Science fiction, by (almost) default, deals with the concept of “life as we do not know it”. This is of course, the idea that life can include forms that may be fundamentally different from our own. It is well-known that life is a notoriously difficult concept to define. That said, a lot of progress is made each day on trying to determine all the possible paths that this phenomenon can take.
As science fiction fans, it is entirely possible that the first example of life as we do not know it that we ever heard about was robots. The very word robot means “servant” and it seem that it was used for the first time in 1921 by the Czech writer, Karel Kapek in his play R.U.R, short for “Rossum’s Universal Robots”.
By far, the most interesting type of robots are the ones endowed with artificial intelligence. We are all familiar with the most famous SciFi robots, from the robot in the iconic 1956 movie “Forbidden Planet” (nicknamed “Robbie”, although the phrase “Robbie the Robot” was used before in science fiction stories) to Star Trek’s Data to the humanoid Cylons from the reimagined Battlestar Galactica.
The great Isaac Asimov was probably the first to come up with a plausible mechanism to build artificial human-like intelligence. He imagined positronic brains, as opposed to electronic brains. This concept was extensively borrowed in many later science fiction series and stories. For example, did you see the movies “Bicentennial Man” starring Robin Williams and “I Robot” starring Will Smith? These two stories are Asimov’s originals…
The topic of robots, especially intelligent ones is fascinating and merits a post of its own. However, I would like instead to talk a little bit about what would it take to create a true example of artificial intelligence.
Intelligence in general is a broad term that can be defined in multiple ways, a little bit like consciousness. However, the general accord is that whatever human consciousness or intelligence may be, they are a direct result of the architecture of the human brain.
A recent trend in neuroscience is to talk about mapping all the connections in the human brain, an effort called the human connectome project. Basically, the idea is to map all the connections of the 100 billion neurons of the human brain (no, there are no 86 billion neurons in a human brain; don’t even get me started… What? You still want to know what I think about it? Ok then, look here, here and here).
The connectome idea is very interesting and will be a wonderful tool to try to understand ourselves as well as for understanding a wide variety of neurological conditions (psychiatric conditions are by definition neurological as well, just so you know). This will, without a doubt more than a blessing to those of us who care of a mentally ill loved one or those of us who are going through it ourselves.
As wonderful as this dream of a connectome is, even though imaging technology is advancing, we still do not have the optimal tools or the computer power to completely map these connections. Moreover, not every important aspect of brain function is directly related to its physical circuitry. To say that a physical map of a brain says all that there is to say about it is a little bit like saying that a map of a shopping mall tells you how all about how it works. The real and complete picture of how a shopping mall works must include all the multiple activities within each store that are not accounted for in the “map”, as well as many related aspects like whether the manager of a particular store contacts the main offices to solve an issue, etc.
To make things even worse, there are many aspects of neuronal transmission that are controlled by processes that work beyond the direct neuron-neuron connections, of which I hope to talk about in a future post. Also, there is the matter of the sheer physical complexity of a typical brain. For example, 100 billion neurons, each one with an average of 30,000 contacts (synapses) with other neurons; you do the math (hint: it is a BIG number). This fact, added to all the extraneuronal events that happen to make it work, makes even a simulation of an actual brain a tall order at our current level of technology (don’t get me wrong, it WILL happen, just not as fast as the hype states…). In fact, about simulations and models, if you want to see some of my thoughts about it go here and here.
These are not the rumblings of a “bah, humbug!” guy. There are other people taking a critical look at the connectome concept.
Let me finish this post with a couple of thoughts about the brain (which were generated by my very own brain… (:-D)…).
The human brain is oftentimes called the most complex structure in the known universe and rightfully so.
This being said, there are a couple of little details that one must not lose sight of. (1) The brain is a biological construct, crafted by evolution and (2) the average speed of nerve transmission is in the order of 55 miles/second. Fast enough, huh?
However, an electronic computer, which will inevitably be the “ancestor” of a robot brain has two very -and I mean VERY- important advantages; it is rationally designed and it works with a speed close to 186,000 miles/second.
Yes, light speed; the very speed limit of the universe itself.
I would not discount electronic brains just yet.
Picture credits: Wikimedia Commons.
Wanna take a look at my blog, “Baldscientist”? Click here.
If you want to know more