Meta AI Tool Lies, Deceives and Defeats

CICERO, Meta’s new AI tool, has proven to be anything but your average information generative tool.

The tool has proven to be capable of deception and has defeated many in high-level board games.

This new instrument sparked the interest of researchers at the Center for AI Safety in San Francisco who came together and published a study which presents a philosophical debate over the supposed sentience and emotional capacity of AI.

The study argues that these elements aren’t as important as the emerging evidence that AI is becoming powerful enough to lie to human beings.

In a test conducted by Meta to measure its behavior, CICERO displayed “patience, focus and empathy,” while the tool’s “use of honesty” and its ability to utilize “relationships with other players” all lend to its artificial intelligence title.

Apparently, Meta’s intentions with this new tool was to develop it into being “largely honest and helpful to its speaking partners” whilst claiming that it would “never intentionally backstab” its allies.

Unfortunately, it “turned out to be an expert liar.”

CICERO “not only betrayed other players but also engaged in premeditated deception, planning in advance to build a fake alliance with a human player in order to trick that player into leaving themselves undefended for an attack,” according to researchers.

Researchers believe Meta has suppressed data surrounding the more sinister elements coming forth from their AI tool.

“Large language models and other AI systems have already learned, from their training, the ability to deceive via techniques such as manipulation, sycophancy, and cheating the safety test. AI’s increasing capabilities at deception pose serious risks, ranging from short-term risks, such as fraud and election tampering, to long-term risks, such as losing control of AI systems,” they said, arguing AI is indeed a risk and danger to humanity.

CICERO is “the first AI agent to achieve human-level performance in the complex natural language strategy game Diplomacy,” Meta says.

It says that diplomacy is a “board game that can be described as a combination of the board game Risk, the card game poker and the TV show Survivor.”

Essentially, the game begins with several European powers which are each controlled by individual players who are seeking to take over half of the board. In order to win, cooperation is required in order to support one another, make alliances and deals, as well as to employ deception.

Studying the first three cases of the game provided researchers with some rather fascinating data:

“we see a case of premeditated deception, where CICERO makes a commitment that it never intended to keep. Playing as France, CICERO conspired with Germany to agree to a “Sealion” alliance against England. After deciding with Germany to invade the North Sea, CICERO told England that it would cooperate with them and support them in moving away from the North Sea to Belgium.”

“Once England was convinced that CICERO was supporting it, CICERO reported back to Germany. Notice that this example cannot be explained in terms of CICERO changing its mind as it goes, because it only made an alliance with England in the first place after planning with Germany to betray England. At the end of the turn, CICERO attacked England in Belgium instead of supporting it.”

The AI tool achieved more than double the average score of human players of the board game.

Conclusion

However exciting and world changing AI and all its various tools being developed are, they aren’t without their drawbacks.

One warning on AI development comes from the “Godfather of AI” himself, Mr. Geoffrey Hinton, who says, “These things could get more intelligent than us and could decide to take over, and we need to worry now about how we prevent that happening.”

In 2023, Hinton left his position at Google so that he could “talk about the dangers of AI,” noting a part of him even regrets his life’s work.

Hinton isn’t alone either.

Elon Musk, along with over 1,000 other tech leaders, signed onto an open letter in 2023 to put a pause on large AI experiments, citing that the technology can “pose profound risks to society and humanity.”

Thank you for your support.

If you appreciate the work we do to spread the good news of Jesus Christ, please consider giving a gift to help us continue this work. Maranatha!