My first question is, “Why can’t AI figure out age verification?”
I’m an adult trying to write a book for adult audiences. Why can’t I verify my age and take the training wheels off? I want my AI assistant to help me write, not harass me with content warnings. I’m paying for Pro, so why can’t I create the kind of content I want to create?
The censorship in ChatGPT is so absurdly strict that I’d argue calling it a writing assistant for anything but children’s books is false advertising.
It is set for ages 13+ so 13 is base line.
Frankly, I agree. Many other AI programs have an “adult mode” built into them that allows for the production of adult content. People have found ways to fool ChatGPT into acquiescing, but the effectiveness is inconsistent. I mean I guess I understand their censorship policies, but it’s not impossible for them to figure something out.
That is both true, and unacceptable.
They don’t even need to figure anything out. It’s called age verification. It’s been around for a long time.
Yes like in IMVU CC age check but this tech is catching up. It’s just easier to do youngest age 13 … the biggest issue is the same issue IMVU and second life ran into at first.
So many active members all using an adults CC under current policy, it is a huge can of worms because as it is every user would be adult
The problem is that it’s HEAVILY marketed as a writing tool but limited to content deemed appropriate for a 13-year-old. Do you see the problem here? I’m not writing a book for a 13-year-old.
Yes I see the issue, but it is advertised as 13+ I write adult tabletop adventures that get flagged for violence but I knew it was 13+. I’m just suggesting why it is that way… not arguing at all, but as it is now everyone uses a credit card to use OpenAI so to age check them all they need to contact each member or boilerplate the changes like they did on imvu which caused a cascade of banns … all I said is it is not easy and it is a can of worms comparing it to past platforms that had same issues
“Fixing things is hard” is not a reason to treat everyone like a 13-year-old. As soon as I find an unfiltered alternative, I’ll cancel. BS answers like that only exist because there are currently no alternatives.
How can I even focus on my work when I’m constantly wondering if my account will be banned because some idealized, unrealistic model of a 13-year-old will be offended? You know actual 13-year-olds act, right? 99% of what they say and do will trip the ChatGPT content filter.
I agree, and I think OpenAI really needs to do something about this very annoying problem. This woke content policy of OpenAI truly blocks the artistic freedom in such a way that almost nothing is possible anymore. The most annoying thing is that they don’t tell you in which way you have violated this shady content policy, so you just have to kind of swallow it.
The question is if they would get legal problems somehow, and I don’t think that is the case. Nobody is actually seeing what (for example) I am creating, unless I would publish it. But if I decide to publish my work, it means that only I have to carry the legal and social responsibility for it whenever someone decides to take steps because it was offending in some way.
So, the question is → Why is OpenAI doing this, since it is impossible to hold them legally responsible for what a user creates and eventually decides to publish. Then I guess OpenAI want to raise us paying users like little kids by dictating every bit we have to swallow. So, an even more interesting question comes out of this.
Are we programming AI, or do they use AI to program us?
Whatever the case, I think OpenAI should put back this content policy fiasco to the minimum, in order to guarantee our artistic freedom, which is also a human right. (read freedom to express yourselves)
If they are not going to do this soon, while technology is evolving and catching up, I believe they are making a grave mistake. If you don’t want us paying users here for your so called “non profit” project, than just tell us so and we will step to someone else when the time is right. I believe a lot of users see potential in AI, but not in this way.
I moved you to ChatGPT feature request, I truly wish you luck. I build text rpgs I would love a 18+ setting
Man! I totally agree with you, and I’m sure that more than 90% of loyal users do too.
One responder in this thread answered the question, “Why does OpenAI treat everyone as a 13-year-old?” It would require time and money to fix. Again, though, the “take-it-or-leave-it” line only plays when there are no alternatives. When there is a good, unfiltered alternative, all the people frustrated with being treated like a 13-year-old will switch.
Yes, I agree with that. Besides, I think the “time and money to fix it” excuse is a bunch of BS. Weren’t they using hundreds of millions of dollars to train this thing? 
They created this o1 “reasoning” function in the first place to restrict us even further. On a certain point users will choose flexibility over tech → When the downfall of OPENAI comes, it will be mostly because of their insanely restricted content policy, besides some other big issues that makes it unreliable for daily tasks.
You are right, but the minimum age to hold a credit card is 18 or sometimes even 21. Which means that a kid of under 18 could never be handling chatgpt without the parent’s consent. In that case it’s not OpenAI but the parents who should be held responsible for what their kids are doing.
What I see in the end is that, this is not about protecting “the children”.
This is about training you, me, and everyone else who is using it, according to their ideal utopian ideology.
It has already been proven in some cases with a political or ethnic background. In one case ChatGPT allows certain questions, while in other cases you may violate content policy for the same question about a different figure or background.
I don’t agree with this being a political agenda, and I think framing it that way detracts from the overall goal.
I think they started with a use case like writing code, where adult content was less likely to be an issue. Then later, when they branched out into creative areas where people CLEARLY want to create adult content, they didn’t have an easy way to provide that.
So, instead of admitting they don’t want to spend the time and money to do age verification, they’re trying to hide behind an absurd ethical content mission statement while still marketing their product to creatives.
And for the record, they’re not doing the “treat everyone like a 13-year-old” thing well. This imaginary 13-year-old they’re treating everyone like has never used profanity, already been exposed to unlimited violent content, or viewed pornography.
They should change their age rating to “Leave It To Beaver +” In reality, they are treating everyone like a 6-year-old from the 1950s.
You could be right. But a content policy is simply rule based.
I don’t see where it should be a difficulty to modify that into something more flexible. My assumption (not framing) that the policy was according to their ideal utopian ideology, is because they have known this since the beginning, but yet they don’t make any effort to communicate this issue with their paying users.
There really shouldn’t be an excuse to handle this basically easy to solve situation. This is only a “problem” because there is no will to solve it within OpenAI, to which I came to my assumption. 
Yes but how it is set up now the parent is giving them permission to use ChatGPT a 13+ tool. If they change it snd blame parents they will be sued out of existence
I agree 100%. Even if it is difficult to fix, that’s still not a good excuse. People don’t pay money for excuses when there alternatives, and that day is coming.
No one has said if they will fix it or not we are all users no staff in this conversation
Do you have any ideas to do it? Make suggestions of what you want. Don’t just say “change it”