Special Guest of Honor Vernor Vinge Interview: ArmadilloCon 2003

Participants: Vernor Vinge, John Gibbons (interviewer)

Vernor Vinge talked about how he got started writing science fiction, about numerical parametrization of group minds, about some characters in “Deepness in the Sky”, and, of course, his brainchild — the concept of Singularity, the ways it could happen, and what current research might lead to it.

John Gibbons. How did you get into writing?

Vernor Vinge. My parents said I was slow to start to read. The first book I “really” read (my mother said) was Between World. I started writing at an early age (before 1954). I really wanted to be a great scientist and over the years I discovered I wasn’t going to be, so I started making more effort writing SF.

The first published SF writer I met was me.

I started graduate school in 1966 and went to a WorldCon — my first SF convention. I was very impressed meeting writers who I’ve been reading for a long time, e.g. John Campbell. Since then I’ve been intrigued by fandom and conventions and been to a number of conventions.

John Gibbons. How did the concept of “The Blabber” come to you?

Vernor Vinge. [Inaudible] I’ve been playing out Heinlein’s “Star Beast” mystery. I wrote that story for an anthology and wanted to name it “Blabber and Other Yammerings”. The editor hoped I was joking.

John Gibbons says something about the zones, the central concept of Vinge’s novel “Fire Upon The Deep” that says that certain levels of intelligence can only be attained in certain regions of the universe, and that in some regions it might be impossible for the lifeforms to attain even human levels of intelligence; whereas in the zone where humans live it might be impossible to attain superintelligence.

Vernor Vinge. It was a major structural thought in my head. The pack minds were in my idea box for a long time, I just didn’t have an opportunity to use them.

There is a numerical parametrization of group minds: how many individuals it takes to make a group mind? For Olaf Stapledon it’s 1013. For Poul Anderson it was a Threesome. So this was a homage to Poul Anderson. Another homage to Anderson was the idea of the zones (Anderson came up with something similar in his story “Brainwave”).

John Gibbons. How did you choose to write “Deepness in the Sky” as a prequel and focus on Pham Nuwen in particular?

Vernor Vinge. There’s an analogy to software. In “Fire Upon the Deep” there are “hooks” that imply all sorts of backstories. For example, Pham has vague, splintered memories of what happened, so I can play it out any way I want.

John Gibbons The prolog to “Deepness in the Sky” was the last thing written. Why?

Vernor Vinge. I don’t have as many people to read my stuff as David Brin. (I didn't know if he meant not as many beta readers, or that his books are not as widely read as David Brin's? -- E.) But two readers pointed out that successful novels do not begin at dinner parties. But Jim Frenkel said “by the time you get to page 70-80, whatever is mentioned of Pham is not helpful. There’s not enough mentioned”. And I just love it when I can collide problems, when I can solve two problems at the same time. So I wrote a prolog [that contained more about Pham].

John Gibbons. It builds up tension about Pham’s relationship with Qeng Ho. I loved it.

John Gibbons. Can we have a brief definition from your point of view, of what Singularity is, and when you started thinking about it?

Vernor Vinge and Elizabeth Moon sign books at ArmadilloCon 2003. More images from ArmadilloCon 2003 can be found in my photo gallery.

Vernor Vinge. I’ve always been an apocalyptic optimist. (Laughter in audience.) Computers were big on my agenda, and survival of the human species was high on my agenda. In 1965, the first story I sold, “Bookworm Run” was about intelligence amplification in a chimpanzee. According to a character in the story, implications of ability to amplify intelligence in humans is the first real weapon of mass destruction in the last 40000 years.

On a panel with Marvin Minsky I brought up a term “singularity”. Later I gave a talk to NASA in 1992 – 1993, expounding on singularity, and later developed it into an essay. Someone from University of Texas mathematics department mentioned to him [Minsky?] that John von Neumann talked about it. Then I found an obituary to John von Neumann written by Ulam [?] and there it was. John von Neumann called it “essential singularity”, and I would like to know what he meant by that, because for people like John von Neumann it had a precise meaning.

Vernor Vinge, a Special Guest of Honor at ArmadilloCon 2003
Vernor Vinge, a Special Guest of Honor at ArmadilloCon 2003

John Gibbons. Posthuman intelligence is a key point of what you envision in a singularity. Could you talk about different categories of singularity in other people’s ideas?

Vernor Vinge. Singularity means that what’s beyond it is unpredictable. Speculation about the run-up to singularity is accessible. Hard takeoff vs. soft takeoff is just a question of how fast the transition will go. Hard takeoff means [singularity happens in] a matter of days, soft takeoff means a matter of 50-100 years, with a lot of preparation and committees and planning. Kurzweil talked about it. Hard takeoff is when it takes about a 100 hours for it to happen, as in, for example, “Blood Music” by Greg Bear.

When asked what types of research are being done today that may bring us closer to singularity, what are the signs to look for that humankind is getting closer to singularity, what scientific achievements may be indicative of the coming singularity, Vernor Vinge answers:

Vernor Vinge. There are hundreds or thousands people working on it, on human-computer interface. They just don’t think about it, it’s very prosaic. On one end, inventing a mouse. On the other extreme end — interfacing with the computer via electrodes [in the brain]. On one end, a human, on the other end, a Bayesian system, and both of them are learning [from each other]. Then of course there’s pure AI prostheses.

I think the work with cuddly robot pets is something to watch. In between there are things like a few cell clumps [in human and animal brains that perform very specific functions — therefore it may be possible to replace them with artificial structures performing the same functions.]

In cases of rats and mice, you could find structures that have very clear inputs and outputs (in their brains). One group is trying to do it for humans, somewhere in hypothalamus, but I don’t know how credible they are.

World Wide Web and ubiquitous computation is another way it could happen. Scott Westerfeld’s “The Risen Empire” is a very nice treatment of group minds. A reality itself could “wake up”.

John Gibbons. Nanotech and biotech are two other possibilities.

Vernor Vinge. Drexler and I both agree that if nanotech reaches fruition, AI will come along real fast, and vice versa.

John Gibbons asks about a possibility of a pure biological progress leading to Singularity.

Vernor Vinge. There’s no doubt we could make ourselves a lot smarter and more effective with genetic engineering. Overall, though my feel about bioscience, as important as it is, it can’t keep up with machine progress. But, bioangle as team-coordination with computers — that could be a plausible path to singularity.

John Gibbons. In some ways, “Fast Times in Fairmont High” is exploring the initial stages of soft takeoff?

Vernor Vinge. It’s not clear that it’s a soft takeoff. Things get nerve-wrackingly tense as progress gets faster and faster. They are rushing down the line towards extreme capabilities.

John Gibbons. You mentioned a religious parallel some people drew, that is, the apocalypse. Does that suggest anything for the left behind? [Vinge and the audience laughs.]

Vernor Vinge. I certainly hope they’ll be treated in a congenial way. Maybe the surface of the Earth is not the best place.