A new training tool for artificial intelligence (AI) can sniff out lies, according to a report from science website Futurity.org.
This tool helps AI programs recognize situations where people might be tempted to fudge the truth for financial gain, like applying for a mortgage or lowering insurance rates.
“Many businesses use AI to make decisions,” explains Mehmet Caner, co-author of a research paper on the topic. “These AI systems often rely on statistics alone. But that can backfire when people have a reason to lie, like getting a bigger loan or cheaper insurance.”
Caner, a professor of economics at North Carolina State, said: “Our goal was to see if we could tweak AI algorithms to account for this incentive to lie.'”
“We wanted to see if there was some way to adjust AI algorithms in order to account for these economic incentives to lie,” Caner said.
How can AI tell if someone is lying?
To address this challenge, the researchers developed a new set of training parameters that can be used to inform how the AI teaches itself to make predictions.
Specifically, the new training parameters focus on recognizing and accounting for a human user’s economic incentives. In other words, the AI trains itself to recognize circumstances in which a human user might lie to improve their outcomes.
In proof-of-concept simulations, the modified AI was better able to detect inaccurate information from users.
The researchers are making the new AI training parameters publicly available, so that AI developers can experiment with them.
Source: Futurity. https://www.futurity.org/artificial-intelligence-lying-3194312/
READ MORE AI NEWS.
