December 19, 2006

Oliver Sachs in space with vampires

bc2.jpgOliver Sachs in space with vampires. I think that sums up Peter Watts’ Blindsight: it combines hard space sf with cognitive neuroscience and a bit of horror. Practically every intriguing brain syndrome is there, from the Asperger protagonist over hallucinations of unseen presences to those conditions we neuroscientists use to harass philosophers with, like patients being fully convinced they are dead or blind people not noticing they are blind.

The basic story is a first contact story. After a strange survey of Earth aliens have been detected in the Oort cloud, and a spaceship manned with a small crew of highly enhanced humans are sent to initiate contact. The Theseus has many similarities with the Discovery in 2001, a happy case of convergent evolution of technological extrapolation. As a first contact story it is well done – the slow build-up to meeting the very alien aliens works great. Sizeable amounts of evolution and neuroscience creep in without being too expository, although this is still very much a cerebral book rather than roaring adventure and deep characterization. And the basic threat/dilemma the crew of the Theseus discover is quite profound.

The enhancements of the crew are plausibly problematic. The synesthetic biologist extends his senses and motor abilities through different sensors and actuators, but at the price of being at a disadvantage in sensing and moving his own body. Another biologist with the same enhancements has practically no body language and consistently uses ‘it’ instead of gendered pronouns. The linguist has multiple personalities, but instead of being regarded as a disorder it is now regarded as a useful and desirable state that can be supported and improved into true multitasking and multimindedness. Perhaps the most outrageous and original enhancement is the vampire captain of the ship. Vampirism is both a past evolutionary failure, an anthropophagic offshoot of our species that died out due to its vulnerability to right angles, and a successful form of posthuman existence thanks to modern biotechnology. Except of course that some of the quirks like sociopathy and scaring humans are intrinsically linked with the superior traits. There is a hilarious presentation online giving some of the background, well worth experiencing.

The Aspergers protagonist appears almost normal – except in the flashbacks to his failed relationship with a woman, where the utter strangeness and tragedy of his highly developed ability to simulate empathy become startlingly clear.

Comparing with Disch’s deservedly famous Camp Concentration I happened to read a few days earlier, I think Watts does a good job. Watts turns the situation inside out: it is not the value and high costs of these enhancements that form a Faustian bargain, but the normal state of Homo sapiens may be the Faustian thing.

The novel reads like a Greg Egan novel. The Aspergeresque theme of looking in at humanity from the chilly world of rational outsiderdom seems to recur. It is also crammed with the kind of big scale evolutionary and sociological speculation Karl Schroeder has been doing. It is no surprise the author admits Blindsight is a kind of response to Schroeder’s Permanence.

What I love about the novel is that it gets its facts so right about tiny things. Parker spirals, maternal response opioids, Sanduloviciu plasmas, taking dimenhydrinate before going to a spacewalk – the list is long. The appendices are highly readable and come with full literature citations. If Disch played with exotic alchemical, art history and theological allusions Watts plays with allusions to state-of-the-art science (and since science aims for clarity giving references is much more fitting). The Chernoff faces mentioned is probably the first vampiric information design application that anybody has ever invented.

Spoilers and discussion below

The basic claim is that consciousness is useless. Intelligence may be highly adaptable, but having a subjective experience and thinking so much about oneself (Watts seems to conflate the two kinds of experiences) may actually be negative. It takes resources, it slows us down, it misdirects energy to various endeavours that do not contribute to our evolutionary fitness.

The vampires accidentally evolved a tendency to have deadly seizures when seeing right angles, a trait that was completely neutral in the right-angle free natural environment. Since the population was small and the tendency was linked to adaptive cognitive abilities this mutation became fixed, in the end dooming the species when humans discovered the fatal weakness. In the same way Watts seems to suggest we might have fixed another useless trait – consciousness – linked to our useful mental capacities. This, just like the vampires’ flaw, was completely random and unique. Most other intelligent species evolve without developing consciousness.

I think Watts asks the important question “what use is consciousness?” but I think he gets into trouble because his conflation of self-awareness and qualia. To my knowledge there is nothing useful about qualia at all, except to give an excuse for philosophers of mind to get good salaries. On the other hand, we are surprisingly attached to them – we want to experience the world, and experience the experiencing – but I don’t know how much this is really valuing the qualia for themselves or mixing them up with general survival and cognition. When thinking of a life without qualia most of us mix it up with being dead (and the zombie terminology used in the normal discussions don’t help), and since we find being dead undesirable we think being without qualia would be undesirable. But that doesn’t tell us much about their real value to us.

My own position is that qualia are simply how information feels when it is processed, and it is not possible to avoid them!

Being aware of a self is on the other hand clearly useful in many situations. It is not just that the information about a self model is helpful in directing behaviour, it is that experience about one’s past reactions and behaviour is useful for predicting what will happen in future interactions with the environment. In social creatures other-models would easily generalise to self-models, especially when second- and third-order thinking about who knows what and how they will react to this becomes important.

Self-awareness is probably less fundamental than qualia, since we know it can be lost (in states of flow, deep meditation) or distorted. Creatures like the Scramblers in the novel might also lack self-awareness since the borders of their minds are very fuzzy. It is never clear in the novel to what extent they form a group intelligence. The border between being a being in a group and being the group might be heavily blurred as communication abilities increase.

In the end I think Watts’ case is intriguing, but I’m not convinced. However, I do think we have a very parochial view of how intelligence should work. All vertebrates have nearly the same large-scale brain topology, and the differences between mammals are even smaller (all the pieces and wiring are there, it just looks like some have more neurons in some systems and localized adaptations). This may give us a very anthropocentric view of what it means to be a being in terms of having a motivational system, emotions, decision making along certain lines, self-awareness and social interactions. I do not think intelligence has to be combined with decision-making in our sense (one could imagine something akin to a super-Google), selfhood could be heavily blurred or shared, subjective risk/reward could run on entirely different schemes and decisions could be based on something else than a modified motor planning system.

Posted by Anders3 at December 19, 2006 12:42 PM