So here’s an interesting short article on the possibility of internet gaining some type of consciousness due to its network based emergence-friendly structure. The author is the famous Ben Goertzel, one of the foremost minds of the futurist/AI school. If you’ve got time you should check out his blog for other articles as well. I’ve found a number of them to be quite compelling. I’ve always been interested in artificial intelligence, though my concentration is with artificial life. In time I’ve come to view the two as the same type of system manifesting in different mediums, and I’ve come to think that intelligence is a trait that naturally comes along with the collection of characteristic called life. Intelligence is life and life is intelligence. In that sense I consider even minuscule bacteria to be intelligent, though not in a way we usually think about intelligence. The very fact that certain collection of molecular machines can work in conjunction to behave in such a way that allows it to feed, evade harm and propagate, even in evolution-aided unconscious manner means that certain system should be considered intelligent. Of course, this is merely my personal view that is not backed by evidence based professional study. This is more of a personal impression with reasonable causes, something that’s on it’s way of becoming a hypothesis but not quite there yet as the things stand. Considering that I consider our current definition of intelligence to be lacking in many ways, I will be at ire of many neurobiologists should I exclaim such opinions carelessly. And for some reason there are a lot of neurobiologists around me so I try to keep my mouth shut most of the time regarding that issue.
Ben Goertzel’s answer to whether the net can become an intelligent construct is somewhat vague, but then he probably can’t help it himself. The question itself is a bit on the vague side when you think about it, including the whole uncertainty of the definition of intelligence that I just wrote about above. He briefly mentions the pervading ethos of the neurobiologists of the recent years, that many of them believe that intelligence/consciousness is a property that will inevitably emerge from any complex system that has the right sort of internal dynamics. I do definitely agree with him on that point, since when you think about it it’s about the only scientifically feasible explanation of the emergence of intelligence/consciousness without attributing some specific part of the brain to the trait of intelligence (like how Rene Descartes attributed ganglia as the sit of the soul). I also suspect that life arises in a very similar manner, and whether that pattern of internal dynamics can be an abstraction that can be applied to different types of physical systems is a major part of my current research as a fledgling science student (the one that’s helping paying my rent). Hopefully I’ll be able to come up with something in my lifetime, since I view the possibility of such a universal theoretical platform to be a big game changer in the upcoming human century, something that might as well change the world we live in along with applications of nanotechnology and modular biology.
Will internet itself become intelligent at some point? I’m sure it will. Dr. Ben Goertzel points out that the internet is way too fragmented to display a coherent vision of an artificial intelligence and instead suggests that there might be a way to construct a sort of unifying backbone using the network infrastructure of the internet itself as a sort of raw data feed/complexity provider for that central structure. It makes sense, in a way that no one really thinks about it before someone else says it first. Most complex emergent systems, when laid out using some elements of graph theory (the graph theory, we are not talking about bar graphs and such nonsense here, for those who haven’t been keeping tabs on mathematics) displays inexplicable tendency to form central clusters around certain limited number of nodes instead of distributing indefinitely. And the change usually isn’t gradual or predictable. It happens rapidly under certain critical threshold as Stuart Kauffman put it very succinctly on his book “At home in the universe.” Internet is very obviously following in that pattern. The last graphic map of the internet I’ve seen displayed certain number of nodes (websites/services) with overwhelming number of links with a lot of nodes with limited number of links. Similar pattern is also observed in the growth of neural pathways and formation of galactic clusters, and who knows what other phenomenon in this universe escaped our notice, considering that complexity science is still a new field. Now I don’t have a very clear idea of what form that central structure would have to take to make the internet truly intelligent to observable degree… I assume it would be something on par with designing CNS for the distributed system that is the internet, possibly with a hint of recursive structure via Douglas Hofstadter, but this is all just some ideas bouncing around and I have no idea what physical/informational form such a construct would take. I’d assume it is something far past the simple matter of linking a lot of links within network nodes or providing raw processing power (that would be like saying any game of go can be won with large enough number of stones, which is just dumb. This isn’t a chess, kiddo)… I should definitely give some more thought to this, the ideas on the nature of the ‘central structure’ might as well be the catalyst I’ve been looking for.
The problem that continues to bother me whenever I think of artificial intelligence is the vague definition of intelligence we seem to share. Just how can we tell what is intelligent or not? Most definition at the moment seem to be about figuring out how human-like other organisms/systems are without regards to the actual ‘intelligence’ of that organism/system. I may not be a professional but I smell a very homocentric perception whenever I read something that pertains to the nature of intelligence. If intelligence is about being able to communicate with other beings then antisocial foreigners are not intelligent. If intelligence is about being able to react to the environment so that you can find sources of food and multiply, then bacteria are intelligent. Maybe even viruses. Both of them do not have any sort of nervous system like we do with ‘higher organisms’ so it makes the problem of intelligence a bit more complex.
Internet may become intelligent someday. This is the year that the internet will have the equal or higher number of hyperlinks as there are synapses in our brains. The real question is, how will we be able to tell if it is intelligent? Are we looking for intelligent traits or are we looking for human traits? How would we be able to tell the difference when the time comes? Maybe the first machine intelligence that blossoms on the world wide web will be trampled on by us as a mere bug in the system. After all, we do it to each other all the time.
On a little side note, the diy-bio NYC had our second meeting this Monday. We made a gel box, extracted DNA, and had a jolly good time. More on that later.