Friday, 22 June 2018

Positive and negative feedback in evolution

Evolutionary theory has its own historical tradition and has developed its own terminology. However I sometime wonder what it would look like if it was (re)discovered by 21st century engineers. This post consists of some musings on that theme.

If evolution was (re)discovered by modern engineers, one concept that would probably be more extensively employed is "feedback". Feedback is the name for systems whose outputs are "fed back" into their inputs. Of course, this happens all the time in biological systems in various ways. The term "homeostasis" is often used in biology, and this is just a type of negative feedback. In particular, gene pools typically have their outputs fed back into them. From this perspective, favorable selection is a type of positive feedback acting on trait frequences (or gene frequencies), while unfavorable selection is a generally similar type of negative feedback.

Framing natural selection in terms of positive and negative feedback seems useful to me. The "feedback" terminology seems more general, and that is often a virtue in science.

Another piece of biological terminology that might be framed differently by engineers is "fitness". "Fitness" is an overloaded term, but "inclusive fitness" could plausibly be replaced by "utility" - the more general term from economics which refers to "that which is optimized". The common term "expected fitness" would become "expected utility" - another standard concept. One slight difference is that "fitness" is usually though of as being "relative", while "utility" is usually thought of as being absolute. It's a minor issue because there are usually many players and so the distinction doesn't make much difference. Anyway, IMO, the way to resolve this is to say that economics has got this wrong.

Saturday, 16 June 2018

Andres Gomez Emilsson's pure replicators

This post is about Andres Gomez Emilsson's proposed concept of "pure replicators". First, I'll let Andres introduce the concept:

I will define a pure replicator, in the context of agents and minds, to be an intelligence that is indifferent towards the valence of its conscious states and those of others. A pure replicator invests all of its energy and resources into surviving and reproducing, even at the cost of continuous suffering to themselves or others. Its main evolutionary advantage is that it does not need to spend any resources making the world a better place.
Conventionally, a "replicator" is something which copies are made of. Agents and minds aren't really replicators, they are large complicated things which can't easily be copied. "Reproducer" might be more appropriate terminology from this perspective.

Evolution optimizes for survival and reproduction. However, that does not mean that it builds creatures that are uniformly devoid of compassion. The evolutionary function of compassion may not be obvious or easy to explain, but it is likely to exist because compassion is widespread among humans. Probably, compassion promotes social cohesion and encourages acts of reciprocal altruism.

Andres warns against becoming a "pure replicator", but he defines this as an agent indifferent towards suffering, and most humans care act as though they care about the suffering of themselves or others - because compassion is built into them by evolution as a proximate goal. Becoming free of compassion does not seem as though it is a likely fate in the first place.

Andres apparently agrees, writing: "Most animals do indeed care a great deal about the valence of their own consciousness". He goes on to explain that "pure replicators" are mostly a future threat. There follows a bunch of speculation about how future intelligent machines might not use the pleasure-pain axis in their motivational systems.

I think that part of the problem here is a failure to properly distinguish between proximate and ultimate goals. A hypothetical agent with the sole goal of maximizing the number of their great grandchildren might still have proximate goals of minimizing the suffering of themselves and others. Many kinds of personal suffering are likely to be negatively correlated with the number of great grandchildren produced. Concern for others could well be adaptive too, though that's a bit harder to understand. That is how evolution can build compassionate creatures.

Andres apparently thinks that compassion is a useless spandrel. That seems tremendously unlikely to me.

However, my number problem is not with the science, it is with the terminology. You can't just hijack the "replicator" terminology and load it up with qualia for no good reason. The proposed termiology is simply ridiculous.

Friday, 15 June 2018

Carl Zimmer on heredity

Carl had a small but nice section on memes in his "Evolution" book. Carl's latest book on "Heredity" is out now. It is called "She Has Her Mother's Laugh: The Powers, Perversions, and Potential of Heredity". It has a section related to cultural evolution, called "The Teachable Ape". Memes get mentioned, but pretty dismissively.Carl cites Ehrlich and Feldman saying that: "The most recent attempts using a 'meme' approach appear to be a dead end".

Carl does cover attempts to expand the concept of heredity to culture, but he avoided citing most of the literature on the topic. Carl managed to convey that he knew something about cultural evolution, but it didn't seem as though he had very much understanding of the topic. I thought the chapter was quite disappointing. It seemed like a step backwards from his 2006 effort.

Wednesday, 13 June 2018

Counter-memes

"Meme, counter meme" is the title of a 1994 Wired article by Mike Godwin. Mike tells the story of how he combatted Nazi comparison memes on usenet by using counter-memes generated using memetic engineering. I don't know if "counter-meme" saw much use before 1994, but Mike Godwin either coined or popularized the phrase.

I think that "counter-memes" is a useful concept. Fighting bad memes with good ones is a common and obvious technique, and "counter-memes" seems like an appropriate name for the concept.

"Counter-memes" are part of the memetic immune system. Some are most effective before exposure to the memes they counter - those are like vaccines. Others are typically used after exposure - those are more like antibiotics.

I think that "counter-memes" is one of the bits of memetics that has the potential to go mainstream and enter the common vocabulary. It's fairly useful and fairly catchy. It could pretty easily happen that people start responding to posted memes with "counter-meme: <blah>". If explicitly saying "counter-meme" when responding to a meme with a corrective meme signals that you are up on the latest internet lingo, it could become quite common. I find this fantasy pleasing and would like to help make it a reality.

Sunday, 10 June 2018

Jordan Peterson new meme critique

Most modern meme critics frequently recycle the same content. Jordan Peterson seems to have come up with a new critique in a recent discussion with Sue Blackmore (starting 10 minutes in):

What do you think of the whoele meme theory? I think it's a shallow derivation of the idea of archetype and that Dawkins would do well to read some Jung. In fact if he thought farther and wasn't so blinded by his a-priori stance on religion, he would have found that the deeper explanation of meme is in fact archetype.
Jungian archetypes are innate, universal precursors to ideas. Memetics is related to the concept (since it deals with ideas), but archetypes aren't really a "deeper explanation of memes". "Meme" is mostly just catchy terminology for "socially-transmitted idea".

Modern scientists may not discuss Jungian archetypes very much - but there are certainly conceptual equivalents. One modern perspective involves the distinction between "evoked" and "transmitted" culture. Evoked culture is the product of innate Jungian archetypes interacting with environmental variation. Transmitted culture is not encoded in genes, it is instead, copied from others. There's a spectrum in between the two concepts. Some evolutionary psychologists are very interested in "evoked" culture and play down the significance of transmitted culture. The enthusiasts for "cultural attractors" are also sometimes involved in studying Jungian archetypes under another name.

Jungian archetypes are probably not mentioned very much due to association with the Jungian notion of a collective unconscious - a mystical notion which was subsequently widely rejected by scientists. Jungian psychology is about as out-of-date as Freud. I think if you tell modern scientists they should "read some Jung" you will generally get back some incredulous responses.

Anyway, Jordan Peterson's critique is apparently of a "straw memetics" that holds that ideas are 100% transmitted and 0% affected by our evolved psychology. No practitions actually believe this. In practice, our evolved psychology has always been on the table. There are plenty of ways of incorporating innate biases into cultural evolutionary models. They can affect the selective environment of memes, or they can influence recombination or mutation operators.

However, one of the central ideas of memetics (and cultural evolution in general) is that there's more to culture than innate predispositions (i.e. Jungian archetypes). Culture is not just about variations in the environment evoking different genetic responses (as in the "jukebox" model). There's also transmitted culture, and it is big and important, just as anthropologists have long been saying.

The sterilization of females

Memes can sterilize you, or dramatically reduce your fertility. This is true especially if you are a woman. The correlation between female education and fertility is strongly negative:

This isn't just a case of memes making people richer and r/K selection adaptively kicking in. The phenomenon is powerful enough to reduce fertility to sub-replacement levels - as seen in South Korea, Japan, and many other countries.

In memetics, the explanation for this is fairly obvious - diverting resources away from gene reproduction fuels meme reproduction. It is thus a common interest of many memes to sterilize their hosts. Dawkins gave this as an explanation of priest infertility in 1976 and the basic idea has held up.

If you consider memes as cultural symbionts, then this phenomenon may seem familiar. Many parasites also sterilize their hosts. Many human STDs do this, for example:

It is widely accepted that bacterial infections with Neisseria gonorrheae, Treponema pallidum and Chlamydia trachomatis can lead to fertility alterations. The impact on reproduction alteration is suggested but not well understood in the case of some viral STI, such as human herpes virus (HSV), adeno-associated virus, human immune-deficiency virus (HIV), human cytomegalovirus (HCMV) and human papillomavirus (HPV).
HPV is one of the better-known ones. It gives women cervical cancer and blocks up the tubes of their reproductive system with cancer cells.

Sybiont foodstuffs such as green leafy plants also frequently attempt to sterilize those that consume them. This prevents further consumption of the plant by offspring, and attempts to consume relatives. Phytoestrogens are a common method. These bind to the estrogen receptor and so defeminize many animals.

At this point, you may have noticed a theme. The targets of these attacks on the reproductive system are frequently female. It seems easier to sterilize females than it is to sterilize males. I notice that most human contraceptives also target women. If you look into forced human sterilization, the majority of victims there are women as well.

Why is it easier to sterilize females? I don't know the answer to that, but I have some ideas. Not all species have males. The female reproductive tract is mostly shared with those that reproduce asexually. Another factor might be that if population reduction is the aim, sterilizing males would not work unless you got almost all of the males. If you sterilize one male another can easily take his place. That is not so true of females. I'm not sure that these ideas explain all of my examples, though.