Our expert voices conversation on "When computers merge with our minds."

Our ability to remember and reason is fundamentally limited by our wetware so the idea of punching through that barrier by expanding our brain with hardware is very appealing. It's possible to read out brain signals using electrodes or imaging and write brain signals through electrical or optical stimulation. A skeptic might correctly point out that there's still much to discover about how the brain works, and then conclude it a fool's errand to hook it up to a machine. However, progress cannot be made without attempt. Research in brain machine interfaces will help us glean insights on brain function and improve the understanding necessary to make brain hardware succeed.

The payoff is high: Imagine, for example, being able to interact with the sum of human knowledge with the same flexibility and ease of revisiting our memories. On the other hand, the challenges are legion, not the least of which is the brain interface itself: current models lack the necessary bandwidth, selectivity, or longevity. This, at least, is a problem that can almost certainly be solved.

The other voices in the conversation:

Sridevi Sarma, biomedical engineer, Johns Hopkins University: How and how much to change the brainSteven Chase, biomedical engineer, Carnegie Mellon University: What to fearAndrew Hires, neurobiologist, University of Southern California: An optical link to the brain