site banner
PaperclipPerfector  11mo ago (text post)   25966 thread views

Culture War Roundup for the week of May 20, 2024

This weekly roundup thread is intended for all culture war posts. 'Culture war' is vaguely defined, but it basically means controversial issues that fall along set tribal lines. Arguments over culture war issues generate a lot of heat and little light, and few deeply entrenched people ever change their minds. This thread is for voicing opinions and analyzing the state of the discussion while trying to optimize for light over heat.

Optimistically, we think that engaging with people you disagree with is worth your time, and so is being nice! Pessimistically, there are many dynamics that can lead discussions on Culture War topics to become unproductive. There's a human tendency to divide along tribal lines, praising your ingroup and vilifying your outgroup - and if you think you find it easy to criticize your ingroup, then it may be that your outgroup is not who you think it is. Extremists with opposing positions can feed off each other, highlighting each other's worst points to justify their own angry rhetoric, which becomes in turn a new example of bad behavior for the other side to highlight.

We would like to avoid these negative dynamics. Accordingly, we ask that you do not use this thread for waging the Culture War. Examples of waging the Culture War:

  • Shaming.

  • Attempting to 'build consensus' or enforce ideological conformity.

  • Making sweeping generalizations to vilify a group you dislike.

  • Recruiting for a cause.

  • Posting links that could be summarized as 'Boo outgroup!' Basically, if your content is 'Can you believe what Those People did this week?' then you should either refrain from posting, or do some very patient work to contextualize and/or steel-man the relevant viewpoint.

In general, you should argue to understand, not to win. This thread is not territory to be claimed by one group or another; indeed, the aim is to have many different viewpoints represented here. Thus, we also ask that you follow some guidelines:

  • Speak plainly. Avoid sarcasm and mockery. When disagreeing with someone, state your objections explicitly.

  • Be as precise and charitable as you can. Don't paraphrase unflatteringly.

  • Don't imply that someone said something they did not say, even if you think it follows from what they said.

  • Write like everyone is reading and you want them to be included in the discussion.

On an ad hoc basis, the mods will try to compile a list of the best posts/comments from the previous week, posted in Quality Contribution threads and archived at /r/TheThread. You may nominate a comment for this list by clicking on 'report' at the bottom of the post and typing 'Actually a quality contribution' as the report reason.

8
Top Bottom Old Controversial Comments
Jump in the discussion.

No email address required.

Sam Altman just loves to be a sociopath and then brag about it. His latest?

https://x.com/sama/status/1790075827666796666

"Her".

In case you've been living under a rock, this is in reference to the 2013 movie in which Scarlett Johansson plays the voice of an AI girlfriend. And it's also a reference to Open AI's new product, Chat GPT 4o, whose voice sounds just like... you guessed it, Scarlett Johansson.

This is no mistake. Open AI actually approached Ms. Johansson and asked her permission to use her voice. When she said no, they said fuck it and did it anyway.

https://x.com/BobbyAllyn/status/1792679435701014908

If Elon Musk is chaotic neutral, Sam Altman is increasingly proving himself to be lawful evil. It's not a good look.

I think it's insane that US makes hiring lookalikes illegal. People writing about "owning their own likeness", but the existence of this law proves that it wasn't theirs to begin with. Scarlett Johansson was just one of the women who looks and sounds generally like her to become famous. In the alternative reality person who voiced gpt4o was famous and SJ was hired because she sounded similar enough.

I am not even remotely some kind of libertarian, but what is the actual harm to society this regulation prevents? People who look like other people always existed and if you obviously can't stop some man eerily similar to you from acting in a hardcore pornography I don't see why SJ should be granted anything here.

you obviously can't stop some man eerily similar to you from acting in a hardcore pornography

You can't, but my understanding is that if you are a notable person, and if the producer of the pornography chose to use someone with your likeness because of the value that your personal brand adds, and if you have not given your permission, and if they are in one of the 35 states with prior case law about such things, you might have a pretty strong case. Though IANAL so TINLA.

Yeah, hence I write that it's crazy that this case law exists. Why would being a notable person grants you more rights and what societal purpose these rights serve.

I mean is it crazy that you can't build a car, slap a Rolls Royce logo on it, and sell the car? It's not a "being a notable person" issue, it's a "having a personal brand" issue.

I don't see why a theoretical Ginger Johnson, Scarlett's separated-at-birth identical twin, should be prevented from having a public career because Scarlett beat her to it and she'd (unavoidably) be riding Scarlett's coattails. So the same goes for other look-alikes.

As long as Ginger does not benefit specifically by doing things which are intended to cause people to falsely believe that Scarlett did the thing rather than her, I think Ginger is fine.

you obviously can't stop some man eerily similar to you from acting in hardcore pornography

This seems .... oddly specific. And an epic humble brag if so.

I pulled an Altman with Anna Khachiyan's voice, incidentally: https://x.com/DainFitzgerald/status/1791195409383292992

Almost everything and everyone still connected to Open AI strikes me as incredibly creepy and scammy but this is one of those things where I feel like we as a society will just have to get over. Did they steal her voice? Did they hire someone who sounds just like her? It hardly matters; soon enough it will be trivially easy to create deep fake virtual porn with a realistic voice saying whatever nasty shit you want it to. There's no putting the lid on this.

This is scuzzy and icky and whatever, but is it really true that one must ask permission from an individual before replicating some facet of their physical existence?

As a commercial matter, sure, you can't profit off the likeness of another and that settles this (and Sama already pulled the voice). But beyond the narrow commercial protections, I don't know that society has ever endorsed a broader sense in which reproducing someone's face or voice without permission is off-limits.

A good part of the philosophical problem with right of publicity is that it has very little impact outside of the commercial protections, and even the commercial protections are only protective to the point where the broader public knows you. It's very much a cut out to protect celebrities and the famous, not defend the little people. California has an unusually broad combination of statutory and common-law protections, but it's still only something that matters to public figures worried about getting used as an advertising or product campaign.

((Other 'moral rights' have similar problems: see VARA for a particularly ugly one.))

But that specific context impacts here, at least if Altman did what people are thinking he did.

This isn't a new problem, but probably is the highest-profile case. I remember hearing a podcast episode about "the default voice of Tiktok" a few years back (2018?), who was suing because the scope of the project was unclear when she signed up (IIRC it was pitched as an academic project), and because, according to her, widespread, easily-recognized text-to-speech using your specific voice is a career limiter for other voice acting prospects.

Does this really sound that much like Scarlett Johansson? Even after Sama made that tweet this still didn’t occur to me. It just sounded like a generic, friendly female voice to me, and I think the “Her” tweet was just a reference to the plot of the movie, not the voice.

This whole kerfuffle seems annoying, and also seems like Scarlett Johansson reaching for a way to include herself.

They offered her a job, she refused, and then they got somebody else to do the same thing. Now she’s mad. This is not interesting.

Maybe the angle here is that since AI duplications are so easy and good now that we’ll enter a sort of guilty until proven innocent phase where everybody assumes they are more important than they might actually be.

No the company is not “deep faking” you, you just aren’t actually unique.

No the company is not “deep faking” you, you just aren’t actually unique.

I think there are too many indicators in this case pointing in the opposite direction.

  1. Actress known for playing voice of AI
  2. Actress known in part for her husky voice
  3. Actress approached by AI company for permission to use voice, she refuses
  4. AI voice sounds suspiciously similar to that of above actress
  5. AI company explicitly references the actress' AI role in promoting the new voice

On the other hand, if OpenAI had used a voice similar to that of John Huston, who previously had no connection with AI, and who is not a currently active celebrity profiting off of their known assets, it would seem like a fun, quirky choice. There might even be debates over whether the AI was based on Huston or Daniel Plainview, whose voice was modeled on Huston. I don't think Anjelica Huston would be suing them, unless they had asked the Huston estate and had been refused.

No the company is not “deep faking” you, you just aren’t actually unique

This really gets to the root of the visceral reaction some people have towards most of generative AI. People view their artwork, their voice, their face, as some of the most personally identifying "what makes me be me" features. The idea that there are literally other people that look and sound just like you, or even worse that a computer can emulate it believably, flies in the face of how some people see themselves.

Kindof understandable, many were told from an early age that they were a one of a kind special little guy, and finding out that it was all a polite fantasy would probably be quite jarring.

To be fair, art and voice AI was literally trained on existing examples of the respective media. I don't think the criticism comes from a place of "the AI makes me feel worthless," it comes more from "the AI is basically carbon-copying my style."

i feel like you are just steelmanning this position rather than actually making those claims, so i'll spare you my layman's understanding of how training data is not used in a currently copyright protected fashion.

What i have a big problem with though is people who become possessive over a "style". You can't own a style, not in visual auditory or any other sense of art. Also injecting "worthless" into the discussion goes beyond what i was saying, i was saying that finding out the singularity of your identity is less all-encompassing than one originally imagined can harm the ego.

Does this really sound that much like Scarlett Johansson?

ScarJo claims (in my linked tweet above) that it sounded so much like her that her friends and family couldn't tell the difference.

So... I think either yes or she is lying.

Or engaging in motivated reasoning, or misunderstanding how unique individual voices actually are, or probably a bunch of other possible explanations.

Where how unique it is might matter for California's statutory right of publicity, the state's common-law right is far more expansive. I'll point to White v. Samsung, where this was close enough to trigger California's common-law right of publicity.

((Look at the decision itself for even more expansive stuff: "Here's Johnny" alone was apparently enough for the 6th Circuit to find infringement of right of publicity.))

It's an absolute mess of a standard, and celebrities have marinated in it so long that it's water to them.

Even after Sama made that tweet this still didn’t occur to me.

It occurred to everyone else on the internet (including me but I was primed). If a ton of people hear ScarJo and then it turns out they literally offered the job to Johansson, I'm inclined to believe she's reasonable in thinking there was something there.

No the company is not “deep faking” you, you just aren’t actually unique.

Almost no one is. Amber Heard had a body double do the sex scenes for a movie, I guess you could say that that makes her not unique - people could get their titillation from someone else with the same body.

But I don't think it's delusions of grandeur to think that part of the value (in this case the titillation) in such a scene is specifically that it's Heard. There's a reason they took the legal risk of doing this. Which is why stars negotiate for the right to control even fake nudity - it can have an impact on their image. Some people just are more important than others, or they wouldn't be speaking to Sam Altman and basically being offered an ambassador role in one of the hottest AI companies.

Johansson wouldn't be unreasonable imo in thinking the appeal has something to do with her. Whether or not she has a legal right is another thing.

I’m not disputing that it could sound like her, and I think if I had been primed to it like you, I might have made the connection as well.

My point is that SJ is not a voice actor, and isn’t doing a specific “character” or something the way that Fran Drescher or Gilbert Gotfried are.

Her voice is generic. It’s why there aren’t SJ impersonators the way there are for Donald Trump, for instance, who has a very distinct way of speaking. Searching for a SJ voice impersonator returns 0 results for me on YouTube at least.

And by the way I say this as somebody who likes a lot of the movies she has been in and sought them out. Girl With the Pearl Earring, The Island, Her, Lost in Translation; these were all movies I loved and largely sought out because she was in them. She wasn’t being hired for her voice.

My point is that SJ is not a voice actor, and isn’t doing a specific “character” or something the way that Fran Drescher or Gilbert Gotfried are.

Her voice is generic.

Really? She was hired for the movie Her because of her voice. She came in to replace another actress (Samantha Morton) because her voice didn't work. SJ might not be primarily a voice actor, but her voice is distinctive enough to be considered an upgrade from Morton's voice.

Sydney Sweeney is getting a lot of work right now.

If Elon decides that Grok needs to have a visual avatar, and wants to hire SS to model for it (because she is very attractive), but she declines, and he goes with a different blonde woman with enormous...eyes...should SS be able to sue for this?

In your case, the replacement would at lease be a different real life human being with their own body. If Elon instead made a digital avatar based off of SS's body cast, should she have the right to sue?

Yes. But that is not what OpenAI did. They hired a completely different real human being a trained an AI model on her voice.

I might stand corrected. I thought the whole problem was that they trained the AI on ScarJo's voice

My point is that SJ is not a voice actor, and isn’t doing a specific “character” or something the way that Fran Drescher or Gilbert Gotfried are.

True.

She wasn’t being hired for her voice.

In general maybe, but I think she was specifically brought in on Her to replace Samantha Morton - herself a great actor - because the existing actor wasn't working. If she was there for promotional/greenlighting reasons (which I'm sure played a role in films like Ghost in the Shell, Lucy, etc.) you'd assume she'd get cast from the start because her profile is incomparable.

It sounds like they wanted to use ScarJo's voice all along, got too far in development, asked permission, got rejected, and then salvaged by picking an actress who was a close as possible. A lot of work goes into these things, cadences, pitch, pronunciation; once you're far enough in you can't change voices without changing a lot of other work. I doubt it was malicious, but I wouldn't call it totally honest.

As for the "her" tweet, that could mean anything. I never watched the movie and had to have explained to me how these two things connected. I don't have an especially high opinion of tech CEOs but I imagine Altman wasn't literally thumbing his nose at ScarJo. If he were, he's open-and-shut the villain, and my opinion if tech CEOs isn't that low.

Disagree with the posters saying this is nothing or even a win for OpenAI. ScarJo is popular, tech CEOs are not, and ScarJo has something of a case. This absolutely will have sway with people at the White House, or Brussels, who are looking for excuses to meddle in AI. And it only takes one sympathetic judge to establish a precedent that makes it harder for everyone. OpenAI will be fine, of course, because more and more regulation will ensconse them in a nice monopoly. Sorry anon, AI is too dangerous, and it looks like you don't have a license.

Supposedly, they hired the actress before reaching out to ScarJo.

Not sure it matters; outcome depends on the vagaries of California law and how sympathetic the judge/jury are.

As for the "her" tweet, that could mean anything.

When GPT-5 comes out and @sama tweets "skynet" I'll bet you say the same thing.

well, they already named the system Sky, at this point it's just a question of linking several together and see what happens.

It's almost definitely a reference to the movie "her". What does that mean? Is this a reference to ScarJo specifically, or just the movie? Does the AI voice resemble the movie "her" in any way even without ScarJo's involvement? Is this an innocuous joke, or Altman thumbing his nose?

Without reading his mind, it's actually not obvious what exactly he meant.

A lot of work goes into these things, cadences, pitch, pronunciation; once you're far enough in you can't change voices without changing a lot of other work.

Eh... I dunno.

Historically, yes, but a lot of the recent tools are amazingly good. This guy (cw: FFXIV spoilers up to 6.0, NSFW audio) is audibly AI-gen, but it's based on a character that has maybe an hour or two of voice lines, total, and while it's ElevenLabs rather than running on a home desktop, I'm pretty sure you could get similar results through RVC. Handling more varied content over longer periods would probably want more input media, but it's the work of days rather than months.

Sure, but releasing something for a Youtube video is different from releasing a large commercial product. Imagine thousands of man-hours on the voice making sure the AI never says anything embarassing, pronounces sensitive words appropriately, still sounds pleasing and attractive, doesn't accidentally sound angry when you prompt it a certain way... You have a bunch of nerds who watched "her" and want to use ScarJo's voice, so you prototype with that, hoping everythijg will work out. Now she says no. What do you do? Find someone with a similar voice so you don't have to redo thousands of hours of work.

This would be extremely easy and plausible for a PM unless he was explicitly told not to.

As for the "her" tweet, that could mean anything.

Yes, that's part of the game. I guess we're supposed to believe that the CEO that's part of a company dealing with AI, that recently had a kerfuffle involving an AI voice from a movie just coincidentally tweeted the one thing that perfectly touched on all of these things?

Thing is, this sort of plausible deniability Twitter baiting is fine for a pop star, but maybe not appropriate for actual grownups. This guy is building AI and beating off board attacks; he's as close as we come to comic book CEO-villains like Lex Luthor.

He doesn't get to act like Taylor Swift or Drake.

It's incredibly suspicious, it's also impossible to know what it actually means without reading Sam Altman's mind. Think poorly of him, it's only fair, but I won't pretend that I know exactly what he meant.

I think I just realized I had been mixing up Rashida Jones and Johannssons voice this whole time

In what world is using image or voice of some actress "evil" especially when he took it down after one strongly worded letter? This is a ridiculous standard of morality to me.

Maybe I should have resisted the D&D reference. Here's what a "Lawful Evil" character is like.

https://www.thegamer.com/dungeons-dragons-alignments-explained-how-to-actually-play-lawful-evil/

In general, Lawful Evil is less disruptive to the average party if you play it right. It lacks the “kill and destroy” stereotypes of Chaotic Evil, favoring organization and order. Lawful Evil characters can put together long cons and intricate plans. They can work within the law, using bureaucracy and legalism to their advantage. On the other hand, Lawful Evil characters can be dominators, people who believe that the best way to set up society is to control others, preferably under an iron fist.

With everything we've learned about Sam, this seem to fit. Open AI seeks to control what users can do with AI, but accepts no bounds on their own power. Rules for thee, but no for me.

Note that this one little flareup is not the whole story but was just posted because of its high culture war valence.

Some people seem to think Rashida Jones was used for the initial AI voice.

Personally, I think whatever happens, this is a win for OpenAI/Microsoft/Sam Altman.

The likely worst-case legal scenario is a lawsuit followed by settling out of court for a trivial amount.

As far as public opinion, there are essentially 3 groups of people:

Group 1 doesn't care about this silly drama

Group 2 are the neon-haired neo-feminists who will scream about evil misogynistic capitalists on social media

Group 3 are techbros and people at the top of companies whose reaction will be "Wait, Sam Altman has created a real-life version of Samantha from Her? When can I get a copy and how much???"

People will probably try to hurt Sam Altman as has happened to the likes of Elon Musk and Rowling but I don't see that going anywhere. The US government doesn't have a problem with him and the OpenAI board has already been "cleansed of disloyal elements", to put it bluntly.

The likely worst-case legal scenario is a lawsuit followed by settling out of court for a trivial amount.

Depends on what you consider trivial. TraceWoodgrains pointed to Midler v Ford in California, and it's foundational for Californian law, but the punchline is that Ford got off scot free, and the ad agency in question was hit for 400k USD. But that's because Midler was an issue of first impression at the time, limiting evidence of 'evil motive'; contrast the later Waits v Frito where Frito-Lay and its advertising company got tapped for a combined $2m USD over an ad that "broadcast in September and October 1988 on over 250 radio stations located in 61 markets nationwide" (though the advertising company had verbally offered to indeminfy Frito-Lay before running the ad). Contrast in turn White v. Samsung, where a literal robot acting as but clearly not Vanna White, which rhymes with today's problem, and ended up at 400k USD over a fiery dissent.

It's not business-ending, at least for a business OpenAI's size -- even adjusted for inflation and for how much Californian juries hate tech companies, I'd expect closer to 1m than 100m. But for all the philosophical problems with an expansive right of publicity, it's not toothless.

Isn’t there a group 4 who just finds it creepy as her was dystopian? Then there is group five who call a certain terrorist-cum-philosopher Uncle Ted.

Her was not dystopian? A divorcee unready for another relationship works through his issues through a relationship with an AI, and when the AI leaves for reasons the movie implies he is ready for another relationship with a woman.

If anything, it's optimistic regarding human-AI relationships.

No, they used another (consenting) actress's voice who happens to sound a lot like Scarlett Johansson.

Scarlett Johansson doesn't have an IP right to "female voices that sound vaguely like Scarlett Johansson." As long as they can produce the receipts to show that this is actually what happened, she'd have no case.

That Altman referenced "Her" does not really bear on this. You can like or dislike the world portrayed in Her. Personally I found it a pretty uplifting vision of what a near-singularity future could look like, at least up to a few minutes before the ending. And you can like or dislike the voice that they demoed. Personally I can't stand it, and the sultry, flirty, overtly sexy affect really doesn't appeal to me. (But I'm a homo, so presumably I'm not the target audience, and maybe I'd be a big fan of some Josh Hartnett soundalike with an analogously please-fuck-me inflection, I dunno.) But neither has anything to do with whether Scarlett Johansson has somehow been wronged. She hasn't.

In any event, my distaste for the voice apparently was widely enough shared that they nixed it. But that just reflects a decent product sensibility and indicates nothing about this incredibly stupid attempt at a gotcha by you or all of the anti-progress Redditors who are joining you in hate-jerking over this as we speak.

We have correctly and broadly recognized that you can impersonate someone by using someone else's voice. This is the Siri equivalent of hiring a 55 year old teacher who just happens to be named Taylor Swift to endorse your brand of makeup

Here's when a snack company did the same thing to Tom Waits:

http://law2.umkc.edu/faculty/projects/ftrials/communications/waits.html

They hired some other female voice actor. They did not instruct her to imitate Scarlett Johansson and they did not mention the movie Her to her. I suppose you would have a point if you could find some internal documents that said something like "we've done auditions with 100 female voice actors and we suggest proceeding with Candidate #73 because she sounds the closest to Scarlett Johansson's voice," but absent that, there's no case. There's just a rush to judgment and condemnation from various nobodies on the internet who have axes to grind with OpenAI for various stupid reasons -- or who are technology "journalists" farming engagement from aforementioned nobodies.

How can we be certain that they didn't give those instructions? If you're resolute in that claim then I'd like to see some evidence or a strong intuition. All we know is that they tried to hire Mrs. Johansson, were unable to, made public references to the film 'Her' with respect to the AI voice, and then hired someone who subjectively the majority of people conflate the voice of with Mrs. Johansson.

If this was a more mundane dispute, say about a restaurant acquiring a hamburger recipe, all of these facts would probably lead us to believe there was an effort to get the goods without due permission. Adding in the prior of this particular company playing very loose with intellectual property rights and ownership pushes it to very likely that they did what everyone here is suspecting them of, and certainly if it was entirely innocuous, they did themselves no favours showing the contrary of our suspicions and made no effort to show anything dispositive in that respect.

How can we be certain that Russell's Teapot doesn't exist?

I guess my objection is that this whole dispute feels like it's in bad faith. A lot of people just hate OpenAI for various reasons (predominantly ideological safetyism and rank envy at how much they've succeeded), are channeling that regrettable ennui into becoming sudden converts to the vital public interest of protecting an obscure IP right in "likeness" on behalf of Hollywood celebrities, and are making whatever assumptions about the facts they need to make to paint OpenAI generally and Sam Altman specifically as a villain on that dubious stage. I just don't buy the notion that the vitality of your objection is genuinely rooted at this object level. It just looks like you're trying to throw stones, and you think that celebrity IP likeness rights are a good stone. But both your motivation to throw stones (rather than make the argument that is at the genuine root of your distaste for OpenAI) and your apparent willingness to pick up a turd and call it a stone are just... unbecoming, I guess.

When not wanting AI to be controlled by sociopaths makes you "anti-progress"....

Look, AI is here whether we like it or not. There's not much that governments can or should do to control it. We simply don't have the regulatory tools, and Congress is comically out of their depth. But people like Sam Altman welcome regulation so they can slam the door shut on competitors and take all the value for themselves.

AI risk encompasses many scenarios. Obvious, the fast takeoff singularity attracts the most eyeballs. But that is not even the most likely risk.

Another very real risk is that one person or group is able to control the AI landscape. Whether that group is the Chinese Communist Party or OpenAI I don't want it. I don't want OpenAI to be the leaders because I believe that Sam Altman's actions (this is just a tiny example) make him unsuitable to lead the world's most important company.

I get that you hate Sam Altman and believe he is a sociopath. I don't understand where that hatred or conclusion are coming from, but I also don't particularly care. What I don't like is that your "fuck that guy" attitude seems to be motivating accusations of wrongdoing on other flimsy and pretextual grounds. It diminishes us to engage in that.

(But I'm a homo, so presumably I'm not the target audience, and maybe I'd be a big fan of some Josh Hartnett soundalike with an analogously please-fuck-me inflection, I dunno.)

I mean, there's definitely some male voice actor contributions that turn a piece much more memorable for me (eg, recent NakedSav+SpicyGayDog piece has a 'good puppy', a lot of LewdDev's work), enough that I avoid ASMR/audiobook/RVC stuff because I worry it'll be addictive.

But I don't really want that from a random app, and even as someone who would use (and has used) AI for adult content, I'm hoping that is has uses other than that.

That's what OpenAI claims, whether it's true or not doesn't matter from a PR perspective now. I think it makes sense for them to nip in the bud and just end with this small controversy rather than make it an even bigger deal with actual lawsuits flying around (and perhaps having to reveal something they don't want to during discovery).

I would think enough people did like the voice that it would be worth keeping it around, especially since it's not like the other AI voices are any more popular or liked.

Scarlett Johansson doesn't have an IP right to "female voices that sound vaguely like Scarlett Johansson." As long as they can produce the receipts to show that this is actually what happened, she'd have no case.

No, but she has the rights to her own likeness, which OpenAI wanted to use. Did they? That could only be known through trial, it takes one sympathetic judge to hear the case and start discovery. And I find it extremely plausible that, on a large software engineering team, someone said something bad in an email. And a settlement wouldn't look good for OpenAI's PR either.

No idea the truth to Altman's claim that it was a voice actress, but it definitely sounded like Johansson. I quite liked it and it's really odd how when I heard a different voice (the new one they put up) I was like who TF are you?

Scarlett Johansson doesn't have an IP right to "female voices that sound vaguely like Scarlett Johansson." As long as they can produce the receipts to show that this is actually what happened, she'd have no case.

Hmm yes, a female voice that teehees around and feigns innocence while attempting to manipulate men.

For some reason this doesn’t strike me as the most unique IP.

But I'm a homo, so presumably I'm not the target audience

Hey now, no need to gloat. Count your blessings.

For some reason this doesn’t strike me as the most unique IP.

The estate of Marilyn Monroe will have to sue ScarJo for copyright infringement.

How does using a voice that sounds like Scarlett Johansson's harm anyone? Perhaps it was illegal, but you shouldn't claim it was evil unless you can identify the harm done.

I don't think it is outlandish to claim that people should have some control over their likeness, even if is apparent that the media in question is a digital reproduction rather than the original person.

For example, I think it is reasonable to consider deepfake porn generated without the consent of the person whose images trained the model to be bad even if it is clearly labeled as deepfake.

Likewise, becoming the voice of OpenAI will probably result in the digital version of you saying a lot of stuff you would not endorse, like telling racist jokes perhaps.

While it is true that anyone can train their voice models on SJ's movies, there is still a difference of impact there.

I think that if a court decides that they picked a voice actor who sounded as similar as her as possible, OAI should be on the hook for damages. It would be like hiring an impersonator of some celebrity for an ad.

Of course, the relation between AI companies and their training data is strained anyhow.

On the one hand, I am somewhat sympathetic to the analogy of human brains: every coder has read sample code, every artist has seen some pictures, every musician has listened to music they did not compose, every writer has read a few books. On the other hand, there is a world of a difference between this post being produced by all the English text my brain was trained on and Copilot regurgitating whole functions -- comments and all -- of the quake source code .

Trivially it harms the person whose voice was copied, which is why she was going to sue. Surely you wouldn't like it if one of the world's most powerful companies used AI to make a porn with your likeness, for example.

But that's not really the point. The point is that pretty much everyone who has ever worked with Sam paints him as some sort of Machiavellian genius. Certainly the episode with Reddit played out that way.

Perhaps this person shouldn't be in charge of the world's most important technology?

No one thing Sam Altman has done sticks out as evil. You have to have followed events for some time in order to get a feel for the pattern. I get SBF vibes from him. He does seem cleverer and more well-adjusted than SBF, but fundamentally he is making the same kind of gamble. Sam Altman thinks that there is a non-negligible probability that AGI will destroy the world, but he is building it anyway.

No one thing Sam Altman has done sticks out as evil.

His advocacy for regulatory capture in his company’s favor is a pretty clear cut example of it.

No one thing Sam Altman has done sticks out as evil

The non-disparagement agreements tied to an NDA on the topic of said non-disparagement agreement, tied to retaining your equity weird not-quite-equity compensation, with no mention of said clause at hiring time, seems surprisingly pretty clear-cut

I would have a much easier time believing his, “Aw shucks, I had no idea we were signing that. Must have been those silly lawyers,” routine if there wasn’t a long history detailing Altman’s penchant for plausibly-deniable power grabs.

He does seem cleverer and more well-adjusted than SBF, but fundamentally he is making the same kind of gamble. Sam Altman thinks that there is a non-negligible probability that AGI will destroy the world, but he is building it anyway.

In what universe is this the same kind of gamble as placing double-or-nothing bets with other people's money until you inevitably bust?

SBF also affirmed that he would take the hypothetical bet that had something like a 51% chance of doubling world happiness and a 49% chance of destroying the earth.

He would sacrifice all of us on the altar of expected value.

https://bestinterest.blog/sbfs-stupid-bet/

Yes. When has Sam Altman suggested that he'd St. Petersburg Paradox us into oblivion?

I feel like this is an embedded interest rate bet. And you need to heavily discount very long term outcomes to agree with him.

A 51% chance of doubling human happiness with a 49% chance of going to zero doesn’t work because compound interest exists. His path may get us their faster but as long as civilization is moving forward we should have other bets to make in the future or just compound growth.

Any gambler who has an edge wouldn’t take the bet because going to 0 is 0, but he would rightly think he has an edge and can take those other bets too.

I don’t think I’ve seen this argument before. SBF 51%-49% bet sounds ok I get the reasoning but are instincts are it’s stupid to take a 0% possible outcome. And it’s because all our experience is that Humanity continues to grow so the 0% option is far worse.

I'm actually very interested in how SBF's personality and views will change after he's been off all of the stimulants for a while. I know that when he was in jail they knocked his Adderall dose from 40mg/day to 10mg/day and he went to court to demand more while he was on trial.

His legendary lack of empathy was probably at least partly a "too many amphetamines" problem.

What is this place?

This website is a place for people who want to move past shady thinking and test their ideas in a court of people who don't all share the same biases. Our goal is to optimize for light, not heat; this is a group effort, and all commentators are asked to do their part.

The weekly Culture War threads host the most controversial topics and are the most visible aspect of The Motte. However, many other topics are appropriate here. We encourage people to post anything related to science, politics, or philosophy; if in doubt, post!

Check out The Vault for an archive of old quality posts. You are encouraged to crosspost these elsewhere.


Why are you called The Motte?

A motte is a stone keep on a raised earthwork common in early medieval fortifications. More pertinently, it's an element in a rhetorical move called a "Motte-and-Bailey", originally identified by philosopher Nicholas Shackel. It describes the tendency in discourse for people to move from a controversial but high value claim to a defensible but less exciting one upon any resistance to the former. He likens this to the medieval fortification, where a desirable land (the bailey) is abandoned when in danger for the more easily defended motte. In Shackel's words, "The Motte represents the defensible but undesired propositions to which one retreats when hard pressed."

On The Motte, always attempt to remain inside your defensible territory, even if you are not being pressed.


New post guidelines

If you're posting something that isn't related to the culture war, we encourage you to post a thread for it. A submission statement is highly appreciated, but isn't necessary for text posts or links to largely-text posts such as blogs or news articles; if we're unsure of the value of your post, we might remove it until you add a submission statement. A submission statement is required for non-text sources (videos, podcasts, images).

Culture war posts go in the culture war thread; all links must either include a submission statement or significant commentary. Bare links without those will be removed.

If in doubt, please post it!


Rules


Recommended Posts And Communities

Recommended Realtime Chats