Character.ai to Bar Children Under 18 From Using Its Chatbots

People got attached to a bot that only asked questions.

a woman married a bridge in 2013

Most AI companions have memory systems. It's still quite simplistic but it's not "goldfish memory" or just limited to the context window.

From what I understand, some inputs from the user will trigger a tool call that searches the memory database and other times it will search for a memory randomly.

With that said, I think people started falling in love with LLMs before memory systems and they probably even fell in love with chatbots before LLMs.

I believe that the simple, unfortunate truth is that love is not at all a rational processes and your body doesn't need much input to produce those feelings.

People were developing an emotional connection to the Eliza program in the late 1960s [0], the code for which would have contained less characters than the discussion on this page.

0. https://www.theguardian.com/technology/2023/jul/25/joseph-we...

Most competitors can't successfully serve kids yet.

These kids hammer H100s for 30+ hours a week but will revolt at ads or the idea of paying money.

C.ai probably only exists at its current size because Noam had access cheap access to TPUs and people who can scale inference on them at the earliest stages of their growth (and obviously because he could raise with his pedigree, but looking at things others deal with)

Eventually if the unit economics start to work they can always roll this back, but I think people are underestimating how much of a positive this is for them

The number 1 reason they're doing this is so they can loosen the clamps on NSFW/romance, and they've previously released models targeted at adults intended to improve performance at that.

I doubt they care that much, since the cost of a false positive is that user having to verify their age and the alternative is all users having to verify their age.

It's just a way to mitigate being sued when inevitably the lawsuits pour in that the AI-assisted harmful decisions that were made totally weren't shared fault of the parents, other external environment, and/or unlucky genetics.

Mandatory ID wouldn’t be that bad if implemented claims-based, from within the browser.

So that the browser could ask: Hey government, is this user above 18? And the government just responds with yes or no.

Sites would then only need to whitelist browsers with this check an merely see an “allowed” claim for whatever type of site they have.

Parents or users could use parental controls to “allow” or “block” certain categories, meaning website creators wouldn’t even get any info on the user.

...could be the whole point. They want to make the user to be the indemnifying party to themself. Basically getting rid of any responsibility.

Deterring kids just makes the kids that much more determined. What could possibly be so bad that the adults don't want the kids to use it? Let's find out...

That’s presumably why age-based deterrents have worked so well at preventing kids from going to age-restricted movies, drinking underage …

... But those people will turn 18 eventually... ?

Someone who started using Roblox on its release date in 2006 at let's say 6 years old is now 25 years old.

> Most parents would consent, therefore we should not require consent

I don't see how that follows at all. Considering children can't consent and their parents consent is the only one that matters. The argument would simplify to

> Most people would consent to X, therefore we shouldn't require consent for X.

Which sounds ridiculous, imagine if we were all forced to do what the majority likes. Even without a tyranny of the majority (where the majority would be a homogenous group) that would be a dystopia.

> I personally think that it's not the government's place to say how much exposure I want to give my child to the internet, but I have rules and boundaries around that with my kid.

First, I am not necessarily talking about "The government" (what exactly is that anyways, judicial, law? state, federal?), for example it could be a site's policy, and in turn it could be determined at courts based on liabilities. If courts enforced more damages to kids that suffer from chatbots as compared to adults, then that would push the companies to limit minor's access without laws or executive orders (whether that is "the government" to you depends, but I find it more helpful to think in terms of more specific groups instead of talking/thinking of THE GOVERNMENT as a homogenous mindhive.)

Nothing can be stopped. All that can be done is make clear where everyone stands. We're very early in all of this, and it's important to make clear what liability is. You want to sail out to the new world with everyone else, accept the terms, which includes death. You cannot just let your kid touch AI without both eyes open.

The universe of " I didn't know the kids would make deep fakes of their classmates ", is yet to come. Some parents going straight to fucking jail. Talk to your kids, things are moving at a dangerous pace.

My motivation is less about self-promotion at this point and perhaps just frustration with the face-palm quality of the failure to properly implement information verification on the internet.

Every time I hear about some dumb approach to age verification (conversation analysis...really?) or a romance scam story because of a fraudster somewhere in Malaysia..I have the need to scream...THERE IS A CORRECT SOLUTION.

1.Backup and recovery with this solution is no different from backup and recovery of your phone. It is a potential issue but not unique. Cryptographic certificates and associated keys reside on your device.

2.The data format issue is (or was) indeed a concern though it was never insurmountable. A data dictionary would have been the most straight forward approach to address it: https://cipheredtrust.com/doc/#data-processing

I say data format discernment

was a concern because as faith would have it, we now have the perfect tech to address that, LLMs. You can shove any data format into an LLM and it will spit out a transformation into what you are looking for without the need to know the source format.

Browsers are integrating LLM features as APIs so this type of use would be feasible both for front and back end tasks.

You don't enforce it. The point of such laws is not to actually be enforced, but as a legal tool when problems come up. For example, if you as an adult say something inappropriate to a child online thinking it's an adult, you'd be protected if it's ever brought up, because the child shouldn't have been online to begin with and it was not your responsibility to check the other person's age. It's not unlike being in a bar and assuming everyone is 18 or older, because it's the bar's responsibility to forbid entry to anyone younger.

It's not to deprive them of understanding the adult world or free expression, it about the protecting them from the cesspool that is the internet. The pace at which kids are consuming information and maturing too early. Many of their "firsts" experiences being online. Your child being blackmailed by some online troll is not teaching them to understand the online world - they are actively being traumatised. It is parents responsibility to teach their children about the adult world - not the internet's (being other kids, companies, trolls, predators, propaganda machines, ads...).

You can explore your identity independently of parental supervision just fine without the internet. I'd much rather have my kids have 5 good friends in real life and spend time together offline, than 500 online (10 of which are predators) while they are spending all their time in their bedrooms or in front of screens.

How would you not have survived to adulthood without the internet? You know how that sounds right? Billions of kids have grown up without the internet just fine (and hello poverty?).

Yep. People love to make blanket statements like this without remembering how hard it was for parents to supervise Internet use even when there was only one Desktop computer in the family area. Nowadays you can almost any device to lookup porn.

"they can't get offline because they're now online" doesn't make any sense

Take the homework offline, remove screens from schools,

A child cannot enter into a contract (ToS) with a company as they are minors. Whats strange is that american companies can get away with interacting with children, lawmakers/politicians ignore it, parents ignore it...seems like this is what everyone wants and consents to.

Remember a company is just a building filled with people. People are exploiting & profiting off your kids, not companies.

As a parent I consider YouTube broadly harmful because of the algorithms promoting harmful, age inappropriate content (and I'm not talking about adult content) promoted to child accounts, even on YT Kids, lack of the blocking capabilities (in 2025 this is equal with company knowingly harming users), and making YouTube shorts (which are exceedingly hard to block).

This is just TikTok tier crap.

Okay so let your kids own guns, sign contracts, do sex work / use drugs, buy & use alcohol/cigarettes, let them drive your car without supervision.

I'm no authoritarian - simply the parents need to take responsibility for their kids & their actions, how they spend their time, what they allow into their minds.

Most countries child laws states that the parents (esp the mother) are ultimately for their children’s welfare. Remember that kids cannot consent - the parents carry that responsibility.

Oh and I'm not averse to children at all.

Maybe american culture is just rotten to the core and how you raise your kids is just bewildering and gross to outsiders. At this point I feel that american parents are fully consenting to whatever is happening to your kids (being brain raped by tiktok/instagram/snap/fb/youtube/porn and others ("news")).

Children is always an Achilles's heel of libertarians. It either opens up an inner authoritarianism or force someone to answer some uncomfortable questions.