Ted Berger squints through a microscope at a slice of rat brain the size of an infant’s fingernail. It’s resting on an array of microelectrodes, which eavesdrop on the murmurs of nerve cells. Berger, a biomedical engineer at the University of Southern California, wants to understand the cells’ language because he is designing a computer chip that might one day bolster the brain’s memory banks.

The most immediate beneficiaries would be victims of stroke, Alzheimer’s and other ailments. Healthy humans might one day use such a chip to bone up on French, quantum mechanics or an F-16 flight maneuver. “It’s down the road,” says Berger, “but the building blocks are being put together now.”

After a decade of stimulating nerve cells and observing their responses, Berger has devised computer programs that replicate the cells’ behavior, and has built chips to run them. Earlier this year he showed (in a petri dish) that a chip can replace rat brain cells in a neural circuit. He hopes to test a chip in live rats within three years, then monkeys and—ultimately—humans.

Before taking that giant step, scientists must learn more about how memories are stored in the human brain—a difficult task, given restrictions on experiments. Some scientists think Berger’s project is pre-mature. The work is promising enough, though, to have attracted funding from the National Science Foundation, the National Institutes of Health and the Pentagon, which is intrigued by the notion of boosting soldiers’ performance in battle. Last year the President’s Council on Bioethics objected to “computer-brain hookups that would enable us to download the Oxford English Dictionary.” The policy clash should be memorable.

More here.