[mfn]

Optogenetics is incredibly exciting. Trying to understand the brain without such a technique seems analogous to trying to debug a complex program without being able to vary its inputs and instrument it to dump state information.

Boyden mentions that optimistically, this technique could be perfected within just 5 years and potentially yield enough information to allow us to construct computational models of (small) neural circuits. Would be interesting to see how quickly this progresses now that it seems that we have the tools.