Hacker News Reader

Comment by swayvil | Hacker News Reader

Your objections are that training is required and that where and how the lie is uttered might bear.

Mere technical hurdles IMO.

I would presume that someone being subject to a lie detector may have different incentives than those running the lie detector, and they may intentionally taint the data.

I think the point is, who would deliberately train a system to detect their own lies?

> who would deliberately train a system to detect their own lies?

Maybe an employee required to do so as part of on-boarding to a corporation or government org?

Or maybe someone in the process of a passport or visa application?

This podcast episode [0] with Sean Carroll and Nita Farahany scared the crap out of me on this topic, it seems inevitable.

This is the time to regulate neuro access prior to it becoming big business, and part of the TSA screening process.

[0] https://www.preposterousuniverse.com/podcast/2023/03/13/229-...

It still seems trivial to game. Make up some "tell" for your lies. Maybe every time you need to lie during the training period you squeeze your muscles, or think of hard math problems. It would be very hard for the AI not to fit to this fake "tell."

Those sound like techniques to fool a polygraph.

There are new infrared sensors in development that should see much more complex behavior in the brain, than say EEG. I’m sure even those could be gamed theoretically, but it will clearly become increasingly difficult as the technology improves.

I was thinking it might be something everybody does in highschool or something. Every morning you spend an hour training your AI shadow. Everybody gets one. So useful, like a cell phone.

Reply on news.ycombinator.com