Mechanical or Biological?

27 posts

Flag Post

So while vika and I were discussing how to play god and get rid of genetic diseases, I slowly started to wonder about medical regeneration. More specifically, I was wondering whether regrowing limbs would be better than grafting a mechanical limb. I could see the appeals to both ideas, and as I understand, they’re both seeing their own early stages of development. So I ask you, which will be more prevalent in the near future? In the distant future? What complications would you expect from each decision?

Here are some examples of where both have been meeting with success.
Growing ears on mice, using stem cells.
Attaching robotic arms to a human being.

 
Flag Post

Or both? One of the problems I imagine with growing a full limb is the amount of raw mass. That amount of tissue takes a lot of time and energy, with things as they are I can’t quite see that amount of miracle grow being successful. That said, I think we’re nearing the point where we can wrap any given prosthesis in living tissue, giving it the look and relative feel of a natural limb.

Not quite limb regrowth but an interesting note to composite ideology and possibility, a performance artist named Stelarc grew an ear out of (mostly) his own stem cells and had it affixed to his arm. He then inserted a mic into it that could wirelessly transmit. He could listen to his arm-ear through a blutooth set, or transmit it to an internet feed, or what have you.

http://stelarc.org/?catID=20242

He’s a bit of a nutter, but quite the novelty. His current bent has been on reconstructing the body and interfacing with emergent technologies. He’s someone I imagine Vika may take some interest in certainly.

 
Flag Post

To be honest, in the near future, I think bionic body parts are more likely than regrown body parts. Partly because they’re already in use (e.g. fully mechanical heart) and partly because the biggest complication I can think of is nerve grafting, which I think is shared with the biological equivalents.

In the distant future, which becomes dominant really depends on how they compare to each other. If one is clearly a superior product, then I will assume that product will become more likely to be used. However, as bionic arms have much more flexibility in use and design, I believe they’ll be more popular. (I explain this a bit more later.)

As for complications? Well, for biological:

  • Time to grow the part will likely exceed manufacture time of bionic equivalent.
  • Likely not as sturdy as bionic equivalent.
  • Likely not as customizable. That is, a mechanical arm could possibly be altered to suit the needs of the user better. (For example, in The Moon is a Harsh Mistress, one of the protagonists had a mechanical hand with tools built into it. However, this assumes that a person’s nerves can adapt to this kind of change, which isn’t certain in my mind.)

For bionic:

  • Vulnerable to a different range of problems that may be complicating.
  • Potentially restricted for sports and other physical competitions, resulting in injured athletes requiring biological needs.
  • Possibly more upkeep required, and shorter life span. (Potentially not an issue in “the distant future,” though.)
 
Flag Post

I can’t quite find or remember the terminology, but despite being a transhumanist, I do not believe that any system, biological or otherwise, can be substituted for other materials. The makeup of silicon is different than carbon, and will necessarily provide different functional experiences simply because of their chemical structure. A brain could be simulated by a computer with advanced processing power some day, but I believe that 1:1 simulations are themselves real by requirement.

That isn’t to say we couldn’t one day create and destroy cellular structures and even create more efficient brains and body parts, but to truly say that we have alternative parts for these things is false. A brain would be a brain, not a “robot” brain.

I’m not sure I answered the thread’s question, though.

 
Flag Post

I think we will first build mechanical parts (which seem easier) and then begin to more and more replace them with biological parts which are closer to the origenal and therefor more of a cure.

 
Flag Post

Mechanical:

+ Quicker to make (depends on how fast the bio arm grows back)
+ Durable
+ Fairly easy to use

- Expensive
- Much more energy for user than needed
- Longer to make large quanities of


Biological:

+ Natural, the user will be used to it.
+ Reliable, shortage of metals you say? Oh well grows back arm
+ Less Energy needed.

- Slower to grow back (depends)
- Possibly weaker
- During healing, arm is open to infections and possible mutations.


I believe that both are equally good, but each shine out in differant situations.

 
Flag Post

To be honest, in the near future, I think bionic body parts are more likely than regrown body parts. Partly because they’re already in use (e.g. fully mechanical heart) and partly because the biggest complication I can think of is nerve grafting, which I think is shared with the biological equivalents.

Regrown body parts are also in use now, to a limited extent. This story is four years old now, and although it hasn’t made the headlines since, I did see a television programme a while back about some very promising results that the US military has had with this stuff. This Wiki article mentions a young man in California who regrew his heel after the application of pixie dust.

I posted this link a few months back in another thread. It’s an interesting story, so I make no apologies for bringing it up again. If we can introduce a spider gene into goats to produce silk, it makes me wonder if we could introduce something like a salamander gene into future generations of ourselves so that we would regrow any body parts lost during our journey through this vale of tears. We all grew from a single fertilised egg cell, so we all start out by being able to generate and regenerate body parts. Somewhere along the line we lose that ability, but I have read that if you amputate parts from an early foetus, those parts will simply regrow as the foetus develops. There’s a ton of research to be done on this, but it does offer some very exciting possibilities.

 
Flag Post
Originally posted by TheBSG:

I can’t quite find or remember the terminology, but despite being a transhumanist, I do not believe that any system, biological or otherwise, can be substituted for other materials.

Substrate chauvinism.

A brain would be a brain, not a “robot” brain.

Not true. Whilst the quantum-level effects would be different, so long as the same thought processes can be laid out on the new material, even if they have to be physically laid out a different way, so long as the way they process data is identical, it is still a brain.

  

Obviously I am going to be biased strongly towards mechanical limbs in my answers here, because of the nature of my work. However, both are quite considerably more advanced than the OP might suggest.

If we consider a current-generation haptic-prosthetic arm for instance:


This is an earlier version of the modern iLimb, as modelled here by the first woman ever to receive one. We have advanced fantastically in the six years since this was first fitted, but there are still a whole host of problems.

  • The limb has to be lighter than the original arm. Max of 30% original weight. This is because bone implanting is still a bit of a hit-and-miss science. We have newer methods in the works, but none have been deployed in human patients as of yet. So, since we are anchoring in soft tissue, it cannot take the same weight bone can.
  • Repairing damage is theoretically possible, but not possible in current models. The best we can do is to use sensors to indicate how fr damage has spread.
  • Cost. Even the cheapest smart prosthetic on the market will still set you back $10,000. More advanced models climb as high as $80,000
  • Neural Codes. We still don’t have a complete catalogue of neural codes for the peripheral nervous system. Whilst work is continuous, the limb is not as capable as the original, yet.
  • Sensor density. At it’s highest, the density of touch sensors in human skin approaches 2,500 per square centimeter. At it’s lowest it is closer to 4 per square centimeter. Both of these are currently beyond our ability to integrate into a prosthetic, because of the bandwidth needed to process input over the whole arm, heat dissipation and weight restrictions. Work is progressing, but it is slow.
  • Freedom of movement. Certain joints such as the knee and the wrist, are a nightmare to replicate from an engineering point of view. Whilst we have reasonaby good wrists now (and have had for 15 months), the complexity of the knee still eludes us.
  • Rejection. Fighting the possibility of infection at the implant site, and rejection of the implanted materials is still a major work in progress. New materials are being tried, in an effort to minimalise the problem.

Now, that said, there are some wonderful pros with prosthetics, that you don’t get anywhere else.

  • Hot-swappable limbs. Unplug one hand, plug another one in. Differnt limbs with different capabilities plug into the same implantation socket. Just broken your leg? No problem. Pop it out, pop another one in.

One lady even has twelve sets of legs, so she is anywhere from 5’.0" to 6’.0" depending on how she feels that day.

  • If the limb does break, you can drop it off for repairs and go do something else for the week. Try doing that with a broken natural arm or leg.
  • Ability to integrate custom electronics into your arm. I’m thinking in particular of the new ones from Holland with a media player embedded in the upper arm. Integrated radio or computer access within your body.
  

On the other hand, biological implants have been tried. This case was the most extreme. Ultimately, the patient had his new hands amputated again, as he couldn’t cope with them psychologically – they felt foreign, like the hands of another man. They didn’t feel like they were part of him, and he could not cope with that.

Like proisthetics, biologial limbs that are grown from your own body as opposed to transplanted, are going to have problems:

  • Long regrowth time
  • Very likely to be rather painful to regrow complex systems such as arms, as spinal regrowth will also be necessary to replace the damaged nerve cells.
  • Weakening of the immune system as new structures are force-grown.
  • Extra strain on the heart for the same reason.
  • Will be absolutely useless for patients with genetic defects, as the replacement limb will grow into the same distortion pattern as the old.
  • Just as fragile as the original limb.

But, on the flipside:

  • Natural connection to the motor neurons and cerebellum means no decoding of neural signals required.
  • Easy to master. Relearning to use the limb fully may take as little as six months.
  • It is an integral part of your body.
  • Perfect match with your build / skin coloration / bone mass is naturally achieved. You will have to work hard to replace lost muscle mass of course.

Ultimately, I would go with prosthetic limbs, but then I’m biased of course. To me they raise fewer possible health concerns, and what defects they do have, we can tackle in time.

 
Flag Post

I think that really depends on the extent to which prosthetics are developed.

It is possible that we might create a machine that performs a function better than the biological organ that it is meant to replace.

 
Flag Post
Originally posted by fma1:

It is possible that we might create a machine that performs a function better than the biological organ that it is meant to replace.

Oh there’s no doubt that we will ultimately get there, it’s just that some people may prefer a biological limb despite that, or it may be dictated to us by virtue of the events we wish to compete in, or status we wish to hold in society.

 
Flag Post

No, substrate chauvinism doesn’t describe what I’m talking about at all, I think you’re misunderstanding me. I’m not making a moralistic argument, I’m making a mechanical, physical argument. The only substitute for a brain is a biological brain. There is no silicon based “version” of a human brain because then it wouldn’t be a human brain by definition. If you had the same efficiency and chemical relationships, it would simply have the same chemical composition. I am saying that the nature of the neurons in our head are based on their makeup, and thus any alternative makeup would be something different, a simulation of that specific relationship that arises from our makeup. It’s not a limiting belief at all, in that I simply think efforts to simulate biology are mislead, while efforts to recreate biology make far more sense to me. I’m not even against the use of simulated biology, just that it is inferior to the goal of reproducing human cognition. I really really wish I could remember the specific term for this argument, because others have said it far better than myself.

I hope this illustrates it: There might be a material that is more efficient than our brain’s materials, and thus we could simulate and recreate all of the features of the brain even more effectively than the brain itself could, but that would still be a separate thing from a human brain in itself. I’m not devaluing the artificial brain, and I don’t even like the term artificial. I’m simply saying that the biological relationships between the materials that make up a given object cannot be directly simulated to contain all of the emergent and interrelated effects of those systems. There is a different number of electrons in silicon than in carbon, so it’s going to react differently to the elements within that system than carbon itself would act.

Second try: If I have a wooden structure that supports a stage, and I recreate it out of metal, it will function differently by it’s very nature, even if it functionally serves nearly the exact same purpose. It might have more less springiness, require less materials, but more braces to support its own weight. Unless we recreated the same object out of wood, we wouldn’t experience the same qualities. I’m not suggesting one is more superior or inferior to the other, but that a perfect simulation of a brain would be a brain itself. God dammit the word is on the tip of my tongue. Something to do with makeup, artifice, simulation, abstraction, god dammit. I hate when I can’t find something.

 
Flag Post
Originally posted by TheBSG:


Second try: If I have a wooden structure that supports a stage, and I recreate it out of metal, it will function differently by it’s very nature, even if it functionally serves nearly the exact same purpose. It might have more less springiness, require less materials, but more braces to support its own weight. Unless we recreated the same object out of wood, we wouldn’t experience the same qualities.

I understand what you are trying to say now, but I still disagree. If the mind contained in that structure is represented by the user upon the stage, then it does not matter whether the supports are made of wood or of metal, they feel the same to the person performing on them. They provide a foundation for the same ideas and expressions, the same modes of thinking et al.

Yes, the quantum interactions are different, but if we have mapped the full range of expressions possible on the old medium, providing we map the expressions the new medium is capable of onto them, we still have the same brain. The specific interactions are different, but they are interpreted as the exact same interactions the original material had, all the way up.

Computationally, it is then the same brain, and will process data in the self-same way, regardless of substrate.

It is almost like in computing, where you place an operating system between the hardware and the application layer, so the application runs the exact same way on many different computers, completely oblivious to what the hardware actually is, and the differences in wiring between different manufacturers.

 
Flag Post

You’re assuming we have complete knowledge of the structure itself, and I would argue that if we had complete knowledge of the brain, we’d be able to just use biological materials to construct it perfectly, instead. Otherwise, there could be emergent interactions in the relationships that we aren’t particularly aware of, and might not even be aware of once we use that medium to upload our minds to, as we can hardly describe the emergent experiences we have with the substrate of our brain as it is. I guess I’d be willing to agree that, assuming perfect knowledge of the brain, a simulation could be created that included all of the relationships of that brain. As it is now though, we don’t have perfect knowledge of the brain, and I’d argue that we never will, simply by the nature of knowledge, not because I think it’s infinitely complex or anything stupid like that.

I might go crazy if I can’t find this fucking wikipedia article, I’m not kidding. I keep trying to get back to work and I suddenly come up with another buzzword that might help me find it, and I fall back in for another 10 minutes. Usually I find stuff because I remember the name of the person who developed the argument, but I can’t remember for the life of me who made this argument, and where I would go to find it. I’m currently scouring AI articles looking for this paradox (if it even is that), with no luck.

 
Flag Post
Originally posted by TheBSG:

You’re assuming we have complete knowledge of the structure itself, and I would argue that if we had complete knowledge of the brain, we’d be able to just use biological materials to construct it perfectly, instead.

You and I both know there are serious issues inherent in constructing biological systems of that complexity and cell density. Besides, there are many instances when it would be advantageous for the brain to be larger, or a great deal tougher. You run up against biological hard limits if you go too far in either direction.

Otherwise, there could be emergent interactions in the relationships that we aren’t particularly aware of, and might not even be aware of once we use that medium to upload our minds to, as we can hardly describe the emergent experiences we have with the substrate of our brain as it is.

I’m well aware there will be problms with the first generation or two of artificial minds. Uploaded as well as silicon-grown.

I guess I’d be willing to agree that, assuming perfect knowledge of the brain, a simulation could be created that included all of the relationships of that brain. As it is now though, we don’t have perfect knowledge of the brain, and I’d argue that we never will, simply by the nature of knowledge, not because I think it’s infinitely complex or anything stupid like that.

I think we will, but it’s going to require a lot of intermediate tech, both to reverse engineer the full complexity of the mammilian brain, and to develop the data handling technologies (and hardware to run them on) to deal with datasets that massive.

I might go crazy if I can’t find this fucking wikipedia article, I’m not kidding. I keep trying to get back to work and I suddenly come up with another buzzword that might help me find it, and I fall back in for another 10 minutes.

Kurzweil’s site might yield better results than Wikipedia. What exactly is it you are looking for?

 
Flag Post

I am looking for the term that describes how a simulation of a given system becomes the system itself when it is perfectly efficient, and is otherwise not a perfect simulation until then. It has something to do with the words material and medium, and is directly related to complexity and computational theory. I think it’s actually a mathematical principle, too, and is somehow related to P=NP.

 
Flag Post

This was the argument I think you were referring to before, BSG, with regards to the substrate-limitations of cognitive systems.

As to the other, I thought at first you were describing pattern theory, but I don’t think you are.

 
Flag Post

Oh christ, thank you. That article included a link to the Kullback–Leibler divergence, which is what I was referring to. I can finally get back to work. Edit: This is a better article that applies the theory to limited and infinite data, and can be extrapolated to include theories about complexity as related to computation.

Also, that article you posted is entirely agreeable, in that I think the effective simulation of a brain is as good as a true simulation if we maintain certain goals. I still think that biological recreation is actually more desirable though, and doesn’t imply complete biological efficiency, and thus provides for improvements and advances using both biological and technological additions. I just think that starting with a simulation is less than ideal, while starting with a fully realized recreation of a brain allows for deeper mechanics and tinkering without disrupting true transition to the new substrate.

4th edit: I would also concede that, as we have learned about our brain, any problems with substrate can be solved with rationalization mechanisms, as mentioned in the article. We already “round down,” or (in the case of drugs) “round up” any unintelligible information. We are incredibly good at generalizing a thing that is not specific, and thus, I’m not sure we necessarily need a perfect simulation for undisrupted cognitive transitioning, as Koene notes.

 
Flag Post

I found this relevant.

 
Flag Post

We’ll probably all end up(if possible. most likely is) on a small machine as a consciousness. Everything else would be a waste.

 
Flag Post
Originally posted by Galdos:

We’ll probably all end up(if possible. most likely is) on a small machine as a consciousness. Everything else would be a waste.

Arguably, we already are. The organic body is after all, a small, complex mahine.

 
Flag Post

Ooh, we could be like the Geth! Millions and millions of people inside computers, thousands inside each body.

(I doubt this will ever happen, but still.)

 
Flag Post

I think they’ll both stick around for a while.
People like choices, regardless of whether or not one is superior- even if we get mechanical substitutes that work just as well as the real thing. I for one would like to keep the squishy bits- I feel attached to them. I’m sure other people will feel the same way, and if I lost my legs in an accident, I’d want to drop the money and time on biological replacements, and if there are enough people out there who feel the same way, then there will be companies providing that service. It’s the flip side for mechanical- unless we see some sort of mass societal movement towards all mechanical bodies, then we’re going to see both. And there is always going to be a counter-culture movement. So it’ll probably be pretty distant in the future before mechanical overtakes biological, if both are an option.

 
Flag Post
Originally posted by Ketsy:

Ooh, we could be like the Geth! Millions and millions of people inside computers, thousands inside each body.

There are real problems with that approach. If you are putting multiple separate individuals in the same body, so their minds flow over the same circuitry, use the same resources, how do you guarantee that these separate minds stay separate minds, rather than bleeding over into each other, and forming a single mind?

It is a genuine concern, and one we will have to deal with, as we increasingly network our brains directly. A multiple embodiment in the same artificial brain has the problem raised orders of magnitude worse than this. How do we keep these millions of separate minds as separate individuals?

 
Flag Post

I think that the difference biological and mechanical currently lies in the complexity and the amount of outsourcing that can be done.

Mechanical replacements and argumentations are less complex but its also easier to outsource tasks that are necessary for their function. They can be built outside of functioning immune system while biological replacements still need a host to grow in. The energy necessary for the use of mechanical parts(if needed) can easily come from batteries, bypassing the rather ineffective food processing of the digestive system(and all the other biological steps normally involved in distributing energy). Often mechanical parts can be replaced without surgery, switching a biological hand on a biological arm is a bit harder than a mechanical hand on a mechanical arm.

Down part is that the less in complexity generally also means a less in number of functions. A biological arm has many more useful functions than even the most complex mechanical arm available.
Some of the less recognized functions are for example the ability to help regulate the body temperature, store energy(f.ex. fat), store blood, helping in producing parts of the immune system.(Sure the human body can generally handle such stuff even when missing all four limbs, but not nearly as good as with them.)

On the other hand to get all these functions done biological systems often have to compromise. Making mechanical system able to out preform them when trailered to specific functions.

In the future we will see a raise in the complexity of mechanical systems and the other hand a greater amount of outsourcing and specialization of the biological systems. A point will come when they become one and the same in their ability.

 
Flag Post

Johnny, I agree with most of what you’ve said, however disagree on a few technical aspects. Mostly because of subject-speciality knowledge which you won’t have had.

The energy necessary for the use of mechanical parts(if needed) can easily come from batteries, bypassing the rather ineffective food processing of the digestive system

would it surprise you to know that battery-power for implanted prosthetics is considered a poor solution?

A battery will eventually deplete, and require a new operation to replace. Far better to have the prosthetic draw power directly from the body’s own stores. So that’s what we’re doing. Piezoelectric sheets draw power from every movement of nearby muscles. These can be implanted elsewhere in the body and charge broadcast by RF back to the implant, which uses a trickle-charge chemical capacitor (These actually exist, they’re being tested in the lab at this time).

The idea being, whilst a battery will only last for ten years at most, the piezoelectric method will keep running for fifty years or more, as long as the patient does. Once the patient dies, the capacitor will slowly drain, and in as little as six-months max, will run dry completely. Once the patient is dead, there is really no need for it anymore, but there’s still enough charge left, for whatever action needs to be taken once they have stopped moving, but are still-alive.

Some of the less recognized functions are for example the ability to help regulate the body temperature, store energy(f.ex. fat), store blood, helping in producing parts of the immune system.

Again, modern prosthetics can do this if they have to with the capabilities we have available. They don’t because at the moment, weight and space are at an absolute premium.

Regulating the body temperature is perhaps the easiest, because the motors generate waste heat. It would be within feasabiliy to pump that back into the body, if a sensor detects an alarming drop towards hypothermia. Excess heat draining would require a more drastic use of engineering, and methods not currently employed by our prosthetics.

Storing energy can be accomplished by deriving energy from blood sugar, as again, has been accomplished in the lab. A chemical storage system could be engineered to return energy to the bloodstream dependent on need. Presumably a biochip system would need to be present in-situ to determine this need.

I know Franhaufer are working on a diabetic implant using this method – automatically monitors blood-sugar level in real-time and activates an insulin pump when required. It’s the same principle in action.

Blood storage is impracticl at this time, for many obvious reasons. An axiliary immune system may be possible at the implant site, using artificial vein tissue around the implant site, and methods to detect and remove harmful pathogens (I read somewhere that one group was working on this method to stop cancerous material circulating in the blood, but I cannot remember what the study was called, for love nor money).

A point will come when they become one and the same in their ability.

Worse. Because mechanical systems can consist of multiple specialised layers fitted together, a point will come when mechanical is far superior to biological from a utility and function point of view. Biological will keep the edge in uncontrolled growth, most likely.