On the other hand, you should consider what evolution can do. Evolution is not the world’s best algorithm for inventing things. However, it is an excellent optimising algorithm. Balancing multiple considerations to decide the optimum amount of substance A in your body is the sort of problem that algorithm should do really well.
Essentially the only exception to this rule is when your cells are reacting to DNA/RNA that doesn’t belong to you. If cold virus RNA is making your nose run, stop it by all means. But you should trust your own body on most other matters—adding extra chemicals is likely to turn out worse....
Note what’s being optimised here—not intelligence, but biological fitness—how likely you are to reproduce successfully. You might improve intelligence somewhat, but there if there isn’t a downside somewhere then Darwin was wrong.....
On the other hand, you should consider what evolution can do.
It frustrates me how often this argument against using mind enhancing substances is used and, more importantly, the weight it is given. Not only is evolution optimizing for different critiera (which DuncanS mentions) it is also optimising for an entirely different environment. Further, our expectations that random chemicals will be bad for us is to a massive extent screened off when we go ahead and test them and find that they make things better!
Yet another situation in which evolution should not be expected to give superior results to what we can come up with with science is when we know what we are going to be doing at a specific time. What is best as a general baseline is not going to be the best state when studying for a test. Which is in turn going to be less good when doing unpleasant and potentially traumatic things that you don’t want to remember.
Consuming chemicals that have been tested is certainly an improvement on consuming chemicals that haven’t been.
Consuming chemicals to make your brain work better seems to me to be a rather similar activity to overclocking a computer. Let’s add more voltage. Let’s pour liquid nitrogen into it. Perhaps it will go faster ! Perhaps it will, but will it still be working in 5 years time?
First of all, note just how crude these efforts are compared to the technological research undertaken by the companies that actually make microchips. The same is true of the brain—it can make dopamine and deliver at synapses—exact points of contact throughout the brain. Yet you see people discussing just adding more dopamine everywhere, and thinking that this is in some sense improving on nature in a clever way.
I have to mention a point against myself—which is that I do take general anaesthetics, which, while not an intelligence enhancer, is definitely an intelligence modifier for specific circumstances. However, turning brain function off is arguably simpler than trying to make it better.
It is possible, definitely, to improve human intelligence by combining it with a computer. So it’s not the case that I’m against the idea that it’s impossible to improve on the natural intelligence we all have—it obviously is.
What I’m pointing out is that all of these drug ideas are bound to be something that evolution has at some point tried out, and thrown away. And they are really unsophisticated ideas compared with those the brain has actually adopted.
Even the situation dependent argument isn’t as strong as you might think—for example your brain has a lot of adaptations to cover the “unpleasant and potentially traumatic things” situation, for example—and these adaptations generally disagree with your view that you shouldn’t remember them. It’s probably the case that intelligence tests are a novel environment, however....
What I’m pointing out is that all of these drug ideas are bound to be something that evolution has at some point tried out, and thrown away. And they are really unsophisticated ideas compared with those the brain has actually adopted.
What I’m pointing out is that all of these drug ideas are bound to be something that evolution has at some point tried out, and thrown away. And they are really unsophisticated ideas compared with those the brain has actually adopted.
Well there could be many reasons why evolution has” thrown them out”. Maybe they are harmful in the long term, maybe their use consumes precious energy, or maybe they just aren’t “good enough” for evolution to have kept them. That is, maybe they just don’t give any signifigant evolutionary advantage.
Evolution doesn’t create perfect beings, it creates beings which are good enough to survive.
There can be harmul side-effects and that topic is not covered by the article; on the other hand, pure evolutionary argument can be doubted because of changed environment.
If I stimulate my brain, it is natural to assume my brain requires more energy now. So I probably need more glucose. In evolutionary relevant context, that would make me more likely to starve—after all, I would need more highly valued energy and thinking clearly wouldn’t make a killed bull magically appear before me.
This is still true for the most of the Earth’s population. It is not true for many of LessWrong readers, though. There are some primarily-mental jobs now (in some places of the world—the places where LessWrong readers come from). Keeping more things in you mind means being a better programmer, teacher, scientific researcher. Being better at your profession often helps you to evade starvation. And getting needed amount of calories—if you already know where to get all these vitamins and microelements—is trivial in these parts of the world.
So, this modification was not a benefit earlier, and it was quite costly; both factors are significantly reduced in some parts of modern world.
Of course, increased mental capability can lead to some personality traits that make it harder to reproduce; but that is again a question of side-effects and not a self-evident thing. If you consider it harmful, you can try to spend effort on fighting these side-effects—some people report significant success..
Same here. As a childfree human, maximizing the number of copies of my DNA is right up there with paperclip maximization on my list of priorities. :-)
One other categories of exceptions: We aren’t in the EEA anymore. In particular, we have much looser
constraints on available calories than in the EEA, and that changes the optimal settings even for
reproductive success.
There’s this really pretty large class of issues called “genetic disorders”, and a wide variety of other ways the body fails just fine without encountering foreign DNA/RNA… I’m assuming insulin for diabetics also has unexpected drawbacks and isn’t really in our best interests?
Or, put succinctly: “Scientists are so ignorant! If it was possible to cure cancer, why didn’t we just evolve to not have cancer in the first place?”
I have heard of genetic disorders, and know that they occur, and why they are found in the gene pool. And in such cases, it’s entirely appropriate to take medication for them. I take your point that perhaps I could have mentioned this class of DNA earlier, and that it is appropriate to take medication for that class of diseases, as it doesn’t improve your quality of life to leave it as it is.
I didn’t think we were discussing genetic diseases particularly. I am convinced of your argument that if you have a genetic disorder that affects your intelligence, you should take medication for it. I don’t see why this is relevant to the more general case of people who don’t have a particular genetic issue. Evolution is a good balancing algorithm—but it works by trimming the outliers. If you are unlucky enough to be one of the outliers, there’s likely something that needs correction. But generally, I don’t see the relevance.
As for your second point—well, we did. Cancer is very rare in the natural environment. It’s only in our much safer modern environment that we live long enough for it to become a problem again.
Given the genetic variance in IQ, it’s obvious that most people don’t have optimal genetics for intelligence. Whether this is a “disorder” is an interesting semantic question, but the point remains that we know that the general class of “human brains” has a maximum that’s higher than where most people are at.
Equally, our brains evolved for a much different environment with much different trade-offs. Just like cancer wasn’t a threat in our ancestral environment, the ability to do a second order differential equation wasn’t a benefit.
In short, medical science suggests that, actually, there’s plenty of room to improve humans, both because we’re extremely inconsistently built, and because we’re not built to handle our current environment.
Actually I agree with all of this—there’s a tremendous difference between average intelligence and the top end of the bell curve, and we have no reason not to think it can’t go higher. We are the first species on this planet to attain general purpose intelligence, and there’s no good reason at all to think that either more isn’t possible, or indeed that the process of human evolution in this respect has stopped—quite the reverse I suspect. We have every reason to assume that at the moment human intelligence is evolving like mad—it’s being very strongly selected for in a very large gene pool.
But my point is that this all has nothing to do with the proposed methods of improving the brain. If we really knew how it worked, and were able to model the consequences of our actions better, then it would be less of a guessing game whether there was a longer term price to the short term gain.
On the other hand, you should consider what evolution can do. Evolution is not the world’s best algorithm for inventing things. However, it is an excellent optimising algorithm. Balancing multiple considerations to decide the optimum amount of substance A in your body is the sort of problem that algorithm should do really well.
Essentially the only exception to this rule is when your cells are reacting to DNA/RNA that doesn’t belong to you. If cold virus RNA is making your nose run, stop it by all means. But you should trust your own body on most other matters—adding extra chemicals is likely to turn out worse....
Note what’s being optimised here—not intelligence, but biological fitness—how likely you are to reproduce successfully. You might improve intelligence somewhat, but there if there isn’t a downside somewhere then Darwin was wrong.....
Or as Eliezer puts it:
But here’s gwern writing about about loopholes in Algernon’s Law.
It frustrates me how often this argument against using mind enhancing substances is used and, more importantly, the weight it is given. Not only is evolution optimizing for different critiera (which DuncanS mentions) it is also optimising for an entirely different environment. Further, our expectations that random chemicals will be bad for us is to a massive extent screened off when we go ahead and test them and find that they make things better!
Yet another situation in which evolution should not be expected to give superior results to what we can come up with with science is when we know what we are going to be doing at a specific time. What is best as a general baseline is not going to be the best state when studying for a test. Which is in turn going to be less good when doing unpleasant and potentially traumatic things that you don’t want to remember.
Consuming chemicals that have been tested is certainly an improvement on consuming chemicals that haven’t been.
Consuming chemicals to make your brain work better seems to me to be a rather similar activity to overclocking a computer. Let’s add more voltage. Let’s pour liquid nitrogen into it. Perhaps it will go faster ! Perhaps it will, but will it still be working in 5 years time?
First of all, note just how crude these efforts are compared to the technological research undertaken by the companies that actually make microchips. The same is true of the brain—it can make dopamine and deliver at synapses—exact points of contact throughout the brain. Yet you see people discussing just adding more dopamine everywhere, and thinking that this is in some sense improving on nature in a clever way.
I have to mention a point against myself—which is that I do take general anaesthetics, which, while not an intelligence enhancer, is definitely an intelligence modifier for specific circumstances. However, turning brain function off is arguably simpler than trying to make it better.
It is possible, definitely, to improve human intelligence by combining it with a computer. So it’s not the case that I’m against the idea that it’s impossible to improve on the natural intelligence we all have—it obviously is.
What I’m pointing out is that all of these drug ideas are bound to be something that evolution has at some point tried out, and thrown away. And they are really unsophisticated ideas compared with those the brain has actually adopted.
Even the situation dependent argument isn’t as strong as you might think—for example your brain has a lot of adaptations to cover the “unpleasant and potentially traumatic things” situation, for example—and these adaptations generally disagree with your view that you shouldn’t remember them. It’s probably the case that intelligence tests are a novel environment, however....
Gwern has a good overview of this argument.
Well there could be many reasons why evolution has” thrown them out”. Maybe they are harmful in the long term, maybe their use consumes precious energy, or maybe they just aren’t “good enough” for evolution to have kept them. That is, maybe they just don’t give any signifigant evolutionary advantage.
Evolution doesn’t create perfect beings, it creates beings which are good enough to survive.
There can be harmul side-effects and that topic is not covered by the article; on the other hand, pure evolutionary argument can be doubted because of changed environment.
If I stimulate my brain, it is natural to assume my brain requires more energy now. So I probably need more glucose. In evolutionary relevant context, that would make me more likely to starve—after all, I would need more highly valued energy and thinking clearly wouldn’t make a killed bull magically appear before me.
This is still true for the most of the Earth’s population. It is not true for many of LessWrong readers, though. There are some primarily-mental jobs now (in some places of the world—the places where LessWrong readers come from). Keeping more things in you mind means being a better programmer, teacher, scientific researcher. Being better at your profession often helps you to evade starvation. And getting needed amount of calories—if you already know where to get all these vitamins and microelements—is trivial in these parts of the world.
So, this modification was not a benefit earlier, and it was quite costly; both factors are significantly reduced in some parts of modern world.
Of course, increased mental capability can lead to some personality traits that make it harder to reproduce; but that is again a question of side-effects and not a self-evident thing. If you consider it harmful, you can try to spend effort on fighting these side-effects—some people report significant success..
Maybe inclusive genetic fitness is not my utility function.
Same here. As a childfree human, maximizing the number of copies of my DNA is right up there with paperclip maximization on my list of priorities. :-)
One other categories of exceptions: We aren’t in the EEA anymore. In particular, we have much looser constraints on available calories than in the EEA, and that changes the optimal settings even for reproductive success.
There’s this really pretty large class of issues called “genetic disorders”, and a wide variety of other ways the body fails just fine without encountering foreign DNA/RNA… I’m assuming insulin for diabetics also has unexpected drawbacks and isn’t really in our best interests?
Or, put succinctly: “Scientists are so ignorant! If it was possible to cure cancer, why didn’t we just evolve to not have cancer in the first place?”
I have heard of genetic disorders, and know that they occur, and why they are found in the gene pool. And in such cases, it’s entirely appropriate to take medication for them. I take your point that perhaps I could have mentioned this class of DNA earlier, and that it is appropriate to take medication for that class of diseases, as it doesn’t improve your quality of life to leave it as it is.
I didn’t think we were discussing genetic diseases particularly. I am convinced of your argument that if you have a genetic disorder that affects your intelligence, you should take medication for it. I don’t see why this is relevant to the more general case of people who don’t have a particular genetic issue. Evolution is a good balancing algorithm—but it works by trimming the outliers. If you are unlucky enough to be one of the outliers, there’s likely something that needs correction. But generally, I don’t see the relevance.
As for your second point—well, we did. Cancer is very rare in the natural environment. It’s only in our much safer modern environment that we live long enough for it to become a problem again.
You may have missed my points...
Given the genetic variance in IQ, it’s obvious that most people don’t have optimal genetics for intelligence. Whether this is a “disorder” is an interesting semantic question, but the point remains that we know that the general class of “human brains” has a maximum that’s higher than where most people are at.
Equally, our brains evolved for a much different environment with much different trade-offs. Just like cancer wasn’t a threat in our ancestral environment, the ability to do a second order differential equation wasn’t a benefit.
In short, medical science suggests that, actually, there’s plenty of room to improve humans, both because we’re extremely inconsistently built, and because we’re not built to handle our current environment.
Actually I agree with all of this—there’s a tremendous difference between average intelligence and the top end of the bell curve, and we have no reason not to think it can’t go higher. We are the first species on this planet to attain general purpose intelligence, and there’s no good reason at all to think that either more isn’t possible, or indeed that the process of human evolution in this respect has stopped—quite the reverse I suspect. We have every reason to assume that at the moment human intelligence is evolving like mad—it’s being very strongly selected for in a very large gene pool.
But my point is that this all has nothing to do with the proposed methods of improving the brain. If we really knew how it worked, and were able to model the consequences of our actions better, then it would be less of a guessing game whether there was a longer term price to the short term gain.