Wired had an interesting article about the successful reverse-engineering of the optic system that allows flies to detect motion patterns. Apparently they don’t fully understand how it works, but have codified it as a system of non-linear equations mimicing the “neural network” of synapses in the fly’s optic center. I have not been able to find the paper on the web, though have found the abstract.
I have often thought that the first successes in creating an “intelligent” AI will be accomplished by mapping the brain and sensory input structure of an existing organism; this as opposed to a new model of knowledge representation and learning. The latter is a much harder problem than initially anticipated in the 50’s and 60’s, and I’m afraid we are no closer to the goal. I believe our ability to read biological neural activity at progressively finer granularity will outpace innovation on the AI modeling front.
Certainly there is a very long way to go before (or if) we are able to observe individual neuron activity in a vast structure (such as a mammalian brain) OR alternatively scan the configuration of neurons. Speculating as a layman, I am going to guess that the amount of noise from the vast number of neighboring or enclosing neurons in the 3-dimensional structure would prohibit the extraction (isolation) of an individual neuron signal. If this problem is insurmountable, may only leave intrusive / destructive approaches to scanning neural structures.
Renaissance in Biology
Biological systems are enormously complex. In physics and chemistry we try to understand the fundamentals that pertain to matter/energy, information, and time, mostly at the microscopic or gross macro-level. With biological systems we have a multitude of “state machines” at different levels each of which can be understood at some micro-level, but result in emergent behavior that we, as of yet, hardly understand.
We’re very early in the game, but it is exciting to see biological systems being explored within the fields of mathematics, engineering, statistics, physics, etc. With computing power came a number of innovations in biology and the approach to biological systems. Protein folding, genome analysis, simulation / modeling of cellular mechanisms, are just a few of many areas that were previously intractable (or at least too complex to fathom without the aid of massive computations).
With new tools, biology has become more interesting to engineers and computer scientists, amongst others. Here is a very interesting example of how biological sciences are changing: