This 'College Protester' Isn't Real. It's an AI-Powered Undercover Bot for Cops

>I-powered bots across social media and the internet to talk to people they suspect are anything from violent sex criminals all the way to vaguely defined “protestors” with the hopes of generating evidence that can be used against them.

so what if the bot radicalizes them?

Then real cops step in and help the radicalized folks plan something illegal, then arrest them for planning it before they can carry it out.

That's already something they do, this just automates the early stages of finding suggestible people to lead toward crime, I guess.

The cops get an arrest, win/win.

The FBI already grooms young men and provides fake explosives for them to do terror attacks, so they can arrest them. They start by talking to them online so all this is doing is making the process cheaper and larger scale

For reference, this is literally something the FBI has been known to do.

Here are some examples, I have read of a bunch more: https://www.theguardian.com/world/2011/nov/16/fbi-entrapment...

"Yay, we get to invoke the Insurrection Act!"

https://www.npr.org/2020/06/01/867467714/what-is-the-insurre...

From a purely law enforcement perspective this blows my mind. Rather than fighting crime, they are generating crime to them fight it. It would be like an SWE intentionally creates a bug and then fixes it in the name of "making the system bug free"

Can you imagine what would happen if we used the same resources to talk people down instead of rile them up?

Some people get sent down a dark path and finding someone to pull them up out of it can really help.

Instead I'd guess that these programs can likely drive them deeper and over the edge.

if your kpi was "bugs fixed a month" and your pay was directly set based on that...

I think it's an old joke, right? spend the morning introducing ones and then the afternoon fixing?

If you view the world in black-and-white with "Good" people and "Bad" people, then this makes it easier to ensnare the bad people and won't affect the good ones.

(Not a viewpoint I agree with)

If you imagine a scenario with a known murderer and you have two options:

* Wait until he does a murder, and then try to capture him AND prove it was him

* Seduce him into planning a murder and then arrest him before he carries it out

The second seems desirable, given the "known murderer" part. And once you've setup something to do that, it becomes very easy to feed others into it.

Whatever gets the budget increased in their view. FBI has long been a political operation too, e.g. COINTELPRO.

That would be a really decent application of AI.

We already have the beginnings of "AI therapists." Not sure how well they'll work, but they probably won't make people's pathologies worse.

As opposed to just about Every. Single. Online. Social. Network.

There's just waaaay too much lovely money to be made, by feeding people's ids.

FBI was created when Teddy Roosevelt used the Secret Service for political reasons and got banned from using them for most stuff. He then created the FBI by moving over all the SS agents he previously had targeting political enemies. It's always been a political operation.

>Can you imagine what would happen if we used the same resources to talk people down instead of rile them up?

The same exact persistence of the establishment and status quo but with less violence and crime and political unrest to justify their budgets with, hence why they're using it to radicalize and not calm down people.

I'm gonna write me a new minivan this afternoon!

https://devhumor.com/media/dilbert-s-team-writes-a-minivan

> The second seems desirable, given the "known murderer" part. And once you've setup something to do that, it becomes very easy to feed others into it.

If you haven't done the first, how do you get to the conclusion they are a "known murderer"?

Like the MS employees that wrote books on proprietary APIs that are otherwise undocumented an undecipherable.

>to talk people down instead of rile them up

Well then, police budgets would go down! And if police budgets go down then you are less safe!

Having no crime is more dangerous for politicians then some baseline of crime. If there is no crime they can't run on a hardball anti-crime platform and ignore everything else. If you run into a situation where there is no crime, it's easy enough to go invent some, generally focused at the young and poor.

This is the Shirkey principle.

"Institutions will try to preserve the problem to which they are the solution"

Another way to interpret it is via the savior complex or self licking ice cream cone.

This is minority report bullshit.

If you have a known murderer that is free presumably he's already paid for his crimes, no? So luring him into doing it again is extremely anti-ethical to me.

There's no pre-crime... yet. This is where you get profiling and abuse of power everywhere.

IDK how we can say manipulating people into commiting crimes is a good option. That's a crime. If we're going to commit crimes in the name of "preventing crimes" then why not go and arrest/kill the suspect directly? It's simply a different crime, right? But since I have the monopoly on violence, I can do what I want and case closed.

How can this be a good option?

I'd say it's Psycho-Pass more than Minority Report. They're fishing for latent criminals.