Meta Teaches an AI to Lie, Strategize

Saturday, November 26, 2022

An AI taught to play a boardgame that involves negotiating with human players and inferring their motives could have applications for enterprise chatbots, Meta says.

Posted by LampLighter at 09:33 AM | 5 COMMENTS | permalink | Comment on This Entry |

Alternate links: Google News | Twitter

Comments

Admin's note: Participants in this discussion must follow the site's moderation policy. Profanity will be filtered. Abusive conduct is not allowed.

More from the article...

...Meta has trained an AI agent to play a boardgame that involves chatting with other players to persuade them to support its strategies " and then betraying them.

The company, which owns Facebook, Instagram and WhatsApp, says that its Cicero AImay have widespread applications in the near future including developing smarter virtual assistants with the combined use of technologies such as natural language processing (NLP) and strategic reasoning, according to a blog post released by the company. ...

#1 | Posted by LampLighter at 2022-11-25 07:49 PM | Reply

So... there's an AI that lulls you into a false sense of security, and then betrays you.

When will we be equipping AI with guns?

Oh, wait...

San Francisco police seek permission for its robots to use deadly force
drudge.com

#2 | Posted by LampLighter at 2022-11-25 07:53 PM | Reply

Meta teaches an AI to lie, strategize.

For the last time, his name is Mark Zuckerberg.

#3 | Posted by censored at 2022-11-25 11:00 PM | Reply

Could this AI feed information to police to de-escalate since they aren't "equipped" for human-to-human interaction with least horrific consequence? What Would Officer Zuckerberg Do?

#4 | Posted by redlightrobot at 2022-11-26 12:13 AM | Reply

Who knew someone would use Donald Trump as a model for AI design. Someday that bot could be President.

#5 | Posted by danni at 2022-11-26 09:49 AM | Reply | Newsworthy 1