Rather than "pleading" for them to stop, wouldn't she have more success going after the ai content creation companies via legal process? I thought actors have the right to control commercial uses of their name, image, voice, likeness, or other recognizable aspects of their persona, thus if people are paying for the AI creation wouldn't the companies be wrongly profiting off his likeness? Although I'm sure some laws haven’t yet been explicitly updated yet to cover AI replicas.
> Rather than "pleading" for them to stop, wouldn't she have more success going after the ai content creation companies via legal process?
But that shouldn’t be the first step. Telling your fellow man “what you are doing is bothering me, please stop” is significantly simpler, faster, and cheaper than contacting lawyers and preparing for a possibly multi-year case where all the while you’ll have to be reminded and confronted with the very thing you don’t want to deal with.
If asking doesn’t work, then think of other solutions.
You can't tell everyone. Barely anyone will know of this being published. And then there will be lots of people thinking "whatever, I don't care". And a not insignificant number of people thinking "lol, time to organise a group of people who will send Robin Williams creepy genAI to her every day!"
Asking solves it for you, biting the hand makes them think twice about doing it to others, but good luck doing that to the many-headed serpent of the internet.
I don't know. I mean, things could have gone very differently (in a better or worse way) and the world may be unrecognizable today if certain key events did not happen.
Like if nothing sparked the World Wars (conversely: or if Hitler won). Or Greece harnessed steam or electricity to spur an industrial revolution 2200 years ago. Or if Christianity etc. never caught on.
> Telling your fellow man “what you are doing is bothering me, please stop” is significantly simpler, faster, and cheaper
It's not because just telling people on the internet to stop doing something doesn't actually stop them from doing it. This is basic internet 101, streissand effect at full power
No, she would not have any success. Take a look at this list and think about the sheer number of companies she would need to sue: https://artificialanalysis.ai/text-to-video/arena?tab=leader.... You'll see Google, one of the richest companies on the planet, and OpenAI, the richest private company on the planet. You'll see plenty of Chinese companies (Bytedance, Alibaba, Tencent, etc.). You'll also see "Open Source" - these models can't be sued, and removing them from the internet is obviously impossible.
The most these lawsuits could hope to do is generate publicity, which would likely just encourage more people to send her videos. This direct plea has that risk too, but I think "please don't do this" will feel a lot less adversarial and more genuine to most people than "it should be illegal for you to do this".
> The most these lawsuits could hope to do is generate publicity, which would likely just encourage more people to send her videos.
It's not fruitless and doesn't only generate publicity. Some states like California and Indiana recognize and protect the commercial value of a person's name, voice, image, and likeness after death for 70 years, which in this case would apply for Robin William's daughter.
Tupac's estate successfully sued Drake to take his AI generated voice of Tupac out of his Kendrick Lamar diss track.
There is going to be a deluge of copyright suits against OpenAI for their videos of branded and animated characters. Disney just sent a cease and desist to Character.ai last week for using copyrighted characters without authorization.
What I'm saying is that successfully suing individual companies or people would have zero impact on her actual problem. If California says it's illegal and OpenAI says they'll ban anyone who tries it, then these people can effortlessly switch to a Grok or Alibaba or open source model, and they'll be extra incentivized to do so because they'll find it fun to "fight back" against California or luddites or whatever. Do you see the difference? Tupac's estate successfully stopped one guy from using Tupac's voice, but they have not and cannot stop the rest of the world from doing so. The same is true for Disney, it is trivial for anyone to generate images and videos using Disney characters today, and it will be forever. Their lawsuit can only hope to prevent a specific very small group of people from making money off of that.
The problem she wants to solve is "people are sending me AI videos of my dad". She will not have any success solving this problem using lawsuits, even if the lawsuits themselves succeed in court.
Is that really the problem she wants to solve? She could just turn off her phone to accomplish that. The problem is multi-layered and complex. Holy shit, It’s her dad. He’s dead. I don’t have her phone number, but let’s pretend I did and we were friends, why would I be texting her videos of her dead father? He’s Robin Williams, sure, but why? why! would I be making AI videos and sending them to her? Forget Sora, if I made a puppet of her father and made a video of him saying things he didn’t say, and then sent it to her, I think I’d still be a psychopath for sending it to her. I think she should sue open AI and California should have it be illegal without a license, and yeah there’s always gonna be a seedy underbelly. I’m sure there’s Mickey Mouse porn out there somewhere. A lawsuit is going to make it official that she is a person and she’s saying hey I don’t like that and that she would like for people to stop it, and that the rest of us agree with that.
Asking people to stop seems like the first step. Especially since this is specific to people sending them to her in particular. People think they are being nice and showing some form of affection, but as she mentions she finds it disturbing instead.
So I don't think there was actually malicious intent and asking people to stop will probably work.
I think there is also a major distinction between creating the likeness of someone and sending that likeness to the family of the deceased.
If AI somehow allowed me to create videos of the likeness of Isaac Newton or George Washington, that seems far less a concern because they are long dead and none of their grieving family is being hurt by the fakes.
How long would the legal process take? How much would it cost? Does she have to sue all proprietors of commercial video generator models? What about individuals using open source models? How many hours of her time will these suits take to organize? How many AI videos of her dad will she have to watch as part of the proceedings? Will she be painted as a litigious villain by the PR firms of these very well-capitalized interests?
Her goal seems to be to reduce the role in her life played by AI slop portrayals of her dad. Taking the legal route seems like it would do the opposite.
It's so frustrating that "just call the cops" is the answer, at the very same time that the cops are creating a massive disruption to our society.
And even if this were a viable answer: legal process _where_? What's to stop these "creators" from simply doing their computation in a different jurisdiction?
We need systems that work without this one neat authoritarian trick. If your solution requires that you lean on the violence of the state, it's unlikely to be adopted by the internet.
Legal process is not an “authoritarian trick," it's the primary enforceable framework for wide scale, lasting societal change as it's the only method that actually has teeth.
Also, calling legal enforcement as “leaning on the violence of the state” is hyperbolic and a false dichotomy. Every system of rights for and against companies (contracts, privacy, property, speech) comes down to enforceable legal policies.
Examples of cases that have shaped society: Brown v Board of Ed, pollution lawsuits against 3M and Dow Chemical, Massachusetts v. EPA resulted in the clean air act, DMCA, FOSTA-SESTA, the EU Right to Be Forgotten, Reno v. ACLU which outlined speech protections online, interracial marriage protected via Loving v. Virginia, the ruling that now requires police have a warrant to access cell phone data was Carpenter v. US, and these are just a few!
> And even if this were a viable answer: legal process _where_? What's to stop these "creators" from simply doing their computation in a different jurisdiction?
Jurisdictional challenges don't mean a law is pointless. Yes, bad actors can operate from other jurisdictions, but this is true for all transnational issues, from hacking to human smuggling to money laundering. DMCA takedowns work globally, as does GDPR for non-EU companies.
Nobody’s arguing for blind criminalization or over policing AI. But perhaps there should be some legal frameworks to protect safe and humane use.
Not really. From what I understood of the interview is that her complain is not about money or compensation (which she may be entitled to), but about how people use the technology and how they interact with it and with her. Legal process or even the companies implementing policies won't change that problematic society behavior.
Since the raise of generative AI we have seen all sorts of pathetic usages, like "reviving" assesinated people and making them speak to the alleged killer in court, training LLMs to mimic deseased loved ones, generative nudification, people that is not using their brain anymore because they need to ask ChatGPT/Grok... some of them are crimes, others not. Regardless most of them should stop.
