Future Current

The AGI-08 Post-Conference Workshop session presented by Professor Hugo de Garis of Wuhan University investigates the possibility of a bitter controversy arising out of humanity’s capability of building massively intelligent machines. It foresees humanity splitting into three major camps: the Cosmists (in favor of building artilects), the Terrans (opposed to building artilects), and the Cyborgs (who want to become artilects themselves by adding components to their own human brains).

The following transcript of Hugo de Garis’s AGI-08 post-conference workshop presentation entitled “The Artilect War” has not been approved by the speaker. Video is also available.

The Artilect War

Last year in China I gave about five big talks, so I am more or less getting into a certain pattern. It is much more fun if there are sparks. This will not be just an intellectual talk; I will try to hit your gut, to make it more emotional. It’s more fun that way.

This is my first book, published three years ago now. It is selling very poorly in America, very well in China. The Chinese press that did this is the equivalent of the MIT Press, so I was very happy about that.

Here is the argument in a nutshell (a lot of it is uncontroversial in this community, of course). 21st century technologies will enable us as a species to make godlike, massively intelligent machines. The mathematical physicist in me is very interested in something called “phys-comp,” the physics of computation. You all know, Moore’s law will take us at around 2020 to about one bit per atom. What are the switching speeds of an atom? Femtoseconds? So if you had a nanotech grain of sand, the total bit flips computing potential of that grain of sand, compared to the human brain, the estimates are greater by a factor of billions.

What I’m trying to say is, AGI will take us up to more or less human level, but there is far more, way beyond that. The physics says that is possible–let me try and give you some of the numbers. Personally, I see the greatest political problem of global politics later this century being dominated by this issue of “species dominance.” It is almost like a life calling to me.

One of the reasons I am in China is I see it as almost inevitable, so long as China does not crash the way Japan did in the ’90s, if China continues the way that it is going, sooner or later it will be the dominant culture on the planet. How do you compete with a giant? 1.3 billion people–Shenzhen is growing economically at 30% per year… How do you compete against that kind of thing?

When I give the talks to the Chinese, I usually have a translator. I’ll say something for a minute, then the translator will translate in Chinese for a minute, and we flip back and forth. It works very well. I say to them, making a nationalist appeal to their Chinese ego, because they have been badly bruised. We Anglo-Saxon, we’re shits. What we did to China was just terrible. They don’t like us, justifiably. I try to appeal to them that I believe this issue of species dominance is the dominant issue of our century. Therefore, it seems appropriate that the dominant culture should be the leader in this debate.

As our machines get smarter and smarter, the debate will be taken outside the technological community. But this issue is not yet really out there with the public. The media has not quite yet grabbed onto it, but that is changing fast.

As the machines, the brain-based industries, get smarter, they will make the household robots increasingly smarter year by year. As the IQ gap, the difference between human level and machine level, as that gets smaller and smaller, more and more people, ordinary people–the masses, if you like–will be asking questions like, “Could these machines become as smart as us? Could they become smarter? Can they be stopped, should they be stopped?” What about the economic momentum, could you stop that?

The US/Chinese rivalry that will be growing–China is pouring money into its defenses, and America is getting worried. The Chinese government is a bunch of thugs, speaking frankly. On the Chinese money is a picture of the man who has killed more people than Stalin or Hitler. According to recent biographies, Mao killed about 70 million people–Imagine putting Hitler on the euro. As China becomes more and more powerful–and America strongly disapproves of China, appropriately–with the military momentum, how are you going to stop the rise of intelligent soldier robots in the political context of this kind of rivalry?

As this gap closes, the species dominance debate will heat up. Should we put a cap on the level of artificial intelligence, or should we just let it go on and on? I see society splitting basically into two, arguably three, major political groups. Let me give you some labels, just for convenience. This is a philosophical, political group of human beings saying that we should build these godlike creatures, because in the big picture our pathetic little human lives snuffed out in 80 years measured against a universe that is billions of years old–that kind of argument. I call them “cosmists,” based on the word “cosmos.”

The second group, will be opposed because, they argue, it is too risky. If the machines do become hugely more intelligent than us, a key word in the whole debate is “risk.” We just don’t really know. If you are a politician fifty years or more into the future, you will hope that they will not be a problem, but you will prepare for the worst. What is the worst thing that could happen? They may feel so superior to us that they just treat us like mosquitoes, or they may wipe us out through indifference. They may do something that has the consequence that we get wiped out, and they don’t give a damn. For all I know, right now I am killing how many bacteria? Probably zillions, but I don’t care.

The gap closes, the debate heats up, the general public gets involved. Then political parties start being formed, assassinations start, sabotage starts, and the general temperature rises. We are not talking now about the survival of a country. If you were Russian in the 1940′s, you were fighting for your life. The Nazis killed 20, 30 million Russians? This time, we are not talking about the survival of a country–we are talking about the survival of a species. Us. I am arguing that the passion level will be so much higher than anything we had in the historical past.

This is what I find particularly gloomy. On this axis, you have the number of people killed in major wars. In the context of the American Civil War, about half a million were killed. I was shocked to learn when I moved to China about the first week, in that same decade was the so-called Taiping Rebellion in China. There, 20 million were killed.

You ask yourself the obvious question, why is the scale of the killing going up? The obvious answer is technology. In the American Civil War, you had rifles. The bullets were about that big, and if they hit you in the arm, you lost your arm. In the First World War, you had machine guns. In the Second World War, nuclear bombs. If there is a major war over this issue of species dominance because of the passion level between these two major deeply opposed philosophical groups, extrapolate late 21st century weaponry up the graph and you get what I call “gigadeath.” I’ll stop there.

This entry was posted on Monday, June 23rd, 2008 at 11:46 pm and is filed under AGIRI, Artificial Intelligence. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.