There’s a reason we don’t know much about AI; Politico, September 16, 2019

Arthur Allen, Politico; There’s a reason we don’t know much about AI

The U.S. government used to study new technologies to flag ethical or social problems. Not anymore. 

"Though many university professors and tech-funded think tanks are examining the ethical, social and legal implications of technologies like Big Data and machine learning, “it’s definitely happening outside the policy infrastructure,” said John Wilbanks, a philosopher and technologist at the Sage Bionetworks research group.

Yet these technologies could have profound effects on our future, and they pose enormous questions for society...

As it happens, there is a good precedent for the federal government stepping up to examine the ethical and legal issues around an important new technology. Starting in 1990, the National Institutes of Health set aside 5 percent of the funding for its Human Genome Project for a program known as ELSI—which stood for the ethical, legal and social implications of genetics research.

The ELSI program, which started 30 years ago, “was a symbol that NIH thought the ethical issues were so important in genomics that they’d spend a lot of money on them,” says Isaac Kohane, chief of the Harvard Medical School’s Department of Biomedical Informatics. “It gave other genetics researchers a heads-up—police your ethics, we care about them.”

ELSI’s premise was to have smart social scientists weigh the pros and cons of genetic technology before they emerged, instead of, “Oops, we let the genie out of the bottle,” said Larry Brody, director of Genomics and Society program at the National Human Genome Research Institute."