Saturday, the NY Times ran a story by science reporter John Markoff titled “Scientists Worry Machines May Outsmart Man.” The article gave examples of a robot that doesn’t need humans to recharge itself, an unstoppable computer virus, autonomous killer drone aircraft, and machines that simulate human conversation and emotions. The reporter wrote:
Impressed and alarmed by advances in artificial intelligence, a group of computer scientists is debating whether there should be limits on research that might lead to loss of human control over computer-based systems.
According to the article, the computer scientists at the conference said there is still a long way to go to create a computer like Hal in “2001: A Space Odyssey.” And they “discounted the possibility of highly centralized superintelligences and the idea that intelligence might spring spontaneously from the Internet.”
Discounted the possibility? I would like to know why. The article doesn’t explain. Do they say this mainly to reassure us? So that those of us not there (at the Asilomar Conference Grounds on Monterey Bay in California, where this conference took place) don’t panic, rise up in fear, and march as Luddites united demanding: No more work on computer intelligence!
The reporter gives no explanation for the “discounting.” On the contrary, to my mind, there’s good reason to think this is precisely what will happen: Supercomputers will become intelligent and the internet will develop intelligence, and both of these things probably are happening now.
After all, what faculties are there in our brains giving us intelligence that do not already reside “out there” in computer systems hooked together by internet cables—just like the functional centers of our brains are hooked together by trunks of axons? There are databases of words, images, and facts (i.e., memory systems); there are systems that receive data from the environment (i.e., sensory systems); there are processors running machines that do tasks and do work (i.e., motor systems); there are computers that delegate and integrate operations (executive and homeostatic systems); there are artificial intelligence systems (such as those learning to speak and interact with humans); and last but not least, there is the internet that connects the computer systems.
Supercomputers are being constructed now, so why should they not be able to execute these functions and become intelligent? And since the internet is connecting the supercomputers, why shouldn’t the internet become intelligent?
This idea is not new. The reporter cites a spectacularly prescient essay, “The Coming Technological Singularity: How to Survive in the Post-Human Era,” written in 1993 by Vernor Vinge of the Department of Mathematical Sciences, San Diego State University. Vinge proposes, “Large computer networks (and their associated users) may ‘wake up’ as a superhumanly intelligent entity.”
The article, therefore, does not break news. Rather, what is eye-catching about the Times piece is, of course, the possibility that these developments may present a danger to humanity and may therefore be cause for “worry.”
According to the article, possible dangers, in addition to those mentioned above, might include criminals controlling superintelligent computers and computers putting people out of work or displacing humans in other ways, including as companions. But I suspect that if the title word “worry” accurately reflects the tenor of the conference, the greatest fear is, in Vernor’s words, “the human era will be ended” and that a post-human world of computer beings will begin. He asks, “just how bad could the Post-Human era be? Well … pretty bad. The physical extinction of the human race is one possibility.”
This is a characteristically human worry, I think: That we may need to defend ourselves from extinction from approaching changes and new situations, and that we ought to prepare ourselves to resist them. It’s like the way we view the possibility of Iran developing nuclear weapons or North Korea developing missiles to hurl their weapons at us.
Although Iran and North Korea do present real dangers, I do not think the same considerations apply to the possibility of machines and the internet developing superintelligence. The difference is that Iran and North Korea are different nations from us, with their own interests that may contradict ours, but this difference does not exist between humans and computers or humans and the internet.
Humans are part of the world of computers and internet, which is developing around us so rapidly. We are not separate from it. In the future, the humans will continue to use the computers, and (although we usually don’t speak this way) the computers will continue to use the humans. As the internet grows, humans will use it more and more. If we discover one day evidence of intelligent internet functioning, that internet-being won’t be separate from humans, we humans will be part of it, and we will experience and benefit from the wonderful things that this humans-computers-internet-being is able to see and do.
Vinge’s remarkable essay even considers this possibility. He writes:
Every time our ability to access information and to communicate it to others is improved, in some sense we have achieved an increase over natural intelligence. Even now, the team of a PhD human and good computer workstation (even an off-net workstation!) could probably max any written intelligence test in existence.
Vinge lists some things human-computer combinations might work together on: technological problems, art, games and pastimes, and he writes of the “worldwide Internet as a combination human/machine tool.” Still, I’m not sure Vinge got what he himself was saying.
I’m going to make the claim without justifying it that what really bothers us is that we humans might be excluded from the superconsciousness that emerges in the post-human era. We fear that the huge computer-internet being will have a consciousness all its own, different from ours, vastly larger and more complex, and humans will not be part of it.
We should not fear this. We do not need to worry. We won’t be separate from the consciousness of computers and the internet. We are now communicating through the internet and with the internet (i.e., with the computers on the internet) continually. For example, as I write this, I am communicating with my desktop computer, and when I post, I will be communicating with the internet.
And communication IS consciousness.
My favorite example illustrating how consciousness and communication are the same comes from the split-brain experiments. (I posted about this previously.) Patients back in the 1980s had their brain hemisphere connections (corpus callosum) severed as a treatment for intractable epilepsy. They emerged from the operation with two separate consciousnesses, one in each hemisphere.
The point of the illustration is this: Before the operation the patients had two hemispheres, but they had only one consciousness. That was because their hemispheres were communicating continually via the callosum.
Just as communication via the callosum expands and unifies brain consciousness, internet communication will expand and unify human-computer-internet consciousness. In the future, we humans will continue to communicate through and with the internet and the computers connected to the internet, just as we do now. Because of this, we should not doubt that human consciousness will be expanded and will be unified with that vast network of computers and connections. The vast and complex consciousness of humans, computers, and the internet will be ours. We will be part of it.
However, there is something we should worry about regarding this future: We should make sure that no humans anywhere get left out. Everyone on earth should get an internet connection, and a computer, and the training to use it. There should be no elites in the future—the few who have access to this unified communication and consciousness—and the masses or the minorities who do not. We must make sure everyone gets included!