Singularity Institute

Glenn Zorpette explains his skepticism of a Singularity in the near future.

Glenn Zorpette is the Executive Editor of IEEE Spectrum.

Singularity Institute: Is the Singularity near?

Glenn Zorpette: There are different versions of the singularity. There are versions where consciousness and life-ever-lasting are central. There are ones that are more economic, in the sense that what happens when we get machines that can reproduce themselves? When they are smarter than we are? Without having to get all caught up with consciousness and living forever. OK.

If your idea of the singularity has one of it's cornerstones to be conscious machines or machines that can sustain human consciousness and let us live on, I think that's centuries away. Centuries. Once you get past one, two or three centuries, I don't think you can make an intelligent estimate about when it's going to happen.

If you are talking about machines that are unbelievably capable and self-reproducing, even that's difficult for me to say. Maybe that's 100 years away, maybe it's a century, maybe it's 75 years? The further out you get with these things, the more fuzzy your estimates are and necessarily.

But I think that that idea of self-replicating machines that are supremely intelligent -- without that, you do not really have a singularity in the sense that most people understand it. If you want me to put a number on it, I'll say maybe a century until we get that, maybe 150 years.

SI: Don't Kurzweil graphs convince you that the Singularity is near?

GZ: We have this expression in journalism -- I'm a journalist -- "Extraordinary claims require extraordinary evidence." The kind of evidence that Ray presents I think is voluminous, but it's not convincing or extraordinary.

To me, it doesn't hang together into a coherent argument that proves his point. It's what we in the journalism business call "arm waving." That might sound more disparaging than I intended to, but it's a lot of evidence that things are changing, that things are happening and that things are advancing. But it doesn't add up to a convincing argument that we will recreate consciousness in machines the near future. The near future being, I don't know what's his latest estimate? 2045? Whatever, it's within 50 years. It doesn't add up to that, for me.

That's why this is the realm of belief. To me, a lot of who people that fervently believe this -- "belief" is a bad word. Any idea that this is religious is -- they cringe at that. But to me, it's in the realm of belief.

Ray has some provocative and interesting arguments. He has marshaled some fantastically large amounts of data. My hat's off to him for that. It doesn't all hang together for me, as an argument that convinces me that I'm going to see conscious machines in my lifetime, or my kids' life time.

SI: Aren't there neuroscientists who say our brain is a machine and who think there can be a singularity?

GZ: Well I don't think all neuroscientists say that. When you say that our brain is a machine, I think depending how you define machine - let's not get into that. Sure a brain is a machine. It's a wet machine. There's nothing magical or mystical about it. But it's a fantastically complicated machine that we do not understand. There is a quote in the magazine I work for, "IEEE Spectrum." We recently did an issue on the singularity.

There is a quote from Rodney Brooks, which I thought was very thought provoking: "The brain is a machine but whether that machine is a computer is another question." Maybe you get into semantics about what is computer, but that's a good point.

We are very good at computers. We know binary logic, Boolean logic, software languages, high-level machine languages and processors and so on. They are very good at math and certain things.

The kind of pattern recognition and subtle reasoning and sensory perception -- there is really no evidence to me that that's happening in computers at any level. Furthermore to me there isn't really any strong evidence that much better software and much faster processors are going to get us there. I could be wrong.

Ray has a lot of interesting ideas about the computers that will follow silicon computers: carbon-nanotubes based logic and quantum computers. I've been a technology journalist for 24 or 25 years and I have covered a lot of these things. None of them are giving me that warm fuzzy feeling that they will be coming on strong in my lifetime, supplanting silicon-based logic.

Ray has arguments about the accelerating pace of change. There is a lot of emphasis on these slicing and scanning systems that will analyze the human brain at the -- slicing it like a salami slicers, not literally, but with electronics in a virtual way.

There is a lot of effort in that. There is a lot of good work. I think it's a great research things. But taking these fantastically large amounts of data that will result from these experiments at Janelia Farms and other places, which are now getting underway, and taking all that data, and assuming that from the data you will be able to reverse engineer the brain; I think that's stupendously misguided.

SI: Why did you focus en entire issue of the IEEE Spectrum to the Singularity?

GZ: OK, I felt that Ray has done a very good job in a number of ways. One of the things he has done very well, I think we can all agree is, he has got a lot of attention on these ideas and his views.

His books have done pretty darn well, for science and technology books. I mean our average science and technology books sell 5,000 to 10,000 copies. Well John's book sold -- what did your book sell, a 100,000?

John: My first one sold 200,000.

Glenn: 200,000, very good! Ray's books do very well. He brought his ideas to a large group of people. The sense that his version of singularity is the version that most people ... that's the singularity to most people.

There are other versions like Verner Vinge and there are other ones. IJ Good, a cryptographer and philosopher from the '50s and '60s. He was British-born but lived in America though, in Virginia or something. He had a sophisticated singularity of his own 40, 50 years ago.

John Von Neumann -- John Horgan, my friend here, and I were talking about this on the plane; he had a singularity vision quite some time ago. Nowadays people, who know about the singularity, generally associate it with Ray and his views. I felt there wasn't really a fairly decent circulation, publication or book out there, that was expressing other views and skeptical views.

And the magazine I work for, "IEEE Spectrum," reaches an influential audience of people who work in technology or are interested in technology, so I thought we were an ideal vehicle for a fairly comprehensive examination of the singularity.

We're a magazine, which means we're limited necessarily by the number of pages we can print, I think we devoted 60 or 65 editorial pages - not as many as Ray has in one of his books.

But, I feel that that was enough to let us cover the gamut of topics and subjects and present different view points - a lot of them are skeptical, and that was kind of intentional, frankly. I wasn't exactly sure what my friend John Horgan was going to say, but I knew it was probably going to be a tough and skeptical article.

We didn't want to make it just that - from the point of the view of the reader, who wants an issue was it's just "'nah, it's not going to happen, no way," article after article. We had Robin Hansen and some other people I suspected would have interesting views, not necessarily skeptical, as some of the others ones.

So I felt that this was a good thing to do, to put out, before our readers, who are people who work in technology, and frankly, many of whom are more involved with the nitty-gritty of technology. If they had any views at all, any understanding of it, it was probably fairly superficial or not all that developed.

SI: What feedback did you receive about the singularity issue of the IEEE spectrum?

GZ: That issue did really well for us - we got a lot of letters from our readers. We got, I would say, probably three or four times as many letters as we would get from a normal issue. And they were all over the map - we had people who really agreed with it, and agreed with, what seemed to be an overall skeptical tone we took, and we had some people who thought we were too skeptical and that we weren't fair. Those were probably in the minority.

What really happened with that issue, that took us by surprise, is that it did tremendously well in the blogosphere. There were just hundreds and hundreds of blogs that picked that up, and also just web publications, more serious web publications.

One of my favorites among the blogs was, the Robin Hansen's article, which was on the economics of singularity, was particularly extensively blogged. Just briefly, Hansen has this fascinating theory that singularity combined with capitalistic imperatives will lead to this bizarre far-future society.

Once you can replicate greater than human level intelligence and consciousness in a technological system, you would... from a capitalist's point of view, you'd want to shrink or minimize that consciousness to make it as tiny as possible, because a tiny being is an efficient being - very little energy is required, both for this being to exist, and also to live, not much space is needed. His idea is that once you can replicate greater than human level intelligence and consciousness in this tiny thing, you would necessarily want to minimize it so it's this tiny creature.

Then you would have these tiny creatures, again for efficiency reasons, living by the billions in these skyscrapers, because that's an efficient way to live. He goes through the article he published with us goes through this whole thing, it's really rather riveting and compelling read. A lot of the blogging comments were about Robin's article.

My personal favorite was this person, who... in his article, Robin speculated that humans would continue to exist, but they would be increasingly marginalized, which is another element of singularity hypothesis, because we're dumb compared to these super-intelligent things, even these super-intelligence tiny things. Robin mentioned that, but he didn't say what humans would do.

The few remaining humans who are left, what would they do? You've got billions and billions of these super-smart brainy robo-bugs and a handful of remaining human beings.

We had someone who blogged, who actually suggested that, since these super brainy robo-creatures would be derived from human brains, (this is one theory), you would reverse engineer the human brain to get this thing. They would still have vestiges of human urges and desires. They would have vestiges of sex drives, for example, but they wouldn't be able to act on them, because, they're half-inch long robo-bugs, or something, or maybe they would, but none of this was in Robin's article.

But this blogger speculated that that would be the role for those remaining human beings, would be making porno movies for the tiny bugs. That was about the best blog I saw.

SI: You're a skeptic, so why are you at the Singularity Summit?

GZ: I am fascinated by this, in much the same way I am fascinated by the idea that there is life elsewhere in the universe. I think the idea that consciousness will spring up in a machine, at some point, whether that point is 50 years away or 1,000 years away is one of the most compelling and profound thoughts facing technology or in the technological realm, which I mentioned before - in the same way that, whether there is life elsewhere in the universe, is perhaps the most profound question facing science.

I find the two questions similar or parallel in that they're extremely profound questions that can't help but captivate any intelligent person, but yet we have no real data, no real data points. These are things people really want to know, really turn people on, but about which we know very little, frankly.

The group that's assembled here, I think, represents the best minds on the subject. So what data there is on this will be bandied about, discussed, and debated in the next 24 or 48 hours. I'm interested in this in much the same way I'd be interested in a SETI conference.

SI: Should we debate the perils as well as the promises of the singularity?

GZ: Without debate, concepts stagnate, they don't advance, they don't move forward, they don't get sharper. I don't know of any other more efficient or faster way to advance an intellectual agenda. In that sense, John and I will play a key role here, being nasty old curmudgeons.