I searched ANKI for "Cognitive Biases" but do not see a deck named "Cognitive Biases and Related Terms".

The two decks that are available are described so similarly that I can't figure out which one is supposed to be the improved version, or whether your deck even still exists.

Would you mind fixing this?

Wow. Okay. That's not what I expected, but it does sound like a plausible depiction of reality.

What are the common problems that GAI programmers run into?

I am also hoping that all parents in the future have the ability to make intentional genetic improvements to their children, and I also agree with you that this would not necessarily result in some horrible dystopia. It might actually result in more diversity because you wouldn't have to wait for a mutation in order to add something new. I wonder if anyone has considered that. I doubt that this would solve all the problems in one generation. Some people would be against genetic enhancement and we'd have to wait for their children to grow up and decide for themselves whether to enhance themselves or their offspring. Some sociopaths would probably see sociopath genes as beneficial and refuse to remove them from their offspring... which means we may have to wait multiple generations before those genes would disappear (or they may never completely vanish). We also have to consider that we'd be introducing this change into a population with x number of irresponsible people who may do things like give the child a certain eye color but fail to consider things like morality or intelligence. Then we will also have the opposite problem - some people will be responsible enough to want to change the child's intelligence, but may lack the wisdom to endow the child with an appropriate intelligence level. Jacking the kid's IQ up to 300 or would result in something along the lines of:

The parents become horrified when they realize that the child has surpassed them at age three. As the child begins providing them adult level guidance on how to live and observing that their suggestions are actually better than their parents could come up with, the child has a mental breakdown and identity crisis - because they are no longer a child but are stuck in a toddler's body, and because they no longer have a relationship with anyone that can realistically be considered to play the role of a parent.

If the parents are really unwise they'll continue to treat that person as a toddler, discourage them from doing independent thinking, and stifle all of their adult-like qualities until they're over 18 - because what they really wanted was to raise a baby, not a super-intelligent adult-like entity in a tiny body.

There must be many other enhancements that could backfire as well. An immoral parent trying to raise a moral child may also cause mutual horror and psychological issues (ex: the child turns in the parents for a crime and becomes an orphan).

I don't think it would be quite as efficient and clean as you're imagining, but I think the problem we'd run into (assuming everyone has access) would not be that we'd suddenly have too much conformity or that the elites would overpower everyone... but that people would do really stupid things due to not understanding the children they created and not having any clue what they were getting themselves into before hand. It could take multiple generations before we'd wake up and go "Ohhh! It needs to be illegal for parents to increase their child's IQ to three times their own!"

I agree with the spirit of "Screw this grossly inefficient and natural selection nonsense." but it's possible that even if genetic engineering can be made accessible to everyone, that people will simply refuse to legalize it for religious reasons or due to paranoia or that they'll have other irrational reasons... and it's possible that if it were legal and widely accessible, that humanity will do really unwise things with it and create big problems (especially if traits like responsibility are being lost). It's also possible that we simply won't perfect the technology anytime soon. It is pretty complicated to combine psychological traits in a functional way... give one kid LLI and they become a genius... do it with another kid and they become a schizophrenic. The scientists know it's complicated -- but who will they test it on? It's not ethical to test it on humans, but testing psychological trait engineering on mice wouldn't do... That's a real obstacle.

So there are many reasons I'm still interested in thinking about less efficient methods.

Wow. The article was fascinating. I devoured the whole thing. Thanks, Kaj. Do you know of additional information sources on the neurological changes?

far too short for strong evolution

This really depends on your concept of "strong evolution". If that is jargon meant to refer to a conglomeration of changes that makes the organism different all over, I would agree. If we're just talking about this in terms of "Is it possible that something of critical importance could significantly change in a few generations?" then I say "Yes, it is possible."

I assume you consider responsibility to be an important trait. Even if a change to the trait of responsibility alone may not qualify as "strong evolution" to you, would you say that it would be of critical importance to prevent humanity from losing the genes required for responsibility in even half it's population?

In a world where 40% of the people get here by accident, and we can tell that a lot of their parents failed to use their contraceptives consistently, are you unconcerned that there could be a relationship between irresponsible use of birth control and irresponsible genes being reproduced more rapidly than responsible genes?

The traits that I am aware of that show strong evolution all have had thousands of years

But today's situation is not the same. We have technologies now that could result in much more powerful unintended consequences just as it results in powerful intended ones. Birth control pills, for instance, didn't exist thousands of years ago. Our lives and environments are so different now (and are continuing to change rapidly) that we should not assume that our present and future selection pressures will match the potency of the selection pressures in the past. To do so would be to make an appeal to history.

I haven't found any evidence that allele frequencies have changed - I just started to look into this, and didn't even have a search term when I started. Due to that, I thought it was obvious that I didn't have anything on micro-evolution, so I gave you the evidence I do have which, even though does not do anything to support the idea that allele frequencies are being influenced, does support the idea that there's potential for a lot of influence.

Hmm. A contraceptive and unplanned pregnancy survey by 23andme would be so interesting... I wonder if they do things like that... If I get a useful response to my request for a credible source on their accuracy, I will investigate this. (I want to get their service anyway but am demanding a credible source first.)

Thanks, Maia, but my interest in this is from the perspective of an altruist who wants to know whether humanity will improve or disintegrate. I am interested in things that might create selection pressures that affect things like ethical behavior and competence. It seems like you've read about this subject so I'm wondering if you know of any research on micro-evolution affecting traits that are important to humanity having a good future.

Human microevolution, ooh. That sounds like a good guess. Google is showing me some results... it will take a while to parse them.

I would be very interested to hear evidence to the contrary, though.

Well the first thing that comes to mind is the incredibly horrible failure rate of common contraceptives, and the unplanned pregnancy rate and birth rate that goes with them.

Evidence:

In not even four years, about 25% of people using condoms became pregnant. Birth control pills were similar. http://www.jfponline.com/Pages.asp?AID=2603

"49% of pregnancies in the United States were unintended" http://www.cdc.gov/reproductivehealth/UnintendedPregnancy/index.htm

"These pregnancies result in 42 million induced abortions and 34 million unintended births" (world population growth was 78 million for contrast) http://www.arhp.org/publications-and-resources/contraception-journal/september-2008

If there's any trait at all that's connected with this - inability to afford more expensive methods, not caring about reliability enough to get an IUD or something more effective, dexterity level too low to correctly apply the product, impulse control issues / inability to think under pressure or when excited, forgetfulness, inability to resist temptation, etc. those traits are likely to reproduce faster than their counterparts. Considering that half our population growth is unintended, I'm pretty concerned about it.

The situation could be that (if a genetic irresponsibility trait exists and is responsible for a large portion of unintended pregnancies that go full term) even if the responsible portion of the population is larger, that the irresponsible portion begins it's generations sooner, and it's growth outstrips that of the responsible portion of the population, overpowering it in a short time.

We're also doing things like removing sociopaths out of the population and putting them into jails. This probably reduces the rate at which they reproduce, though I'd expect far slower evolution there, if any, than I would with something that influences contraceptive failure.

We select certain types of people (or they select themselves) for the military. When they go off to war, they're more likely to die before reproducing. Since Americans tend to send their soldiers away, they're also a lot less likely to reproduce before dying in a war than soldiers defending a home territory where they have access to lovers.

If welfare creates a perverse incentive to have more children, any trait that might make welfare appealing to a person could end up being reproduced.

People who get a 2 or 4 year degree have more free evenings in which to find a lover and take care of a child. Contrast that with people who get a higher level degree. They have to wait longer before they'll be ready.

People in certain industries work very long hours. They might not get a chance to meet someone or might decide they can't have kids working as many hours as they do.

For these last two groups, if they're determined to have kids, they'll probably find a way to do it -- but they may be significantly delayed compared with someone who gets a 4-year degree, works a 40 hour week and can start having kids when they're still in their early 20's. The delay of a few years probably wouldn't make much of a difference one or two generations away, but if there are any traits that result in one getting a higher level degree or working longer hours, those people probably won't reproduce as fast as others.

Can anyone tell me the name of this subject or direct me to information on it:

Basically, I'm wondering if anyone has studied recent human evolution - the influence of our own civilized lifestyle on human traits. For example: For birth control pills to be effective, you have to take one every day. Responsible people succeed at this. Irresponsible people may not. Therefore, if the types of contraceptives that one can forget to use are popular enough methods of birth control, the irresponsible people might outnumber responsible people in a very short period of time. (Currently about half the pregnancies in the USA are unintended, and probably 40% of those pregnancies go full term and result in a child being born. As you can imagine, it really wouldn't take very long for the people with genes that can cause irresponsibility to outnumber the others this way...)

Any search terms? Anyone know the name of this topic or recall book titles or other sources about it?