Hacker News Reader

Unfortunately, the price will go down pushing the supply/demand curve out, and we'll get ever more garbage. Some of it will be dangerous or addictive to susceptible portions of society, mostly just boring and stupid to the rest of us.

Wait for first kid who dies trying an AI generated "challenge" or the first violent mob killing caused by AI generated outrage porn. AI generated video porn may look like triple breasted whores of Eroticon6 today, but with sufficient influencer content (playground videos) and porn (dungeon) footage, I suspect you can generate more than enough novel and relevant (child S&M) porn for everyone.

To play devil's advocate: If AI is producing content that would be morally objectionable because it harms someone, but nobody was harmed in the making of it, are we still right to find it morally objectionable?

In terms of enjoyment, yes. In terms of overall harm reduction, no.

It's still a bad thing for humanity at large but it may have a knock-on effect of pacifying people who would otherwise pay significant amounts of money for new content to be produced. If we can placate those people, at least the money dries up for those other sources and maybe they would move on to doing something other than harming children.

Tricky spot, to be sure.

How do you know if an AI porn character is 17 or 18?

How do you know when a porn character is 17 or 18?

You find out in court: https://nypost.com/2010/04/24/a-trial-star-is-porn/

I don’t understand why procedurally generated porn of made up humans is harmful to someone. How is this different from video games that allow you to shoot and stab thousands of virtual people without any proven deleterious effect on real life?

I think sexuality is different to violence. I've heard plenty of stories about people developing weird kinks from porn, or having to turn to weirder and weirder stuff to get off. But I haven't heard of any cases of video games turning people violent.

This is extremely unscientific, I know.

Unfortunately I don't think we have data here to make conclusions, only a few moral intuitions, which historically haven't been the most reliable compasses for navigating these grey areas.

Nah, compare violence in media by sex/age breakdown of the victim. Excluding sexual stuff completely, it's many, many orders of magnitude more acceptable to show the gory killing of a male character on screen than it is to show the same of a woman or child.

There are all of these unwritten but extremely obvious rules that dictate what is acceptable & what is not.

OK, let's assume your argument is correct. (which I don't think is true, but let's go with that for now) Let's say that media depicting some kinds of sexual material warps what gives some people sexual satisfaction.

...so?

Should my preferences of what I find acceptable and not acceptable in that domain somehow influence what content others are allowed to consume? If the creation of that content actively causes harm, then sure. That content should be restricted. If we are restricting it because you think it's icky and it might lead to other things you find icky than we took a wrong turn somewhere.

That's like saying we should outlaw mozzarella cheese because it might lead to the consumption of triple cheese sausage pizza. Neither of those acts hurt me, so why should I care about people choosing to participate in either?

The method of gratification is different from porn is different than to video games.

I don’t have to get off to enjoy obliterating people back in the quake days, but porn is a different psychological mechanism entirely.

I’m trying to assume you’re asking the question to spur conversation, and not that you actually hold the view that watching child porn, whether synthetic or not, is morally equivalent to playing video games.

Is it? Are they not all dopamine pathways in the end, with addicts emerging from all of them? Where's the evidence against that?

People spent decades arguing that rock & roll, Dungeons & Dragons and videogames would lead to moral corruption and decay. Turns out there is no evidence to support any of it in the end. The immorality was always "self-evident". That's not enough to justify banning something though as evidence has shown again and again.

Also, whose morality are we talking about here? Morality is all over the map depending on which culture you ask. Each cares more or less about sanctity, divinity, loyalty, freedom etc.

And you're correct, my personal opinion on this is irrelevant.

>I’m trying to assume you’re asking the question to spur conversation, and not that you actually hold the view that watching child porn, whether synthetic or not, is morally equivalent to playing video games.

I think you're less trying to assume this, and more trying to cast aspersions without breaking HN rules. Barely.

See I somewhat agree with you, but you have a limited perspective in a way, that maybe as a gay man I can help expand on without being shouted at for being MRA trash just because I notice things.

You say "video games allow you to shoot and stab thousands of virtual people", this isn't true. In actuality, video games allow you to shoot & stab thousands of virtual _men_. Replace the baddies in a game with women and or children and people will go mental.

Same thing with media; Television/movies - plenty of men are brutally gored on screen with no one batting an eye, but generally the camera will cut away if there's violence against women &| children. The only time you do see it is when they want to make a very strong impact ie "this guy killing this woman is a V E R Y bad guy, and you know that because it's not just some anonymous dude he's goring".

There's plenty of hypocrisy in it; Game of Thrones people famously complained about the Cersei rape scene (also in the books, although more ambiguous) meanwhile the average episode spends a good portion of the time showing us the insides of three dozen guys' ribcages.

Or Altered Carbon where in the book series where Takeshi is forced into a simulation as a woman, where he is tortured; media at the time said they were glad that it didn't make it into the show because it would be "mysogynistic torture porn" - apparently none of the gratuitous violence against male characters was considered to be bad.

And _that's_ why it's different. Because certain types of porn could be generated that infringe in the protected classes that society creates.

That all sounds about right. The point I was making is that the poster was looking at it from the standpoint of "reducing harm", whereas in reality they were merely following their moral intuitions about what's sacred and taboo in society, and then backwards-rationalizing it through the "harm" moral axis.

Very common in the liberal West where we think of ourselves as being above taboos and caring about divinity, except we still do very much. E.g. digging up the corpse of a dead relative and having sex with them doesn't harm anybody, but is highly taboo. Having sex with a dead family pet, doesn't harm anybody, but highly taboo. Siblings on contraception having sex, zero harm, but very taboo. Hentai of protagonists who could be underage, zero harm, very taboo. etc. etc.

Women, children, LGBTQ members, the involuntarily unhoused, minorities, Muslims and many others are all currently sacred groups whose sanctity you cannot violate even in a work of art and fiction.

So back to my point, it has nothing to do with harm, it's all about sanctity.

It would normalise that behaviour. There are plenty of studies that show pornography warps the watchers' minds.

Even video game violence would do it but a vast majority of the experiences are easily identifiable as cartoonish. So it may give you the idea that you slaughtered a 1000 NPCs but each kill is nowhere close to the visceral reality. The games which have a focus on realistic killing, either do so to aliens/monsters or they are so off-putting that they never manage to find a large audience.

In any case I look at the public discourse about war and violence and I find that people are eager to jump into fights and don't think twice about supporting their government in bombing and killing the shit out of other populations. They couch it in some shallow moral argument. Often the deterrent to war isn't some moral concern but the fact that the other side also has significant weaponry and might inflict casualties on your side as well.

In any case I hear a lot of this overton window concept and there is a lot of merit in the argument that excessive amount of AI generated deviant porn content will eventually make people think that all this depravity is normal and we should just look the other way.

> It would normalise that behaviour. There are plenty of studies that show pornography warps the watchers' minds.

Is there any evidence that it would normalize that behavior? That sounds like a moral intuition, not something backed by evidence. And even if the mind of the watcher was warped, is that any of your business as long as it's not increasing harm to society? In the West we let adults of sound mind harm themselves in lots of ways they choose, from eating terrible food to doing drugs to drinking to smoking to extreme sports. It's their body, their choice. As long as society doesn't have to bear the cost of it, like say with drunk driving, in the West we have accepted that adults have the right to do what they want to themselves.

> Even video game violence would do it but a vast majority of the experiences are easily identifiable as cartoonish. So it may give you the idea that you slaughtered a 1000 NPCs but each kill is nowhere close to the visceral reality. The games which have a focus on realistic killing, either do so to aliens/monsters or they are so off-putting that they never manage to find a large audience.

Great, then really deprived AI-generated porn will likely see the same lack of adoption and stay a niche phenomenon for a few people in a basement somewhere. What's the problem?

> In any case I look at the public discourse about war and violence and I find that people are eager to jump into fights and don't think twice about supporting their government in bombing and killing the shit out of other populations. They couch it in some shallow moral argument. Often the deterrent to war isn't some moral concern but the fact that the other side also has significant weaponry and might inflict casualties on your side as well.

Were people not eager to jump into fights and support their jingoistic government rhetoric prior to the creation of violent videogames? Do you have anything to back that popularity of violent videogame led to a rise in ultra-nationalism and aggressive foreign policy?

> In any case I hear a lot of this overton window concept and there is a lot of merit in the argument that excessive amount of AI generated deviant porn content will eventually make people think that all this depravity is normal and we should just look the other way.

Is there a lot of merit to that argument? How did you decide that?

I think it's important to differentiate something that is personally icky to you and something that is 'bad for humanity at large' unless you can demonstrate how it's bad for humanity at large and demonstrate why that makes it morally objectionable.

Morally objectionable acts require (IMO) a victim.

I think it's essential that before we decide that we are going to limit what people can do we determine if what they're doing is actually hurting someone to the extent that society should limit that behavior or if we're just limiting their behavior because it's different from ours and we don't like it.

In my personal moral worldview, if a person can simulate a universe with no actual people in it in which they commit whatever debauchery they feel necessary without causing harm to anyone else, I say have fun. It has literally no measurable impact on me, so it would be immoral for me to insert my personal preferences into the matter.

If your 10 year old dies because of an AI created Tide-pod or Street-surfing challenge, I think you'll find it objectionable. Same for highly targeted fake news clickbait that causes a race riot.

Why does said 10 year old child have unfettered access to the internet?

Also why is an AI ultimately responsible for a child choosing to perform some challenge? What if the 10 year old child played amogus & then decided to re-enact irl?

I'd say it's less the source of a challenge or false factoid etc and more a cultural problem of no monitoring kids enough; parents give their kids phones & let 'em use TikTok to their heart's content cause it keeps the kids quiet. And immature kids love TT because it's easy to generate clout and therefore dopamine.

> Why does said 10 year old child have unfettered access to the internet?

Q: Of the kids in my 13 year-old's class, what % would you guess have WhatsApp on their phone?

Spoiler: "If you live in a country in the European Economic Area (which includes the European Union), and any other included country or territory (collectively referred to as the European Region), you must be at least 16 years old (or such greater age required in your country) to register for and use WhatsApp"[0]

[0] https://faq.whatsapp.com/695318248185629

Why does it need to be unfettered? They have peers. Kids smoke/drink and it's literally illegal to sell them cigarettes/alcohol.

A significant percent of adults are also susceptible to crazed conspiracy theories... see QAnon. Now allow AI automation to target individuals and small groups to "optimize engagement" with apparently personal communications using A/B statistics. Everywhere all the time, because it's cheap, because it pays the rent. Some of them will be drawn to artificially generated violence inciting agit prop, because it works. There are negative externalities to that.

Why this.. why that..

My 10 year old child just got fucked and I deserve vengeance. I want to see heads roll and I want to see the world burn.

So...someone was harmed? How is this related to my question?

Most of the people you describe have little to no moral compass. Most of the time they are above the accepted morals of society (a very Nietzsche perspective). These are the marketers of the world who encourage you to "pollute the web" and the media manipulators whose secrets are to "con the conmen". The reality is, they make more money than any of us and sleep just fine every night because they know that nobody seeks honesty or reality anymore. The more unbelievable of headlines and articles, the more it warps our compass.

No sane person doing this will push reasonableness, complexity, or mixed emotions.

> they know that nobody seeks honesty or reality anymore

It’s only true for them. Not for us.

What's the difference between not seeking, and not knowing how to seek?

From a moral perspective, a lot. From an amoral pragmatic perspective, not a lot – unless you think it'll somehow benefit you to give people the ability to effectively seek such things? Hah.

Us is them. At least I say this in a general sense that a healthy portion of posters here are in marketing, or they are looking for some way to make a living in any way possible. Trying to making this a us vs them is pretty much meaningless as it's completely ineffective in solving the problem.

> Trying to making this a us vs them is pretty much meaningless as it’s completely ineffective in solving the problem.

i don’t mean this as confrontational as it’s going to come across, but this is nonsense. it isn’t ineffective at all.

would you mind expanding on what you mean?

No, this is something I can't agree with.

The whole point of social norms is that they define some boundaries of what is acceptable and not in some community. If someone violates the social boundaries because "they are looking for some way to make a living in any way possible" (which is not an excuse, I mean, "any way possible" would also justify stealing, robbing and murder to make a living) then they themselves choose to be "not us" and deserve to be shunned by our community.

This also does have a certain effect at solving this problem - if you know that telling people what you want to do is going to result in losing reputation and refusing to assist you, then that does act as some deterrent. The social pressure reduces the likelihood that people will choose to join that industry, and it reduces the likelihood that people in the industry will refuse some activities even if they are profitable. Even from pure game theory and evolutionary psychology we can observe that 'punishing defectors' is a viable strategy that gets some results.

It is important that we do not legitimize or normalize unethical behavior just because someone is trying to make a living through marketing; so whenever someone says "ah, we're all in the same boat, isn't everyone doing this?" it's important to loudly remind everyone that no, we're not all acting like this - ethics is a thing and proper people refuse to do improper things.

> Most of the time they are above the accepted morals of society (a very Nietzsche perspective).

That sounds more like marquis de sade than nietzche imo.

i think nietzche is amoral only when what is moral is arbitrary and self deprecating.

We can hate them for choosing to pollute the world with that kind of work, but they're still human beings being tortured in a mental coal mine.

No they're not. They're exploitatively torturing other people, while deploying machines to mine the coal. They deserve any bad thing that happens to them, because they have the education and resources to do better but choose not to.

I don't care that they're human. If they have agency and resources and leverage those in such willfully zero-sum fashion as you describe, they've chosen to gamble on profiting from the suffering of others. Empathy and kindness are good things, but empathy for willfully abusive people is maladaptive.

> I think there's a bright side if people can't compete with machines on stuff like that. People shouldn't be doing that shit. It's bad for them.

I don’t know. People already pump out a ton of bullshit from content farms then litter their web pages with ads and last-click attribution.

End user value isn’t what drives a lot of “information” businesses. See any recipes site or “news” that’s regurgitating what someone “newsworthy” tweeted.

It will be interesting to see how search engines adjust. Maybe someone will make the GetHuman (https://gethuman.com/) equivalent of search.

I was just thinking about that today before reading this article. It took me like 45 mins one day to find a health site that wasnt trying to sell a product. Most of these for-profit sites conflate there content to jive with the products they are selling, which can prevent someone from finding information that actual does some good.

On one hand I completely agree, but on the other hand, from their perspective it may be "well I paid $500 for this turnkey point-and-click app and now it makes money for me in the background while I sit on my couch making music all day". This new streamlining makes it more soulless in general but less soulless for the individual people responsible for it because they're doing and seeing less of the actual bullshit themselves and deferring it all to the automation pipeline.

They may (and, frankly, should) still feel something about what they're putting out into the world, but they can more easily blind themselves to it and just tell themselves almost everyone's doing something dumb to make a living and they're not even the ones actually "doing it" themselves.

"I think there's a bright side if people can't compete with machines on stuff like that" - Hadn't thought of it like this. Good point. Perhaps it will be akin to email getting better spam filters. And perhaps there is a better way than a 3,000 word article about how long to boil rice.

In fact that's exactly what I want LLM to be doing - read the whole internet and write articles on all topics, answer all imaginable questions, make a 1000x larger wikipedia, a huge knowledge base. Take special care to resolve contradictions. Then we could be using this GPT-wiki to validate when models are saying factually wrong things. If we make it available online, it will be the final reference.

How do we know which sources contain factually right things? What happens when the facts change? It used to be a "fact" that the Sun revolved around the Earth, and that stomach ulcers were caused by stress...

The model can describe the distribution of answers and their confidences. So there will not be one right answer for everything.

Nonsense. There is no reliable way to model confidence on such issues.

That's why repeatable experiments are so important to science. Anybody can independently verify a testable claim.

You shouldn't be so confident without knowing how these things work, there is a absolutely a simple and built-in way to model this... LMMs for example are simply calculating the next word or phrase sequence that is most likely given previous results and modeling information. So they can definitely tell you the combined liklihood that the answer is 'Peru is a cat' vs 'Peru is a country' and provide you the exact statistical likelihood of each.

Actually I am confident exactly because I do know how LMMs work, and your comment fails to address the issue at all. Such models can't tell you anything useful about the probability that a particular statement is accurate.

"So they can definitely tell you the combined liklihood that the answer is 'Peru is a cat' vs 'Peru is a country' and provide you the exact statistical likelihood of each..."

...in the context of the texts that the LLM is built on. Not in the context of the real world, where P('Peru is a country') = 1.0 and P('Peru is a cat') = #cats named Peru / #things in the world (or something).

Most of your world is text. Sure there's a sliver that isn't, but the reality you directly see is a tiny fraction of the reality you know from reports. Come to think of it, I've only ever seen reports of Peru, never actual Peru.

That's how likely they are to occur near each other, not how likely either statement is true. Rude of you to preface your comment the way you did while making this error.

parent post isn't arguing which thing is capital T true (if such a judgement is even universally possible). They are talking about modeling statistical confidence, which is purely an emergent numerical property of data and makes no commentary on objective truth.

Hahaha. Reliable confidence, isn't that a contradiction in terms? If a model makes mistakes but has great confidence estimation why wouldn't it make fewer mistakes in the first place, since it knows when it is wrong.

But if a LM looks up a topic and sees contradictory answers, and none of them is much more reputable, maybe it can still use that information to say it is inconclusive. Knowing a topic is controversial or not present in search engines is useful information. chatGPT would hallucinate, a search + chatGPT solution would refrain from hallucinations. It could also give references.

How is that different with LLMs versus badly-written human generated content? Most clickbait/SEO articles are as poorly researched as they come, and shouldn't be assumed to be accurate anyway.

LLMs don't judge truth. They judge which answer is statistically correct. There is a difference.

They judge the maximum likelihood answer. I take issue with your usage of the word "correct" because it can too easily be confused with "accurate."

The statistically correct answer is not necessarily the true one (if there is such a thing as 'truth'). Many people can believe something to be true, and if I query those people i can calculate which answer is statistically correct. That's the 'wisdom of the crowd'.

Calling a flat earth or a geocentric universe "statistically correct" at past historical points is really inane, don't you think? In doing so, you abuse the notion of what statistics is supposed to represent, which is, generally, a statement of an estimate (and/or distribution), as well as the precision of that statement. Since "correct" is binary, it carries an implied precision of 100%, which renders the notion of "statistically correct" is pretty absurd.

The "wisdom of crowds" is mostly bullshit. It works fine for trivial things like estimating the number of beans in a jar. So what. It completely fails for anything requiring deep expertise.

Seems like this is what you get when you have programmers try to be statisticians, I suppose.

"Statistically correct" gobbledygook that signifies nothing.

Even humans use few relatively simple heuristics to decide what to trust.

One is that objective truth is internally self-consistent. If one AGW denier claims it's the sun, and another AGW denier claims the NASA falsifies the data, and they support each other, then you can judge these are conflicting claims and decrease your trust.

Also, false claims usually focus on attacking competing claims than to come up with a coherent alternative. And they tend to be more vague in specifics (to avoid inconsistency), compare for example vague claims about all scientific institutions faking data vs Exxon files containing detailed reports for executives.

I think you get what you pay for. Free information isn't free.

Profound

It is very nice of you to be so concerned about these folks' inner lives and psychological well being! Are you going to pay their rent, and feed them too?

Honestly I don't sympathize with either of these sentiments. If the only work people can find is by making the world a miserable place, perhaps we have too many people.

1. If "bullshit motivational quotes, carefully constructed outrage takes, or purportedly expert content about topics [the author] knows nothing about" makes your world a miserable place, you are part of the globally privileged. Unplug and get some fresh air, ffs.

2. Lots of people like that stuff. Who are you, and who are OP to decide what content gets produced and consumed? The morality police?

3. The irony of complaining about that stuff on a site dedicated to the industry that platforms that kind of stuff is just astounding. Perhaps the real problem is not the content, but the medium that allows its mass dissemination?

4. The material misery that would be created by shifting entire industries out of work (if even for a "few" years to who-knows-how-long) would be measurably greater than the micro-miseries of the kinds of things OP seems to complain about.