Civilian AI Is Already Being Misused by the Bad Guys | Hacker News Reader

Civilian AI Is Already Being Misused by the Bad Guys

spectrum.ieee.org

2 pointsrbanffy4 comments

   > Last March, a group of researchers made headlines by revealing that they had developed an artificial-intelligence (AI) tool that could invent potential new chemical weapons. What’s more, it could do so at an incredible speed: It took only 6 hours for the AI tool to suggest 40,000 of them.

 Why is this not classified research?

Well, of those 40,000, how many of them actually work? You don't know. The AI thinks they work; is the AI correct?

How many of them can be manufactured? How many of them can be deployed without killing friendly troops?

There's a whole lot more that goes into creating an actual, usable weapon than finding chemicals that are deadly.

Because the AI model used is essentially a drug discovery tool. They ripped out the optimization criteria and replaced it with negated ones, then let it rip.

This news is months old.

Also shame on you. Classifying is the most BS answer out there. Oh, lets just do skulduggerous and keep it a secret from the proles... Great idea. This can only lead to good and ethical things.

Good point, researchers who did this research should have never done it in the first place if it is relevant to making chemical weapons. If it is irrelevant then ok.

Reply on news.ycombinator.com