transhumanists at h+ vs. a skeptical biologist

January 7, 2010

H+ Magazine tends to be very enthusiastic about the Technological Singularity espoused by Ray Kurzweil and transhumanists who like his ideas, so it caught my eye to find an article rather critical of some proposals by a number of Singularitarians published alongside the usual futuristic optimism. And no wonder. Its author is biologist Athena Andreadis who studies gene regulation and writes science fiction and pop sci posts across the web. Her biggest problem? The sheer disregard many Singularitarians have for the biology involved in our brains and the fact that body and mind are a package deal. You don’t have one without the other and the ideas of uploading minds to a computer while tossing the bodies into a waste disposal unit simply doesn’t work. In other words, you can replace parts of your body until you’re mostly machine, but your brain is there for good.

By this point, the article is a few months old. Having found it soon after it was published, I saved it alongside my bookmarks with post ideas but managed to lose sight of it until now. Nevertheless, this is an article that’s just as valid today as it was in mid-October and there are even more things to talk about after some of the big online promoters of the Singularity made their marks in the comment section, including Jake Cannell, who’s made an appearance on this blog as well to argue the exact same proposition he argued at H+ a few weeks before. He seems to be a pretty driven man. And why not? To be able to discard your body in favor of anything your heart desires is an awesome proposition. It would be like having a real life avatar you could style after a cartoon character, give yourself the kind of body that would turn a bodybuilder green with envy without a single trip to the gym, or experiment with the looks of a gorgeous porn star. The possibilities are only limited by what shell bodies are in stock. To believe that something like this is possible in our lifetime is definitely hard to let go. Sadly for transhumanists though, the reality of the matter is far more complex than that.

But why let science spoil the fun when one can snarl at scientists for being wet blankets with no appreciation for the wise and powerful beings of the future who can surely figure out how to do the impossible, according to a number of commenters on the article. As one of them says in his complaint…

This whole discussion is ludicrous, like cavemen discussing the future possibilities of the human race, envisioned entirely in terms of whether stone weapons can solve the cave-bear problem. We have literally no idea whatsoever what the lives of our descendants will be like in 10,000 years — not what their problems will be, nor what their capabilities will be.

There’s a point here. We really don’t know what life will be like ten millennia from now. Our descendants could be exploring the galaxy, or have managed to fulfill Einstein’s dire exercise in prediction and be fighting World War 4 with sticks and stones. But to say that such a fundamental thing like the limitations of our bodies might not be an issue in the future and the biological barriers to such things as mind uploading will disappear, is an exercise in boundless futuristic optimism which has a very mixed track record when it comes to making our dreams come true. I’ve pointed out before why mind uploading wouldn’t work and reiterated those reasons in my debate with the Singularity Institute’s Michael Vassar. Computers do not work like human brains. That’s not how they were designed. Rebuilding them to simulate biological forms would mean taking the current and long proven von Neumann architecture, throwing it out the window, and starting from scratch.

There are some efforts to give computers a capability similar to thought via brain modeling projects but these experiments are more hype than anything else at this point and even if they turn out to be successful, the end product would have little to no practical use. However, instead of trying to be more like machines, we can do what’s already being done in medical and computer science labs across the world and make computers work with us. We already know how to make thought controlled computers, how to build robotic limbs that function as if they were our arms and legs, and even how to turn thoughts into synthesized speech. Becoming part machine is how we could extend our lifespans, replacing organs that wear out for synthetic substitutes which can be repaired, along with the biological modifications suggested by Dr. Andreadis. Of course there’s a limit to how much we can replace. Touch any of our vital organs like the heart, liver or lungs, and you risk killing the brain. Mess with the circulatory system too much and you’ll risk stroke and starving the brain of oxygen.

Ultimately, we don’t know what the future will bring. Still, to task our descendants to make us into immortal bits of software and replying to very valid and pertinent scientific and engineering objections by saying that a future society will surely solve whatever scientific hurdles we face in making it happen, is just another way of pointing to the sky and saying that someone, somewhere will answer your prayers and build the magical technologies so we can transcend being human. But that’s not going to happen. We evolved as humans and this is exactly what we’ll be, no matter how many machine parts we add to our bodies or how far away from Earth we venture in the distant future. And our messy organic brains will always be with us.

[ cyborg illustration by Fydor ]

Share
  • Jason

    I wonder if we’ll ever create “life” that mimics how we think and what we do, without our biological limitations. If it were able to evolve/adapt and create new things the way humans do, it could become humanity’s successor. No more need for our squishy meat brains. Robot overlords FTW. ;)

  • Greg Fish

    I really don’t think we can create anything that would qualify as living without using organic materials. Machines function with digital signals and by using a static set of orders modified by us. Living things use something we could loosely call analog signals and their brains are highly plastic. Try and build a robot with that setup and you’ll end up with an unpredictable mess.

    I’ve heard much noise about evolving robots but I’ve yet to see any real evidence for one. The closest thing to self-evolving programming with which I’ve had the chance to work was a a highly unstable homunculus of spaghetti code.

    Machines are built to do a specific task within certain tolerances. Change what they do and how on the fly and they’ll start going to pieces. By contrast, living things are persistent biological reactions that evolved to keep going despite constant changes in their makeup and even use them to their advantage. This is why machines can’t work like living things. They’re fundamentally different.

  • DamianD

    “Try and build a robot with that setup and you’ll end up with an unpredictable mess.”

    Some might argue that an unpredictable mess is a pretty good description of most human beings. :) In all seriousness, though, I agree with you that constructing a machine that can mimic the way a human brain works is probably a pipe dream. If nothing else, the ability physically heal damage simply wouldn’t correlate with the way you fix a broken component in a computer. The brain would be forever altered by an injury, even if the alteration is slight. A computer could simply have a damaged component replaced, returning it to the exact configuration it had before the damage. As you said, they are fundamentally different.

    And I agree that at present, there just isn’t enough evidence to support the hypothesis that a program could mimic the process of human thought, but that has more to do with technological limitations inherent to the current level of technology we’re working with than anything else. If there’s going to be a synthetic facsimile of the human mind, a software based model is the place it’s likely to happen. And while we may not see that in our lifetime, I wouldn’t rule it out as our technology gets more complex. The article you posted on brain mapping is an example of an interesting step in that direction, though only time will tell if that particular venture bears fruit.

  • http://www.starshipreckless.com/ Athena Andreadis

    Many of the responses at H+ were spoiled brat tantrums at having a pinch of reality thrown at wet dreams. As for Jake Cannell, why should any time or attention be given to a game designer’s “theories” about how the human brain works? This is even more hare-brained than asking the first person you encounter to fix your car, give you a haircut or fly you to Miami.

  • Robert Mugabe

    I enjoyed the article, but you seemed to get held up with the whole mind-uploading idea – that’s not the be all and end all of transhumanism. Yes, it is often cited as a goal for humanity but it’s not a fundamental tenet of the movement.

    Also, you said, “touch any of our vital organs like the heart, liver or lungs, and you risk killing the brain”. What about things we already have, like pacemakers and ventricular assist devices? Admittedly they are not replacements, but it is not an inconceivable leap of the imagination to think that, some years down the line, whole organs could be replaced.

    On your last point, you’re right that we can’t expect our descendents to magically solve all the scientific obstacles involved, but what we can do is strive to do as best we can. If some of the more outlandish transhumanist goals are not reached we will surely have benefited from the journey. And who knows, little by little, generation by generation humanity may slowly reach those goals – this hope is not “just another way of pointing to the sky and saying that someone, somewhere will answer your prayers”, its believeing that each generation will improve upon the developments of the last, that our actions will have an impact on our descendents.

  • http://www.enterthesingularity.com Jake Cannell

    In regards to Athena’s comment: I’m not a game designer, I’m a software engineer, and I’m not a theorist proposing a new understanding of how the brain works. To any extent I’ve written about that, I’m summarizing the work of others.

    gfish:

    I agree with you completely that current brain modelling projects are more hype than anything else, but you seem to believe that intelligence is inherently non-computational in nature: that it is a form of information processing that is forever incompatible with human designed computing devices. I find it hard to follow that. What is it in particular? Is there something magical about neurons? Its certainly not just analog vs digital or parallel vs serial. We can and have built analog computing systems – digital systems are easier to mass produce now with careful design tolerances and a digital system can emulate any analog system. The serial von neumman architecture is definitely a bottleneck, but that’s been known for a while now and we are moving towards increasing parallelism.

    “Machines are built to do a specific task within certain tolerances. Change what they do and how on the fly and they’ll start going to pieces”

    This is true of simple machines, but not learning systems. The field of AI is concerned with developing systems that have learning and intelligence: capacities to perform a wide range of tasks and the ability to acquire and improve those capacities based on experience and interaction with the environment.

  • Bill

    The sheer disregard many Singularitarians have for the biology involved in our brains and the fact that body and mind are a package deal. You don’t have one without the other and the ideas of uploading minds to a computer while tossing the bodies into a waste disposal unit simply doesn’t work. In other words, you can replace parts of your body until you’re mostly machine, but your brain is there for good.

    Who’s arguing that the mind can somehow be separated from the brain and uploaded to a computer? The only proposals I’ve ever heard involve uploading the structure of the brain and emulating the whole thing in the computer. You’re simulating the brain in order to simulate the mind. It’s still a package deal.

  • Bill

    Machines function with digital signals and by using a static set of orders modified by us. Living things use something we could loosely call analog signals and their brains are highly plastic.

    1. Not all machines/computers are digital.
    2. Completely-digital machines can simulate analog/”plastic” processes just fine.

  • gfish3000

    That’s not really a viable idea. See the following.

  • gfish3000

    For the purposes of answering the basic, broad question, the simplification was enough. One can write an entire multi-volume thesis on computing and philosophy of what to call living or not and whether simulation of something is sufficient to be compared to the real thing.