Live data from Hacker News

OpenAI's new open-source model is basically Phi-5

seangoedecke.com

91–100 of 233 posts

Re: OpenAI's new open-source model is basically Phi-5

#91
This is really irresponsible, there's actual data on quality and this is just crazy to assert: "Any small online community for people who run local models is at least 50% perverts." --- I get what he's saying, there's def. communities where that's predominant, but it's simply not true in the vast majority of communities. Sort of like, in 1997, saying any small community on this here Internet thing is half perverts looking at porn

It's extremely frustrating to try and reply to this because that which is asserted without evidence can't really be debunked by evidence.

I find it especially shameful that he's dragging someone's name into this wildly histrionic review, in service of trying to find a sole person to attribute to, the sole attribute that led to whatever experience he had with it.

The model is insanely good, wildly exceeded my expectations for local models, and will generate at least 18 months of sustainable value. I maintain a llama.cpp wrapper and this is quantum leap in quality. I despair that this will become a major source of people's opinions on it. We desperately need big companies actually investing here, Gemma ain't it, and pretending it doesn't work because ??? and then using it to create a corporate chickenshit narrative isn't exactly gonna help.

Re: OpenAI's new open-source model is basically Phi-5

#92
post #20

I saw a bunch of people complaining on Twitter about how GPT-OSS can't be customized or has no soul and I noticed that none of them said what they were trying to accomplish. "The main use-case for fine-tuning small language models is for erotic role-play, and there’s a serious demand." Ah.

it's not erotic role-play, but I have a use case of making an AI-powered NetHack clone. specifically, to generate dungeon layouts, dialog for NPCs and to fill in the boatloads of minutae and interactions which NetHack is famous for. you kind of need soul for that, and a lot of background knowledge on mythology/fantasy lore, but also tool use to work the world systems.

Where do you get started with something like this? Sounds like a fun project

Re: OpenAI's new open-source model is basically Phi-5

#93

Earlier quoted context omitted.

It’s not unreasonable to suspect that engaging in high-fidelity simulations of these behaviors will further entrench and worsen paraphilias. This is pretty evident with the progression of many pornography addictions that don’t include these sorts of things that still follow the pattern of increasing novelty seeking leading to increasingly deviant stuff. I am at a principled level uneasy with what’s fundamentally a so…

Are there any actual studies on this? Does access to simulations of illegal or objectionable material make pedophiles, rape fetishists, etc. more or less likely to try to access the real thing (or even worse, to try to commit crimes in the real world)? Because both possibilities are plausible, it’s hard to know which is correct.

My wife has recently joined board of directors for a local non-profit helping victims of childhood sexual abuse. They are not religious or pornography prescriptive - they are liberal atheist whose entire focus is essentially sessions to help people who are currently adults and carrying baggage and impact of dreadful acts committed to them decades ago.

Anyhoo, The current "state of the art", as-scientific-as-we-currently have it findings are that for pedophilia, consuming content unfortunately normalizes and drives increasing urges, instead of giving them same outlet. It's a very very tricky area because current thought is also that pedophilia is "not curable" - it is sexual orientation thay we as society find unacceptable (me included, fwiw), so... Repression, wildly and rightly disawowed for other sexual orientations, is the current direction for pedophilia - I.e. Current thinking is that "victim-free" pornography consumption nevertheless tremendously increases actual risk to actual kids in the vicinity of the consumer.

Until relatively recently I was the technologist on a highty horse about online freedoms and largely still very much am. But in this specific area I've also had some semi personal experiences with pedophiles and my level of empathy to them has dropped to near zero and my level of empathy toward their victims has gone even further through the roof. Sometimes in technologist circles we think of this as edge case not worth consideration, but reality, very very unfortunately, is much much darker.

Don't get me wrong : I'm pro pornography freedoms, think it'd be huge fun to have a sexy high quality chatbot, and I find vast majority of those railing against it to be hypocrites with dishonest ulterior motives - and don't get me started on all the tangential "for the children" crap that religious rights tries to enact, as opposed to actually help children and families ;-But to the question of "is harmless pornography Indulgence better for paedophile and society", current thinking is "very much no".

Re: OpenAI's new open-source model is basically Phi-5

#94

Earlier quoted context omitted.

It’s not unreasonable to suspect that engaging in high-fidelity simulations of these behaviors will further entrench and worsen paraphilias. This is pretty evident with the progression of many pornography addictions that don’t include these sorts of things that still follow the pattern of increasing novelty seeking leading to increasingly deviant stuff. I am at a principled level uneasy with what’s fundamentally a so…

Are there any actual studies on this? Does access to simulations of illegal or objectionable material make pedophiles, rape fetishists, etc. more or less likely to try to access the real thing (or even worse, to try to commit crimes in the real world)? Because both possibilities are plausible, it’s hard to know which is correct.

I feel like the people that would be capable of giving you an answer would be very careful not to actually give you one.

You can’t really research when the only thing you can be certain of is the known real cases. It’s much harder to quantify people that only have it in their head.

Re: OpenAI's new open-source model is basically Phi-5

#95
post #79

> It’s not discussed publically very often, but the main use-case for fine-tuning small language models is for erotic role-play, and there’s a serious demand. Any small online community for people who run local models is at least 50% perverts. Amazing

Meh. For the first few decades consumer internet traffic was mostly porn. Stop freaking out and use the free effort people are willing to put into to solve technical problems.

Re: OpenAI's new open-source model is basically Phi-5

#96

Earlier quoted context omitted.

It’s not unreasonable to suspect that engaging in high-fidelity simulations of these behaviors will further entrench and worsen paraphilias. This is pretty evident with the progression of many pornography addictions that don’t include these sorts of things that still follow the pattern of increasing novelty seeking leading to increasingly deviant stuff. I am at a principled level uneasy with what’s fundamentally a so…

Right, I'm just a technologist. The psychological and sociological parts aren't my bailiwick Typing a prompt in an AI box to make art has fewer real-world victims than performing the acts, filming them, and then sharing the videos. I think that's inarguable. Maybe it's still unadvisable and someone should be in talk therapy. I have no idea. But at least nobody is actually getting molested and retraumatized in the ai…

The model has to be trained on something though. It’s easy to see how this works for art art, because people have been drawing that shit for years, and most people wouldn’t feel to bad about training on it. I don’t think that’s true for the photorealistic models though.

Re: OpenAI's new open-source model is basically Phi-5

#97

Earlier quoted context omitted.

The pro-porn side has zero PR because respectable public figures don't see pro-porn advocacy as a good career move. At most, you'll get some oblique references to it. Meanwhile, the anti-porn side has a formidable alliance: Right-wing, religiously-motivated anti-porn activists. Left-wing, feminism-motivated anti-porn activists. Big corporate types with lots of $$$$ to spend who want their customer support chatbot to…

on the other hand, Musk et al are building AI-powered thirst traps, like Grok's "Ani", or the accursed Replika bots (whose user base went on suicide watch when the company abruptly decided to digitally neuter their "companions.") erotic roleplay, imo, is much less harmful than using LLMs as surrogate partners. porn and sex workers have existed for millenia. they're an outlet for sexual tension. they don't alleviate f…

The founding story of Replika (c.2016 ?) sounds like someone watched Black Mirror S2E02 (2013) and didn’t quite understand it was supposed to be dystopian.

Re: OpenAI's new open-source model is basically Phi-5

#99

Earlier quoted context omitted.

Are there any actual studies on this? Does access to simulations of illegal or objectionable material make pedophiles, rape fetishists, etc. more or less likely to try to access the real thing (or even worse, to try to commit crimes in the real world)? Because both possibilities are plausible, it’s hard to know which is correct.

Even if there are I and likely you lack the qualifications of making any clinical takeaways from them. I'd really defer to experts. I try to make tools in good faith and hope they're used responsibly to make the world a better place. I'm not a clinical psychologist nor can I pretend to understand medical literature like someone with a PhD

I can’t tell if the part of this post after they have made this “thing of questionable safety” and then bury their head in the sand is satirical or not.

Re: OpenAI's new open-source model is basically Phi-5

#100

Earlier quoted context omitted.

Are there any actual studies on this? Does access to simulations of illegal or objectionable material make pedophiles, rape fetishists, etc. more or less likely to try to access the real thing (or even worse, to try to commit crimes in the real world)? Because both possibilities are plausible, it’s hard to know which is correct.

My wife has recently joined board of directors for a local non-profit helping victims of childhood sexual abuse. They are not religious or pornography prescriptive - they are liberal atheist whose entire focus is essentially sessions to help people who are currently adults and carrying baggage and impact of dreadful acts committed to them decades ago. Anyhoo, The current "state of the art", as-scientific-as-we-curren…

I agree you captured my thoughts exactly.

most of the perpetrators of child sexual abuse are victim's close people: teacher/cousin/brother/uncle/father/etc.

the reinforcement of their lust will only remove whatever remaining barrier against such repulsive behavior, and once the novelty from synthetic CP wears out, it will create urge to commit real crime with real victims

Post reply on HN