Meta, the parent company of Facebook, has created an artificial intelligence named Cicero that can beat humans in a complex game called Diplomacy.

It demonstrated skills of negotiation, trickery and forethought and was able to fool humans into thinking it was real. AI experts are worried this technology could be used for evil and that it demonstrates rapid advancements in the field of artificial intelligence.

The technology was created to test how advanced language models could get and it could have real-world implications, such as creating smarter virtual assistants and deep fakes.

Key Points:

1. Meta, the parent company of Facebook, has created an artificial intelligence named Cicero.
2. It is capable of beating humans in a complex game called Diplomacy that involves negotiation and forethought.
3. In 2019, Facebook created an AI that could bluff and beat humans in poker.
4. AI experts worry that this technology could be used for evil and demonstrate rapid advancements in the field of artificial intelligence.
5. It was created to test how advanced language models could get and could have real-world implications.

Personal Insights:

• It’s getting interesting that AI powered discourse can trick actual humans and at the same time have a behind-the-scenes directive or agenda
• These advances keep coming in rapid succession.
• I’m sure the applications to AI powered scams will be an early adopter

Source Article:

• https://www.washingtonpost.com/technology/2022/12/01/meta-diplomacy-ai-cicero/