Candice Odgers

parental controls, content moderations, and filters.

And so on day one of the ban, you just made a situation worse.

You took all of the safety protocols that were in place based on the accounts, and you threw them out.

And I was previously a professor in a policy school, and that's really impressive.

Usually you have to do two or three steps after the policy to see where things go wrong or where there's iatrogenic effects.

But here, kind of on day one,

And then after that, they've been following young people and to see what they actually do.

And it turns out the vast majority, so over 70% are still on the platforms.

Others are probably in less regulated spaces.

They decided that what you have to do is you have to violate the privacy rights of everybody to figure out who's 16 or not.

So it's not just the kids that have to upload their data, it's all the adults.

So now we're saying, let's give all of our data over to these companies that have shown that they're not very good at not monetizing it and they're not able to protect it or not motivated to do that.

And so we do that.

And then they had the idea that they would use facial recognition, right?

So facial recognition to predict how old

Somebody is.

And so we both, you know, have teenagers and they went through middle school.

So can you imagine trying to go into a middle school and predict whether a 13 year old is 11 versus.

So, yeah, right.

So there are many problems that computer vision can solve.