False Prophecy

(…being another reprint of a months-old Nowa Fantaskya column, because I’m still in Vancouver and haven’t yet had time to do my epic comparison of Fury Road and Kingsman)

I’ve been called a prophet now and again. Articles about neuron cultures running robots or power grids generally provoke a comment or two about the “smart gels” from my rifters trilogy. βehemoth is likely to get a shout-out with each new report of mysterious sulfur-munching microbes, deep in the bowels of hydrothermal rift vents. Recently The Atlantic posted a piece about Louis Michaud’s work on energy-generating tornadoes; readers of Echopraxia pricked up their ears.

I didn’t foresee any of it, of course. I just read about it back before it made headlines, when it was still obscured by the jargon of tech reports and patent applications. In fact, my successful “predictions”— submarine ecotourism, Internet weather systems, smart gels— are happening way sooner than I ever expected.

Predict the future? I can barely predict the present.

I’ve only made one “prediction” (although “insight” would probably be a better term) whose rudiments I haven’t stolen. I’m really proud of it, though. Screw those recycled factoids about head cheeses and vortex engines: I’m the guy who wondered if Consciousness— that exalted mystery everyone holds so dear and no one understands— might not just be some kind of neurological side-effect. I’m the guy who wondered if we’d be better off without it.

I may not be the first to pose that question— I’m probably not— but if I reinvented that wheel at least I did it on my own, without reading over the shoulders of giants. And the evidence in support of that view— the review papers, the controlled experiments— as far as I know, those started piling up after Blindsight was written. So maybe I did get there first. Maybe, driven solely by narrative desperation and the desire for a cool punchline, I threw a dart over my shoulder and just happened to hit a bullseye that only later would get a name in the peer-reviewed literature:

UTA, they call it now. “Unconscious Thought Advantage”. The phenomenon whereby you arrive at the best answer to a problem by not thinking about it. I like to think I got there on my own.

So you can imagine how it feels to stand before you now, wondering if it was bullshit after all.

The paper is “On making the right choice: A meta-analysis and large-scale replication attempt of the unconscious thought advantage” by Nieuwenstein et al. The journal is Judgment and Decision-Making, which I’d never heard of but this particular paper got taken seriously by Nature so I’m guessing it’s not a fanzine. And the finding? The finding is—

Actually, a bit of background first.

Say someone gives Dick and Jane a problem to solve— something with a lot of variables, like a choice between two different kinds of car. They’re both given the same data to work with, but while Dick gets to concentrate on the problem before making his decision, Jane has to spend that time doing unrelated word puzzles. The weird thing is, Jane makes a better decision than Dick, despite the fact that she didn’t consciously think about the problem. Conscious thought actually seems to impair complex decision-making.

I first encountered such findings almost a decade ago, while correcting the galleys for Blindsight; you can imagine the joyful dance my hooves tapped out upon the floor. In the years since, dozens of studies have sought to confirm the existence of the Unconscious Thought Advantage. Most have done so. Some haven’t.

Now along come Nieuwenstein et al. They wonder if those positive results might just be artefacts of sloppy methodology and small sample size. They point out a host of uncontrolled variables that might have contaminated previous studies— “mindset, gender, motivation, expertise about the choice at hand, attention and memory” for starters— and while I’d agree that such elements add noise to the data, it seems to me they’d be more likely to obscure a real pattern than create a false one. And though it’s certainly true that small samples are more likely to produce spurious results, that’s what statistics are for: A significant P-value has already taken sample size into account.

Still. Sideline those quibbles and look at what Nieuwenstein et al actually did. They used a much larger sample, applied stricter protocols. They avoided the things they regarded as methodological flaws from previous studies, reran the tests— and found no evidence of a UTA. No difference in effectiveness between conscious and nonconscious problem-solving.

Shit.

It’s not a fatal blow. In fact, Nieuwenstein’s study actually found the same raw pattern as previous research: the responses of distracted problem-solvers were 5% more accurate than those of the conscious-analysis group. The difference just wasn’t statistically significant this time around. So even if we accept these results as definitive, the most they tell us is that nonconscious decision-making is as effective as the conscious kind. Consciousness confers no advantage. So the question remains: what is it good for?

The authors tried to talk their way around this in their discussion, arguing that “people form their judgments subconsciously and quickly, then use conscious processes to rationalize them”. They speculated that perhaps these experiments don’t really compare two modes of cognition at all, that both groups came to their conclusions as soon as they got the data. Whatever happened afterward— focused contemplation, or distracting word-puzzle— was irrelevant. It’s a self-defeating rationale, though. It’s not a defense of conscious analysis, only an acknowledgment that consciousness may be irrelevant in either case.

The jury remains out. A day after “On Making the Right Choice…” came out, the authors of the original, pro-UTA papers were already attacking its methodology. Even Nieuwenstein et al admit that they haven’t shown that the UTA model is false— only that it hasn’t yet been proven. And these new findings, even if they stand, leave unanswered the question of what consciousness is good for. The dust has yet to settle.

I have to admit, though, that Nonconscious Isn’t Any Worse doesn’t have quite the same ring as Nonconscious Is Better. Which, personally, kind of sucks.

Why couldn’t they have gone after my smart gels instead?

 

 

 

This entry was written by Peter Watts , posted on Wednesday June 03 2015at 11:06 am , filed under blindsight, sentience/cognition . Bookmark the permalink . Post a comment below or leave a trackback: Trackback URL.

36 Responses to “False Prophecy”

  1. Are you doing book signings etc in Vancouver?

  2. Thought I was alone thinking my sense of self was a gestalt-ish side effect.
    Then I read Blindsight and Echopraxia! Thank you for making me think while having fun!

  3. The answer appears to be: Ex post facto rationalizations that make you feel better about the decision you just made.

  4. Or maybe, we’re looking for the conscious mind in the wrong decisions. Making judgements about a choice that’s directly in front of you is something that any mobile animal has to do at some level, even ones with only 100 neurons to rub together — abstracting it out to “which car is the best one to pick” might not change the mechanism by which we make that choice.

    The thing that our brains are supposed to be really good at is forward planning, the frontal-cortex stuff: can we, do we, do that unconsciously?

  5. As always a thought provoking post, for example this is where I went.

    I think that if consciousness is an illusion then psychotherapy would become at best a placebo, and at worst a deception played by the therapist upon the patient. As I don’t believe the latter, because while the evidence suggests that 48% of the outcomes are decided by the therapeutic relationship, which is an enormous confounding variable when testing the efficacy of modalities, there are still more effective versus less effective interventions.

  6. My objection is that it seems you get to have your cake and eat it, consciousness is an illusion? Ok, so the “I” isn’t in charge at all, it’s just a spectator fooling itself saying “I did this!” while limbs and lips go on doing their thing. But then you say it gets in the way, it makes things more inefficient, it interferes! How can something that has no control interfere? Surely it has SOME input.

    Whenever I self monitor doing physical tasks with this idea in mind, I find there’s a sweet spot in attention, sure you need to disengage and let your automatic processes carry the load when you’re say, trying to balance on a narrow kerb but you can’t disengage completely either.

    It’s an intriguing subject but I’m not sold on the whole “we’re better off without our frontal lobes” motion.

    It’s a good topic for certain kinds of parties though…

  7. But I suppose it doesn’t exclude the potential existence of other forms of life that pretty much looks and behave like the more complex animals here on earth; limbs, sense organs and a sophisticated neural system, but with a complete absence of awareness, sentience and consciousness?
    Just too bad we don’t have organisms like that here on earth, with a similar biochemistry like us. No need to feel guilty from eating meat from such creatures, since they would not be able to feel neither pain, fear or suffering.

  8. I certainly can’t say with certainty what consciousness is good for, but I can say something that I’m pretty sure UTA is great for. My personal experiences may all be placebo effect or confirmation bias, but my experiences are consistent with the reality and utility of UTA – not just for decision-making, but for information synthesis and organization.

    I am a lawyer, consultant, and educational content producer specializing in e-discovery (tech + law). I regularly embark on extensive research efforts into new legal and technical topics or developments for the purposes of creating new educational or reference materials or writing guidance memos for clients.

    The standard practice for this I have developed (and used for years now) is: to spend the day gathering all potential source articles, cases, and other materials on a topic; to read through all of those materials, filling my mind to the brim with them, in the evening; to spend no time consciously analyzing them; and then, to go to bed. Without fail, the next morning I awake with all of the information organized in my mind into a clear relational structure, useful patterns revealed, potential outline for explanation available – like picking up a completed research memo from a subordinate on my way into the office. I can’t be the only one who’s noticed that this is possible.

    Even if it turns out that we’re all just ghosts in biological machines, born accidentally as a side-effect of structural complexity, it’s still damn useful for getting through the week to know how to work with your natural cognitive processes rather than against them.

  9. I think (not originally) that consciousness brings attention to corners of some complicated subject, per ‘Global Workspace’ theory. Once you’ve put that sub-tangle on center stage, and used that time to work out its relationships to the rest, you can take it back out of consciousness. Massage all the complex or unknown parts this way and you’re done with consciousness, and persisting in it might distract from the larger picture you need to get back to. This description would predict that the more parts there are to a problem, perhaps not-too-familiar parts, ie /the more aspects that need chasing and integrating/, the more conscious thinking will help.

  10. I’d go with ‘Global Workspace Theory’, that consciousness is to bring less-familiar (pre-worked-out) or, especially complex aspects to center stage to help you untangle or work out the consequences of, or the relationships to other parts of the problem. Once done you don’t need consciousness, and it might even give you tunnel vision or be a distraction. /Not an original thought.

    This would predict that problems involving un-familiar things, or, containing especially complex sub-parts, would do better under consciousness.

    (I sent an earlier version but didn’t have cookies enabled – if you see both, I think this later comment is better.)

  11. […] Watts wonders what consciousness is actually good […]

  12. So, if Charlie R or what’s his name is still around, here’s a link for him:

    http://rstb.royalsocietypublishing.org/content/368/1631/20130080.full

    plainly states that research has found that women don’t do direct confrontation and aggression like men, and instead prefer passive-aggressive shit like spreading rumors and messing with others psychologically, because it’s physically safer, and women are more conscious of that for reasons elaborated upon in the book of Oogenesis*, and thus less likely to engage in heroic behavior.

    Which usually gets people killed. and women don’t want to get killed as their biological programming evolved to put survival higher than winning big, in contrast to males.

    Also, speaking of women in male roles, apparently one woman has managed to graduate from Ranger school (50% of soldiers wash out). Every other XX candidate. could not hack it, even after getting recycled and given a second chance.

  13. might not just be some kind of neurological side-effect.

    Consciousness impairing mental performance .. why can’t that be the case even though it is not a side effect, but an important element of the mind?

    Could be the same thing as say, boss in any organisation. Important position, crucial, can make or break it, but if he’s breathing down the neck of some division, could be those people are more apt to make mistakes.

    So the question remains: what is it good for?

    Okay-.. has anyone ever gotten people to unconsciously solve complex problems? Play chess? Write computer code?

  14. Duffelblog is kind of like the onion for the US armed forces

    “I was really impressed with the way she performed in the SERE portion,” said Staff Sgt. Juan Gonzalez, an artilleryman from 3-319th Field Artillery. “She lasted longer than any of us, because when the instructors got close with their tracker dogs, Ranger O’Keefe covered herself in period blood to hide her scent. It threw the hounds off long enough for her to slip away and last out in the woods another 12 hours before they got her.”

  15. Nestor:
    Duffelblog is kind of like the onion for the US armed forces

    Yeah, I know. It is hillarious though!

  16. Is there any research looking at different kinds of problems and seeing if UTA is more effective in some of them?

  17. Nestor,

    And, like The Onion, sometimes stomps its wang when attempting to perform a Swingin’ Richard routine. Frex, the female Ranger “article.” Sigh.

  18. Re: janinmi

    I dunno, I’m a woman who technically supports introduction of women in combat roles in cases where appropriate level of fitness has been demonstrated, and LOLed at that duffb’ article (I also laughed at philotrans article, but in a waaay less friendly manner, because frankly, evopsych is getting pretty old pretty fast and is becoming very much like an annoying old relative in early stages of senility)

    Re: Y

    Everyone gets you have surprisingly strong opinions about women in military service, but don’t we already have a cozy thread where this stuff is being discussed, heh ? :)

    Re: Nestor

    Well, it could be that the only thing consciousness does is “waste cycles” and is thus both useless and a little bit harmful (an ATP thief of sorts).

    It could be that the only reason we’re conscious is because non-conscious intelligent agents with same kind of problem-solving powers are even harder to evolve naturally, or incompatible with architectural layouts of the kinds of brains that happened to evolve in higher mammals on Earth.

    It’s exceedingly speculative, but hey, we’re on a scifi blog.

  19. 03: Everyone gets you have surprisingly strong opinions about women in military service, but don’t we already have a cozy thread where this stuff is being discussed, heh ?

    I thought it’s been already locked..

  20. I personally think you’re overthinking it. Go with your gut, man! Blindsight is a, if not THE, favorite book I keep around at all times. I just want more vampires.

  21. Gord Wait:
    Are you doing book signings etc in Vancouver?

    Nah, I was meeting with someone. And, as it turned out, getting a tie.

    Chris: The thing that our brains are supposed to be really good at is forward planning, the frontal-cortex stuff: can we, do we, do that unconsciously?

    Yeah, evidently. The problem is, once that suconscious decision has been made and shunted upstairs to the pointy-haired boss, said pointy-haired boss takes credit for the decision and retcons all sorts of rationales to “explain” it, even if they have nothing to do with the real decision (people whose arms are involuntarily moved via remote control, for example, will, when asked, come up with “reasons” that they “decided” to move their arms when they did.) So no matter where the decision was made, it always feels as though it was conscious.

    Nestor: How can something that has no control interfere? Surely it has SOME input.

    Some have argued that that’s exactly what consciousness is for: to act as an interfering “veto power”, a kind of Senate that can cancel decisions made nonconsciously. Why that particular senate has to be self-aware, though, is not something that’s been answered as far as I know.

  22. Biff: Just too bad we don’t have organisms like that here on earth, with a similar biochemistry like us. No need to feel guilty from eating meat from such creatures, since they would not be able to feel neither pain, fear or suffering.

    Yeah, well, Descartes started by assuming that as an axiom and presumably felt no guilt at all as he vivisected screaming creatures. I think a lot of people still do.

    As for encountering such intelligent-but-nonconscious creatures elsewhere in the cosmos, though, that’s something worth considering. Someone oughtta write a book about it.

    Y.: …plainly states that research has found that women don’t do direct confrontation and aggression like men,..

    Dude, why is this subject popping up in this comment stream?

    Y.: Okay-.. has anyone ever gotten people to unconsciously solve complex problems? Play chess? Write computer code?

    (Presumably nonconscious) computers have been playing chess for years. And just this last week, a paper came detailing how an AI derived it’s own genetic algorithm for brute-forcing a solution to the planaria-regeneration problem. Whether or not we use consciousness for those things, it’s pretty obvious that consciousness isn’t a prerequisite for them in principle.

    bookworm1398: Is there any research looking at different kinds of problems and seeing if UTA is more effective in some of them?

    That first study I cited ran their subjects through a number of problems of differing complexity, varying from should-I-wear-gloves-at-the-opera to which-car-should-I-buy. So they tested various levels of complexity, at least.

  23. Peter Watts: Whether or not we use consciousness for those things, it’s pretty obvious that consciousness isn’t a prerequisite for them in principle

    I know, but I consider the idea that conscious thought is a sort of ‘training camp/wheels for humans that allows them to solve novel problems and which can lead to formation of unconscious ‘modules’ or routines that can do the same thing more effectively.

    So I’m curious if you can teach people to solve complex problems subconsciously, without ever having them pay conscious attention to the complex problem.

  24. I’m with Chris on this one. Sentience is, according to most, a really exclusive club here on Earth, so if we want to know whether it carries a disadvantage, shouldn’t be testing the UTA against things we’re (supposed to be) better at than those outside the club (the relative utility value of those things is, of course, also up for debate, but that stuff is constantly shifting anyway in the modern world).

    Peter Watts: Dude, why is this subject popping up in this comment stream?

    Presumably because the other thread has petered out, and Y. finally found a study that backs up his argument, so he wants to make extra-sure that the person he was arguing with reads his amazing point so everyone knows he’s right and his opponent is wrong.

  25. Maybe consciousness is just what gets us through the day until we can have an opportunity to do our real brainwork, dreaming. I’m another of those people whose solution to complex problems is “sleep on it”. It seems to be pretty characteristic of people whose jobs are mostly as problem solvers, and interestingly you see a lot of executive-types with the habit of cat-naps taken as often as possible. JFK was particularly known for this.

    I haven’t really seen the topic come up here, yet, but there might be interesting material in consciousness in musicians during performances and/or deep practice. Speaking as a not-particularly-great guitarist, I’d say that any kind of recital performance is a mental state very different from, say, cooking dinner or washing the car or thinking about network architecture. Once you get “into the zone”, you’re doing things that you probably couldn’t do if you were actually thinking about it. This is especially true of improvisation. To be silly I could suggest that maybe the Scramblers are like an awesome jazz jam session. Nobody has any real idea going into it what they intend to get out of it, probably there’s not a lot of “normal consciousness” going on during it, and whatever comes out of it might be so powerful and unexpected that you could almost call it magic. And how does it communicate to the listener? Probably not to the conscious mind, but something is getting across, for a lot of people. Maybe we should think of the Scramblers, or some kinds of AI, as a jam session that makes product — starships or programming — instead of music. Thinking, but not thoughts.

  26. I presume at least some of the people on this thread will have read Ken MacLeod’s The Night Sessions? Artificial Intelligence and self-awareness as a side-effect of a really good theory of mind. Arising, in the book, accidentally in combat robots that have to anticipate the actions of humans, but that’s a plot device.

    Really cleverly done, and it makes for quite a good argument on the conscious mind as gatekeeper/lead cat-wrangler/gestalt rider. Which is, not coincidentally, a major role of the cerebral ganglion in insects. Remove the head of a praying mantis and it will happily fornicate for hours with no inhibitions. I remain unconvinced about the conscious mind being the last to know about decisions until we can produced a much more convincing demonstration of the _exact_ timing involved, along with a better understanding than I think we have about which bits of our brain count as conscious. Bird brains gave us a nasty shock on that front.

  27. Peter Watts,

    “Yeah, well, Descartes started by assuming that as an axiom and presumably felt no guilt at all as he vivisected screaming creatures. I think a lot of people still do.”

    There was obviously something very wrong with Descartes. People are affected by TV-series and movies despite knowing what they see is not real. Even animated movies can make some people cry or at least create strong emotions. Coldly watching an animal strapped to the table while it’s screaming and kicking when being tortured is not just the result of someone assuming animals are just soulless objects. It also mean there is something about your mind that is just not right.

  28. I believe the primary function of the conscious mind is to provide guilt and regret.

    In this model, consciousness is essentially an audit function. We have to recognize that the conscious mind is part of the environment for the unconscious mind. Thus, the unconscious mind anticipates the emotional response of the conscious mind, and treats minimizing subsequent conscious pain as part of the problems it faces. This explains why the unconscious decisions come first. This does not imply that consciousness does not affect the outcome, improve decision-making. It does so in th same way that tax auditors improve the accuracy of tax filings: by hurting you if you do a bad job.

    Note that regret-minimization describes the actual behavior of people making decisions under uncertainty better than the economist’s standard tool, expected utility maximization. See, e.g., Daniel Kahneman’s “Thinking, Fast and Slow”

    I believe it is also the case that the unconscious mind is relatively poor at choosing inaction from a menu of options So to guilt and regret we must add confusion and paralysis to the gifts of consciousness.

    Finally, I think the conscious mind is somewhat better at choosing short-term pain providing long term gain. Presented with ice cream and broccoli, the subconscious decision is ice cream every time. The conscious mind provides essential elements of self-judgment, self-loathing, and priggishness.

    All these things could have been stated in functionally-equivalent but more positive-sounding terms. I chose this formulation in an effort to appeal to the Watts esthetic.

  29. Biff:
    Peter Watts,

    “Yeah, well, Descartes started by assuming that as an axiom and presumably felt no guilt at all as he vivisected screaming creatures. I think a lot of people still do.”

    Biff replied –> There was obviously something very wrong with Descartes. People are affected by TV-series and movies despite knowing what they see is not real.Even animated movies can make some people cry or at least create strong emotions. Coldly watching an animal strapped to the table while it’s screaming and kicking when being tortured is not just the result of someone assuming animals are just soulless objects. It also mean there is something about your mind that is just not right.

    Now me:
    The famous French physiologist Claude Bernard wrote: “The physiologist is no ordinary man. He is a learned man, a man possessed and absorbed by a scientific idea. He does not hear the animals’ cries of pain. He is blind to the blood that flows. He sees nothing but his idea, and organisms which conceal from him the secrets he is resolved to discover.”

    Bernard’s most famous work, the occasion of his celebration of his deafness to pain and blindness to blood, was done on dogs.

    After a period of time studying the research procedures in Bernard’s lab, physician George Higgins wrote: “Having drunk that cup to the dregs, I cry off, and am prepared to see not only science, but even mankind, perish rather than have recourse to such means of saving it.”
    The Spectator, Volume 48, 1875.

  30. Andrew Hoerner: I believe the primary function of the conscious mind is to provide guilt and regret.

    So that is why psychopaths are so happy, hardly ever anxious and so prone to going completely off the rails. The guilt and regret might be provided, but like with emotions, or cues (sad faces) – they don’t attend or care about such stimuli, they’re metaphorically blind to such.


  31. Discovery of Quantum Vibrations in “Microtubules” Inside Brain Neurons Corroborates Controversial 20-Year-Old Theory of Consciousness

  32. Re: whoever
    Still does not explain how we don’t experience a severe consciousness malfunction inside a fMRI unit

  33. Y.: So I’m curious if you can teach people to solve complex problems subconsciously, without ever having them pay conscious attention to the complex problem.

    I bet you could technologically; we’ve already got rudimentary memory implantation in rodents using hippocampal prostheses, so the kind of implanted expertise described in Gibson’s “microsofts” and the Matrix movies is possible in theory at least.

    Whether we can do it without that kind of tech— through sleep-learning, or the flashing of subliminal messages— is a different question. Again, I think it’s possible (there’s plenty of evidence that the brain can process input it’s not consciously aware of), but as I recall all those attempts to employ subliminal advertising back in the sixties and seventies came to naught.

    Of course, we’re way better at it now.

    Mr Non-Entity: To be silly I could suggest that maybe the Scramblers are like an awesome jazz jam session.

    It may be silly, but it’s also a pretty cool analogy.

    Biff: There was obviously something very wrong with Descartes. People are affected by TV-series and movies despite knowing what they see is not real. Even animated movies can make some people cry or at least create strong emotions. Coldly watching an animal strapped to the table while it’s screaming and kicking when being tortured is not just the result of someone assuming animals are just soulless objects. It also mean there is something about your mind that is just not right.

    I dunno, Biff. Have you ever seen commercial fishers slamming live fish around on the deck and gutting them while they gape and flap? Either that profession selects for sociopaths, or people just naturally get desensitized to brutality after a while. Personally, I think it’s that second thing.

    Andrew Hoerner: I believe the primary function of the conscious mind is to provide guilt and regret.

    Huh. I had not thought of that. I must think on it some more, now.

  34. Peter Watts: Either that profession selects for sociopaths, or people just naturally get desensitized to brutality after a while.Personally, I think it’s that second thing.

    (Really should stop watching GoT, obviously.)

    Loads of good links and lots to read. Thanks Amazon for a kindle I can send websites (read: papers) to.

  35. re whoever
    Also, the only scientist who has “found” these fancy “warm quantum” vibrations is the Bandyopadhyay person, with little success at reproduction, which makes it a one-author-pony along the same lines as Cold Fusion or hafnium grenades

    If “warm qubit in biological systems” stuff was not bogus, DARPA guys would be legs-over-head rushing to have a stab at it and build a goddamn quantum computer that actually does useful work (and might be “self-aware and conscious” as a freebie per OrchOr hypothesis), which they are obviously not.

    It’s sad to see Penrose partake in what appears to be a crank circlejerk

  36. Epiphenomenalism has a long history in philosophy of mind. So do zombies (putative human persons who act exactly like real humans but without consciousness).

    In sf, one of the earliest stories I can recall dealing with the topic is Bruce Sterling’s Swarm (Magazine of Fantasy & Science Fiction, April 1982). Reprinted in Schizmatrix plus and no doubt one of his other story collections.