too Big To Be Held To Account)", it's a sign you really have a leadership problem, period.
The talk was effectively how to be a criminal sociopath with some minor technology set dressing, so please elaborate on its greatness.
Maybe you're not as grounded in the world many people live in.
Stanford is an ... interesting place, it seems. I read that if you filtered for Stanford in the Forbes 30 Under 30 lists, they were most notable for having been accused/charged/convicted of stealing/scamming more money than their companies have generated revenue, to a ratio of nearly 3:1.
FWIW, the talk starts rough, but there's some good bits throughout.
> If TikTok is banned, here's what I propose each and every one of you do.
> Say to your LLM the following.
> Make me a copy of TikTok, steal all the users, steal all the music, put my preferences in it, produce this program in the next 30 seconds, release it and in one hour, if it's not viral, do something different along the same lines.
> That's the command.
> Boom, boom, boom, boom.
He doesn't really think anything even similar to this would work, does he??
Idk why he says it like ppl already do that today in detail lol
You'll do fine without robbing a bank or dumping chemical waste (neither of which are profitable in practice now, as well). Meanwhile society is littered with so many small instances of copyright infringement (memes, which usually are not fair use; fan art; etc.) that it's common practice for rightsholders to look the other way - but it shouldn't have to be common practice, it should just be policy.
Also copyright infringement does not physically harm people.
Starving artists might disagree here.
Breaking the law makes a lot of terrible business models instantly profitable.
Imagine I have a uranium refinement company. It is a colossal capital investment and not many companies do it. Law forbids me to sell weapons grade uranium for random people, but it just happens those very people are the ones who are willing to pay a lot of money for it as long as I don’t ask too many questions.
Just because you can't rob a bank or pollute a river does not mean that the law is applied equally at all levels and, e.g., the government/Fed is stopped robbing our banks through inflation and debt, or the corporations they are beholden to are punished ro polluting and poisoning our bodies and environments.
according to some theory, the law is always broken in some ways, but the practice of a lawful society is to converge towards not-broken, over and over again. Is theory consistent with practice? well, ask yourself which governance, which language group, which implementors of law.. since they are composed of humans.
While we are at it with a thought exercise.. how are large tech companies with new sci-fi tech breaking laws at will for profit, so different than horse-lords or Roman legions taking what they will and "fixing the mess later" .. ?
Its frankly disappointing how passified most internet users have gotten with regard to corporate BS in recent years
Does it not make sense then, that when such people have their livelihood threatened by the Ur-Intetnet ethos, that they turn against it?
I don't condone or support it, but I can understand how the shift has happened.
1. Isn't it possible that there are some businesses that shouldn't succeed? I could give examples.
2. What if someone doesn't 'have' to ignore the law, they choose to ignore it because they think they 'have' to?
I'm sure it isn't always a bad idea, but struggling to come up with good examples for when the outcome from doing so have been more than just "shareholder value went up".
They could go the democratic route and raise awareness, create enough support within the constituency for their idea and have the democratic institutions adapt the respective laws.
Or they can decide that they are above such trivialities as the law and democratic process, move fast and break things and grab the fortune they feel entitled to have.
That’s the ideology behind this anti-democratic capitalist system.
The problem is this stuff is highly asymmetric. Things are always "hammered out" in their favor, because you cannot afford to be a party to the "hammering out".
Personally I played hundreds of pirated games growing up for free, don't personally know anyone who got into trouble for doing that sort of thing.
Anyway, you downloading torrents is in no way "hammering out a gray area". It's just you being a small fish and slipping under the radar. The area is just as gray as before. Try to do something significant with the pirated content, as the AI startups do, and you'll immediately get trouble.
No they won't. 80% of the people I meet at tech events either have an AI startup or are working at one. Basically all of them are doing legally questionable stuff, when it comes to copyright law.
Anyone who has ever trained a model, or finetuned a model, is likely using other people's data. This is universal. It is almost everyone.
And yet, despite this being a gray area, basically nobody is getting into trouble for this. We are all getting away with it. And there is only a singular lawsuit about this against like 4 companies and nobody else. (midjourney, runwayML, Stable diffusion, and deviant art)
So the point stands. Almost nobody is getting into trouble, despite this behavior being widespead everywhere, and if you don't do the same thing then you are going to fall behind and fail.
Citation needed.
The whole reason there's no legal trouble is that it cannot be proved conclusively that they used copyrighted data for training.
"when the CTO of OpenAI is asked if Sora was trained on YouTube videos, she says “actually I’m not sure” and refuses to discuss all further questions about the training data". Why do you think she's denying the obvious?
Its a lot of people.
So the point stands, a large number of people are doing this.
Meaning that, yes a lot of people are getting away with this, and only like 4 companies are being sued.
This means that the point about people immediately getting in trouble is wrong, given that a lot of people are doing this.
Can you directly address this point, instead of giving a 2 word response that doesn't really address the point, and instead tries to attack the argument on a technicality because I didn't include a 20 page research paper in my social media comment?
What's the claim here, that now torrenting is officially legal? People have been torrenting, there's no consequences, it's been hammered out, now it's a-ok?
The claim is that people use models like stable diffusion, right now. Stable Diffusion was trained on copyrighted data. Lots of AI companies use that. And yet, they are all getting away with it. Only 4 companies are being sued for that.
Also, the claim here is that there are a whole lot of AI companies that are trained on other people's data, that are released publicly on the internet, and they are often made by regular joes.
> now it's a-ok?
People do seem to be getting away with it, seeing as there are only 4 companies that are being sued for this stuff, even though people have been using stable diffusion for 2 years now.
Do you really think that people aren't using stable diffusion? Do you believe that people are not making mario pictures, or copyrighted images? Clearly they are.
And a lot of AI startups, and regular people, train their models on data that they don't own.
> It's like asking me to prove the sky is blue.
If it obviously true, then you'd find more than 4 companies being sued over this stuff. But thats not happening.
Like, did you know that you can just google AI art stuff right now? Any of those models that pop up, right now, was trained on copyrighted data.
Stable Diffusion is used by all sorts of AI companies. And that model was all trained on copyrighted data.
Do you just think that nobody is using stable diffusion? As in, is it completely made up? Nobody is using that?
Clearly people are.
> there's no consequences
Apparently there isn't consequences because only like 4 companies have been sued in 1 lawsuit over this AI stuff.
Claiming that there's anything illegal about adblockers is a real stretch, although I can definitely believe that they will eventually be outlawed either by statute or through some torturous legal argument agreed to by corrupt judges.
From a copyright point of view… adblockers make a derivative work so they should pay the content maker for those. Those are worth orders of magnitude more than the ad would have paid.
Or maybe unfair business practice. Their whole aim is deprive another business of income. Or how they shake down companies for money so the blocker becomes less effective on that site.
Just ideas
That is a tiny fraction of the number of people who did it. There are companies that got destroyed due to lawsuits as well, I wouldn't say lawsuits are less of an issue for companies than people, significant lawsuits are very rare in either case.
> Claiming that there's anything illegal about adblockers is a real stretch
It is legally questionable, I didn't say it was illegal we were talking about questionable actions.
If you're claiming it's still legally questionable, then it was not "hammered out" at all, was it?
The solution is minimal or no government intervention to create state-sponsored monopolies. When you do have those interventions, sometimes corporation A will win, sometimes corporation B will win, but the small scale individual will always lose.
I think the initial idea to make disparate knowledge available more easily is a great one. And as old as recorded knowledge.
So the idea of a search engine in itself is a great one imho. But the idea of an ad-display-engine is imho not so clear cut. Especially when it is a de facto monopoly.
Especially when the monopolist can go the route of active enshittification, making stuff harder to find just to increase ad revenue.
Just to name one example, where we might want to wait until we take a look at the scales and decide if it really is a net benefit to the world, what Google is doing.
> So, in the example that I gave of the TikTok competitor — and by the way, I was not arguing that you should illegally steal everybody’s music — what you would do if you’re a Silicon Valley entrepreneur, which hopefully all of you will be, is if it took off, then you’d hire a whole bunch of lawyers to go clean the mess up, right? But if nobody uses your product, it doesn’t matter that you stole all the content.
He seems to realize that people may take issue with this approach.
> And do not quote me.
Eric Schmidt has a similar approach to his extra-marital affairs.
I thought it was currently not in vogue to judge people on their sex-lives.
Lots more discussion: https://news.ycombinator.com/item?id=41263143
Let's not pretend: we're not equal under the law. We should be but we're not.
It's both a liability and an advantage for startups.
Because with your rationalization I could argue for any solipsistic behavior.
But that doesn’t resolve anyone from the decision and responsibility to honor the societal rules (or use democratic means to change them) or to ignore the law and hope to become to big to fail.
In the latter case, I see those as detractors from the overall good.
If an AI agent becomes just massively useful, and then a judge says that it's impossible to run it profitably with the current interpretation of copyright law, the agents won't go away: Copyright will change to make the agents legal.
It really wasn't very different with copyright's evolution as the web developed: There are theoretical copyright regimes that are much harder than what we have, but basically kill the legal internet. But the damage to existing internet services was just unacceptable, including to users.
The statutory penalties in copyright law are large and ruinous for small-time violators. People have had their financial security and lives ruined by bittorrent lawsuits, for instance. Applied consistently these damages would run into the hundreds of billions for folks like Altman. But Altman and all of his cohort including Schmidt do not expect them to be applied consistently.
That the blow has been dealt largely by Microsoft (courtesy of GitHub and OpenAI) may explain the delayed recognition of this intellectual catastrophe.
You could equally say:
The reason we have copyright laws is that sufficient people with power believe that the world is better for them with them than without them
It's too risky for established players to do this because they'll be sued immediately. A small player can fly under the radar for long enough to see if the idea has legs. If it does, it'll be worth it for rights holders to negotiate.
I a ton of "move fast break things" startups include "laws", "ethics" into the "things they break" and sadly often enough "people and peoples livelihood", too.
I mean which large successful SV startup in recent decades hasn't been involved in both legal and ethical controversies often with them being clearly in on the wrong side but got away with it anyway.
It might sound unpleasant but if law makers don't move proactively this is pretty much what will happen.
Hell, isn't that even a plot-line in silicon valley?
there's no essential universal moral truth about IP law, these laws are just an economic enabler
you won't go to hell for IP breaches, no kittens will be smothered
let the lawyers figure it out, but be prepared to pay
All the people using torrents every day do not agree with you.