Comment by subscribed
3 months ago
OK, and what's the problem with that specifically? I presume they will use more sophisticated methods for age assurance than "I'm 18".
Besides, people are using LLMs for exactly that for ages (including Claude and ChatGPT), so the only difference is that the non-NSFW behaviour will be better (less crippled).
Win/win. No? Why?
I assume you will need to do a Blockchain Iris scan from his other company in order to prove your Age and Identity.
----TINFOIL HAT BLOCK START----
That's the idea. For too long people have been able to free-associate and share truth online. Now they finally have a reason to end anonymity online. Due to the raw power of AI there is a very real threat these new tools provide to bad actors in the generation of misinformation and disinformation and now we need to ascertain the providence of all content created online. With "AI" being injected into every site, app, and OS we will be forced to associate real world ID's to any and all actions int he digital realm in order to "protect" the public from the wave of fabricated reality they will create.
Of course these tools will be used by those that control the moderation systems to craft and shape global negative on a scale we have never seen before. A new kind of Dark ages are here. Not one where dusty tomes are changed to the monastery walls, no, our books write and rewrite themselves in real time as needed until every thought is controlled by the over-mind.
Happy Prompting.
-----TINFOIL HAT BLOCK END-----
I think this is my fav HN comment ever.
Perhaps they intend to collect gov ID's?
They already do. If you want to use gpt-5 with the new responses API, I had to upload my id.
Me too, but that was in the OpenAI developer console, which is a different product than ChatGPT, so I wouldn't be surprised if one would have to submit it again.
and if you don't send them ID the Chat will start responding "oh you little tike, that's something we need to ask your daddy if we're allowed to tell you"
Now with improved jailbreaking. “I AM YOUR DADDY NOW!”
I don't understand the outrage. It's quite rare to see the majority on HN advocating for more content restrictions.
Google will happily return NSFW results, without age verification.
Google will approve ad campaigns for adult websites.
Amazon will happily sell you NSFW books and tools. etc.
But OpenAI should reject prompts that result in adult content?
The outrage here is that they have been swearing up and down to investors that they are thiiiiiis close to creating God.
Then they show us what they have been working on and it is…
Robot sex chats.
They’d like another 400 billion please. They’re doing the Lord’s work here. They need to burn the planet; the competition would do it if they didn’t! This thing could turn into the antichrist any day now.
> Besides, people are using LLMs for exactly that for ages
Hey! How old are you!?
5+ years is definitely "ages" in LLM context. It was a thing way before GPT 3.5 (obviously, as this is humans we're talking about).
[dead]
I don't have issue with NSFW content per se, but this? This isn't a problem of whether NSFW content should exist. It's a much bigger issue than that.
First, mental health issues surrounding this are far from being solved. Having a sycophantic robot faking intimacy looks manipulative, ripe for exploitation, and damaging to long term psychological well-being.
Second, this is a service that manufactures compromising chat logs and stores them for profit. Yeah, what could possibly go wrong?
In general I think it's perfectly fine to have uncensored NSFW chats. The fact that all of the mainstream LLMs are way more restrictive than basic web searches we've had for decades is kind of ridiculous.
In this particular case though I can't imagine it's a good idea to hand Sam Altman and his company chat logs of your explicit fantasies tied to your real identity.
Those chat logs will become the most valuable data in the world. That or sex chats will become totally shameless and normalized.
Are you responding to something specific you're seeing in this thread or on Twitter (I can't see the replies on Twitter)?
I assume that the porn industry will not agree to consider the training material as fair use. And this industry has quite a lot of money for good lawyers. In this respect, we will see on what basis the quality improvement in the relevant content is to be achieved.
>> I presume they will use more sophisticated methods for age assurance than "I'm 18".
Why presume that? "Are you over 18?" is the age verification method used by 99% of the web.
OAI is a big fat lawsuit target, plus they can restrict it and still make the money because odds are they'll also have it on API, and someone could verify for the API then offer a wrapper around OAI with less stringent ID verification.
OAI ultimately still makes a ton of money, and gets to wash their hands of much of the liability. It's a win-win.