August 20, 2010

Kurzweil responds to PZ Myers

Ray Kurzweil has retorted to PZ Myers's claim that he does not understand the brain:
For starters, I said that we would be able to reverse-engineer the brain sufficiently to understand its basic principles of operation within two decades, not one decade, as Myers reports.

Myers, who apparently based his second-hand comments on erroneous press reports (he wasn’t at my talk), goes on to claim that my thesis is that we will reverse-engineer the brain from the genome. This is not at all what I said in my presentation to the Singularity Summit. I explicitly said that our quest to understand the principles of operation of the brain is based on many types of studies — from detailed molecular studies of individual neurons, to scans of neural connection patterns, to studies of the function of neural clusters, and many other approaches. I did not present studying the genome as even part of the strategy for reverse-engineering the brain.

I mentioned the genome in a completely different context. I presented a number of arguments as to why the design of the brain is not as complex as some theorists have advocated. This is to respond to the notion that it would require trillions of lines of code to create a comparable system. The argument from the amount of information in the genome is one of several such arguments. It is not a proposed strategy for accomplishing reverse-engineering. It is an argument from information theory, which Myers obviously does not understand.
Be sure to read the entire response.


William said...

Kurzweil is still gravely wrong in one way that PZ highlighted - the genome does not contain the sum total of the information needed to create a brain.

Kurzweil says "It is true that the information in the genome goes through a complex route to create a brain, but the information in the genome constrains the amount of information in the brain prior to the brain’s interaction with its environment". This is wrong. It's dubious to talk about such a thing as a "brain prior to the brain's interaction with its environment", but putting that aside the development of a brain (as PZ said in his original article) is overseen by the genome but that has evolved to manipulate the environment's complexity, which cannot be ignored. How, exactly, do molecules interact with each other? Unless you can model the laws of physics, which we don't even understand fully, you cannot model a human brain simply by using a model of the genome alone as Kurzweil believes.

I thought that was clear from PZ's post.

doubtertom said...

I disagree with William. RK has never said that the genome would be used to build a brain, only that the initial design of the brain is contained in the relatively small genome. Since the genome is small, the amount of coding should not be impossible. This doesn't mean that the code for an artificial brain will be a copy of the genome! For starters, protein folding will not need to be simulated; we need to replicate the connections and signaling between neurons, not the atoms inside the neurons.

For example, Markram's Blue Brain Project has been able to model a neo-cortical cortex column without modeling atomic level physics, and they've been able to replicate the signaling that is observed in mammal NCCs.

Secondly, Kurzweil's use of the phrase "interaction with its environment" does not refer to atomic level physics, but information processing. The brain rewires itself as it receives input. Infant brains in particular go through a pruning process. To model this does not require modeling molecules.

Even if the modeling of protein folding were required, projects like Folding@Home are working to map out these processes. Once a protein folding mechanism has been mapped, it can be coded, compressed, and reused in an artificial system. The initial mapping is lengthy but can easily reused once completed. The additional processing power should not be prohibitive.

Martin said...

I'm glad he clarified some things, but he still doesn't get it:

As for the time frame, some of my critics claim that I underestimate the complexity of the problem. I have studied these issues for over four decades, so I believe I have a good appreciation for the level of challenge. What I would say is that my critics underestimate the power of the exponential growth of information technology.

*sigh* Elucidating the nature of the brain will depend on empirical research -- boots on the ground, or pipettes in the hand, as the case may be -- and has little to do with changes in information technology. For all the acceleration of acceleration of technology, mice still breed at the same rate they always have. In situ hybridization experiments still take as long as they always have. PCR reactions still take as long as they always have. There have been improvements in some experimental techniques, but they aren't accelerating exponentially like infotech. Kurzweil still views this from a comp sci perspective. He fundamentally sees the brain as a comp sci / infotech problem, when it simply isn't. It's a biological problem, and it won't be solved with more RAM or faster processors.

If Kurzweil wants to play the argument from authority game ("I have studied these issues for over four decades..."), I'll go with the professional neuroscientists and biologists who say he'crazy.