The AI Now Institute of New York University is sounding the alarm about the dangers of artificial intelligence in the workplace. But unlike the typical sci-fi scenario of accelerated sentience, the researchers are concerned not about how well it works, but about how well it doesn’t.

In a December report, researchers at the AI Now Institute called for a ban on so-called “affect recognition” technology, which purports to use machine learning to analyze facial expressions to detect certain emotions and/or character attributes. The problem? It doesn’t work.

From the report: “Regulators should ban the use of affect recognition in important decisions that impact people’s lives and access to opportunities.”

Despite the lack of evidence of success, affect recognition software has been adopted by companies in hiring searches, by law enforcement surveilling suspects, prisons watching inmates, and casinos keeping tabs on customers. The AI Now Institute’s report cites several studies indicating potential harm such tools can inflict on people, including one study using photos of NBA players that consistently gave negative scores to black players, despite how often they smiled or showed outward positivity.

The researchers were blunt in their overall assessment. “There remains little to no evidence that these new affect-recognition products have any scientific validity.”

Possible Preaching Angles:

The promises of technology cannot compare to the promises of an omniscient, omnipotent God. We must not become too enamored of every new tool or trick available, but rather, with maturity and discernment, weigh the merits of anyone's proposal by their track record and the fruit of life they have on display.

Source:

Douglas Perry, “Emotion-recognition technology doesn’t work, but hiring professionals, others are using it anyway: report” The Oregonian (12-16-19)