Training machine-learning systems to recognize human actions requires showing them myriads of video examples. New research shows “synthetic” videos generated by a computer can help — while avoiding some of the privacy, copyright, and ethical issues of real data.

In the New York Times, economists Daron Acemoglu and David Autor discussed the forces driving working-class voters toward the Republican Party. “Elites are making choices that are not good news for non-college workers,” Acemoglu said.

Neuroscientists often use AI systems to model tasks the brain performs, but some researchers are urging caution. A new study suggests neural networks do not necessarily generate useful predictions of how the brain works.

Via @MIT on Twitter: “@mrobinsnow PhD ’17 has been named to the EBONY Power 100. Congratulations!”

The human brain is infinitely fascinating. With her 2021 book “Smarter Tomorrow,” author and consultant Elizabeth Ricker ’07 wants to remind us that it is also infinitely varied — and perhaps, infinitely modifiable.

A new machine-learning model simulates how a listener would hear a sound from any point in a room. “I think this work opens up an exciting research direction on better utilizing sound to model the world,” Yilun Du says.