Hi. I'm Tim Tyler and this is a video about optimization. It will mostly be about the significance of optimization.
Firstly, what is optimization? Optimization involves problem solving. Particularly it involves solving problems where you have to find a good solution from a space of possible alternatives. Since practically any problem can be framed in those terms, optimization is a very general concept.
A classical example of optimization is where you want to make a cylindrical tin can which encloses the most volume using the smallest quantity of metal. There's a ratio of diameter to length that is most cost effective which is independent of the size of the can. This is the optimal solution to the problem - and the process of finding it is known as "optimization".
Optimization problems are common. Whenever you try to get from one place to another as quickly as you can, you are solving an optimization problem. Trying to get proper nutrition while minimizing your calories is an optimization problem. Trying to acquire money with out breaking the law is another optimization problem.
Optimization techniques were discovered in the 19th century and were considered to be part of mathematics. The field rapidly split into discrete optimization, and continuous optimization - which often involved calculus. For some problems, the best possible solution available is needed while in other cases, you just want a solution that meets some criteria. Different approaches are often needed for these different classes of problem. All optimization problems feature a utility function that says how good solutions are. Some have an additional function that specifies when to stop the search. This can involve time limits, resource limits or a specification of what solution qualifies as being satisfactory.
Some optimization problems are linear - and special techniques were developed for solving those. However most optimization problems are not linear. Some optimization problems were best described in terms of constraints. The effects of adding and removing constraints has been studied.
Optimization can involve either maximization or minimization. It makes no difference mathematically.
The goals of agents can be represented using what are generally known as "utility functions". The agents then behave so that they maximize their utility function. This idea has been used extensively in economics. It has also been also used to formalize some ethical systems.
As the concept of utility maximization is very general, it provides a framework for modelling and comparing the goals of arbitrary computable agents.
Conflicting goals can be modelled using utility functions and this provides a useful framework for examining conflict. In particular the idea of "Pareto optimality" can be applied to conflicts. A solution to a conflict is "Pareto optimal" if no agent can get more of what they want without some other agent getting less of what they want. There's often a set of such solutions - known as the "Pareto set".
Some optimization problems have a temporal dimension, raising the issue of how long-term gains need to be balanced against short term ones. The solutions to some of these optimization problems change over time, and optimizing agents sometimes need to track optima whose location changes over time. Such problems require a dynamical system to track them - and then the stability of optima can become a significant factor. Oscillating or orbiting around optima becomes possible, and optima themselves may decline or completely collapse. The addition of a temporal dimension can result in a much harder optimization problem.
Though optimization techniques can be applied to specific problems, optimization is a general purpose skill that can be applied to a broad range of problems. Competence at optimization is closely related to intelligence - and the idea that optimization capability is a general purpose skill is related to the empirical observation that high intelligence results in increased competence across a broad spectrum of tasks - among humans.
It's often possible to describe properties of an ideal optimizer in a specified problem domain. For example, for the game tic-tac-toe, optimal strategies for both sides are known. In this case, if both players play optimally, the result is always a draw. In economics, this idea is known as economic efficiency. Deviations from maximum efficiency can arise as a result of stupidity or internal conflict. Conflicts can arise as a result of battles with peers or parasites, for example. Stupidity and internal competition are not always easy to distinguish from one another as causes of inefficiency. Internal competition looks a lot like stupidity from the perspective of an external observer.
We have a grand unified theory about how optimization techniques work. The basic idea is due to Charles Darwin and is widely known as Darwinian evolutionary theory. Optimization involves trial and error. It has the classic form of a Darwinian evolutionary process, where variations on existing solutions are generated and then tested - with the more successful solutions being retained and forming the parents of the next generation.
It is true that there are some optimization techniques do not closely resemble this model. For example, random search is an optimization technique - but it has no concept of memory or inheritance. Exhaustive search is another optimization technique that doesn't look very much like Darwinian evolution. It uses memory, but it only has a single simple lineage instead of a tree of variants. However, these techniques are trivial - and are only useful on tiny problems.
Techniques for solving more complex problems tend to look much more like Darwinian evolution. With more complex problems, trying solutions at random - without paying attention to what has already been tried - is not a very attractive option. You are forced to perform local searches. These more complex cases represent the vast majority of real-world problems and the process of solving them more closely resembles Darwinian evolution.
The Darwinism involved is of a very general kind. It permits the use of interpolation and extrapolation during recombination. In addition to retaining successful variants to show where to search, failed variants can be retained to show where not to search. It is a form of Darwinism which incorporates the principles of intelligent design. It more closely resembles Darwinism plus genetic engineering - or the kind of Darwinism that is involved in cultural evolution. Some think that the term "Darwinism" is a misnomer - although it is hard to deny that Darwin originally came up with the basic idea.
Paradigmatic optimization techniques include genetic and memetic algorithms - which are explicitly modelled on gene-based and meme-based evolution respectively.
Optimization techniques are very useful tools for solving problems. However, optimization has also turned out to have some interesting scientific applications. It is possible to model the behaviour of organisms using optimization models in which the organisms behaved as though they are maximizing the number of their distant descendants.
Ecosystems can also be modelled as maximizing a function - they behave as though they maximize entropy. Entropy maximization is a different idea from the second law of thermodynamics. The second law just says entropy tends to increase. Entropy maximization is a very different idea. It says that if there are a range of possible entropy increases, larger ones are more likely than smaller ones, on average. There is more than one reason why entropy is maximised, but the easiest one to understand involves the basic statistical fact that high entropy states are more numerous than low entropy ones - and so undirected changes are likely to lead away from low-entropy states.
The concept of entropy maximization turns out to have a broad domain of applicability. In addition to organisms and ecosystems, electrical discharges, drainage basins, propagating cracks and stars all behave as though they are maximizing entropy. It turns out that maximization is important in physics and chemistry - as well as in biology.
Lastly, the concept of 'optimization' is significant at the moment partly because it is a foundational concept for those interested in building intelligent machines. A synthetic intelligence would be a powerful optimization process. Such entities will probably be the most powerful optimization processes produced by evolution to date - as far as we know. Understanding how to build intelligent machines is essentially the same project as learning how to optimize effectively. It is a challenging project. We know that intelligent humans can fall prey to addictions and religions. We want to design an optimization process that isn't vulnerable to such things.
Although is started out as a relatively small area of mathematics, it has become clear that optimization is a subject of enormous technical and scientific importance with a correspondingly large social impact.
One old and persistent fallacy relating to differences between organic and cultural evolution involves the claim that order of exposure to ideas matters - whereas that isn't true for DNA genes.
My foil in this case will be Felix Aurioles, who recently wrote:
It is because of this ambiguity that we must distinguish between behaviors or cultural traits, and the pieces of information that form them. There remains a likeness with genetics; when a phenotype is successful it propagates itself along with the information that encodes it. The difference lies, in that while a certain combination of genes always codes for a particular phenotype; the forms of cultural traits depend not only of the “memes” that compose them, but on the order they were absorbed and on the particular social circumstances they entered the culture.
My reply is as follows:
Gene expression is often context dependent - and it is often untrue that "certain combination of genes always codes for a particular phenotype". Instead, genes and environment interact during development to produce a phenotype. This happens in both organic evolution and cultural evolution.
It is also untrue that exposure order is not relevant to DNA-genes. For example, exposure to Hepatitis D followed by exposure to Hepatitis B has quite different outcomes to the reverse order of exposure. That is because Hepatitis D is a satellite of
Hepatitis B - i.e. it requires its presence to reproduce.
In organic evolution, a pathogen-caused immune deficiency disease (e.g. AIDS) might leave a hole in your immune system through which another pathogen (e.g. tuberculosis or pneumonia) might find easy entry. Similarly in cultural evolution a "faith is good" meme might compromise your memetic immune system - and leave you vulnerable to a "the end of the world is nigh" meme.
To effectively compare organic evolution and cultural evolution, you have to have some understanding of how both processes work. Only understanding one of them is not enough. Many critics of memetics from the social sciences often seem to think that whatever smattering of evolutionary biology and genetics they have picked up is enough to allow them to venture forth public criticisms based on their knowledge of these topics - and quite often they are mistaken.
In the first three videos Al parodies popular songs. In First World Problems, he hitch-hikes on a well-known internet meme.
"Weird Al" gave a talk at Google recently - where he explains his approach and marketing strategy.
Critics of memetics sometimes claim that we don't have a predictive science of cultural fitness. Some folk may be ignorant about which cultural items spread, but some other people clearly can tell what will spread - and some have based their careers on it. This criticism is bunk.
In my 2011: year of the meme article I charted the beginnings of the internet meme explosion.
In 2014, it looks as though it might be happening again. Check out this graph of "memes" seaches:
That's a pretty spectacular 2014 "meme" spike! Looking at the meme news it looks as though it may have something to do with the memes that are hitch-hiking on the world cup. However, memes do seem to be on the rise again - and perhaps it will go on to rival the enormous 2011 meme explosion.
Hi. I'm Tim Tyler and this is a video about Universal Darwinism and the modern evolution revolution.
The term "Universal Darwinism" is the most popular name for the ongoing revolution concerning the expansion of Darwinian evolution's domain. Most scientists understand that Darwin's pioneering work applies to the evolution of animals, plants, fungi and microorganisms. However, in the twentieth century Darwinian evolution has also been applied to other areas including physics, chemistry, computer science and social science.
This expansion of Darwinism's domain of application is one of the most dramatic revolutions evolutionary theory has seen to date.
Not only do DNA-based creatures evolve in a Darwinian manner, but so do economies, lightning strikes, technologies, crystals, religions and propagating cracks. The basic elements of Darwinism - copying with variation and selection - are ubiquitous in nature - even outside evolutionary theory's traditional domain of biology.
Whenever you see a tree-shaped structure in nature, the chances are high that it's either a family tree or a bunch of family trees that have grown together. The same is true for structures that don't look like family trees - until you plot their history over time. For example a landslide doesn't closely resemble a family tree - but if you look at its history over time, almost all the rocks involved have one or more rocks which they inherited velocity from - whose influence caused them to gain speed and participate in the landslide.
The term "Darwinism" itself is somewhat vague. This can be confusing to some - which sometimes leads to controversy over what Darwin actually believed and what set of evolutionary processes deserves to be described as being "Darwinian" - as opposed to, say "Lamarckian". However, such discussions aren't central to the current revolution - the whole idea can be explained using specific core ideas from Darwinism instead - such as selection, adaptation and cumulative adaptive evolution. Science is about the ideas, more than it is about the people who came up with them - but there is a strong tradition of giving credit to originators and Darwin's name is used mainly to give credit where credit is due.
As a brief history, the term "Universal Darwinism" comes from Richard Dawkins in 1983 - who used it to mean something different. The term was given its current meaning by Henry Plotkin in 1984 - and the idea then popularized by Susan Blackmore, Daniel Dennett, David Hull, Geoffrey Hodgson, Gary Cziko and many others. Many of these folk were also pioneers of memetics and cultural evolution.
Another related term which refers to the same phenomenon is "generalized Darwinism". There's also a related concept called "universal selection" - which is best seen as being part of Universal Darwinism.
Universal Darwinism has faced criticism - mainly from social scientists and philosophers, who are uncomfortable with Darwinian expansionism for various reasons. One idea from critics is that unconstrained mutation can explain anything and so predicts nothing. However, Darwinism's traditional way around this is to constrain mutations to those that are naturalistically plausible and don't involve miracles. Another claim is that generalizing the idea of evolution so that it covers all change trivializes the idea of evolution. However, not many enthusiasts for Universal Darwinism actually endorse this idea. Macroscopic reversible systems such as a Newton's cradle have little to do with Darwinism. There are plenty of aspects of physics that aren't explained by conventional Darwinism. The easiest way to avoid this criticism is not to generalize Darwinism to cover all change in the first place. Most of the critics are wrong. In many cases, their motives are obvious - they are trying to defend traditional social science turf from being biologicized. This is generally a futile and counter-productive endeavour.
It is instructive to compare this current revolution with previous ones in evolutionary theory. It is smaller than the one that started in 1859 - but arguably bigger than most we have seen since then. The competitors include the revolutions associated with Mendel, Watson and Crick, and the discovery of the significance of symbiosis. Universal Darwinism seems to represent a more painful paradigm shift than any of the other revolutions that evolutionary theory has seen in the last a hundred and fifty years.
Another issue is how far through the revolution we are. Most energy expended on Darwinian evangelism seems to be devoted to those who don't understand Darwinism at all. In other words, it is devoted to the first wave of Darwinism. Later waves seem to receive relatively little attention. From this perspective, the Darwinian revolution looks as though it is less than half way through. However, hopefully science should go a bit faster when it is running on internet time. Technology improves connectivity, and links humanity together into a massive global brain. This tends towards a situation where once a few humans understand something, their knowledge becomes within easy reach of everyone - over the internet. Plus automation and outsourcing can help to take care of much of the more tedious work. So: probably the rest of the Darwinian revolution will take place in more of a rush than we have seen over the last 150 years. However there is obviously still a long way to go.
Being in the midst of an enormous scientific revolution makes for exciting times for those involved. Because of the lack of scientific manpower that has been devoted to the topic, low-hanging fruit abound - and it is relatively easy for budding scientists to make big discoveries and go down in history. Evolutionary theory has substantial social and political significance - so this is also a area of science where people can make a big difference with relatively little effort. Lastly, this is also an area of science that is being monetized. In particular, empirical discoveries related to applied memetics and memetic engineering are prone to being rapidly turned into cash by marketers and advertisers and progress in memetic algorithms looks set to play an important role in the development of machine intelligence.
These are exciting times for Darwinism. However this evolution revolution is currently not known about by many. If you have made it to the end of this video, you probably know more about this topic than most biologists do. However, the truth will out, and it now seems inevitable that Universal Darwinism will have its day.
MemeophobePeter Turchin took some steps towards publicly explaining his opposition to memes recently. What follows is my attempt to untangle Peter's meme muddle:
A cultural trait is similar to a meme, a word coined by Richard Dawkins, which is typically explained as “an idea, behavior, or style that spreads from person to person within a culture.” Dawkins proposed that memes are cultural equivalents of genes—self-replicating units of cultural transmission. Cultural traits, however, are a more general category than memes, because they also include quantitative traits that cannot be easily represented as discrete alternatives (for example, an inclination to trust strangers, which I will discuss in tomorrow’s blog).
Traits are different from genes, just as cultural traits should be considered to be different from memes. However, practically nobody argues that traits are more general than genes – and then concludes genes aren’t useful. That argument would be nonsensical: genes are heritable information, traits are the results of their expression. Similarly, cultural traits and memes are best treated as different names for different things: memes are the heritable cultural information, and cultural traits are the result of their expression.
A closer synonym for meme within academia is “cultural variant”. However the term “meme” is better, shorter and massively more popular. It won, long, long ago.
Why speak of "memes" rather than just discussing "cultural traits"? Because these are different things! Traits are on the phenotype side, while memes are in the genotype side. Or - using more words - memes are heritable cultural information while cultural traits are their expression in some environmental context.
No one argues against genes on the grounds that traits are more general than genes. Well, apart from Charles Goodnight, maybe. The reason is that doing so would be ridiculous. So: why argue against memes on the grounds that cultural traits are more general than memes?
I think the answer is that because often in science the more general theory is to be preferred - if it is equal in other respects. However, when comparing trait-oriented theories with gene-oriented theories, there can be no pretense of equality in other areas. Traits are the product of developmental processes and environmental interactions. Those are enormous complexities. One of the useful things that was discovered during the gene revolution was all the mileage that it is still possible to get if you strip away all that complexity and just focus on the heritable information - i.e. the genes.
As a brief summary, genes (i.e. heritable traits) are what persists in evolution, while other aspects of the phenotype are lost in each generation, don't persist, have less influence and can often be usefully ignored.
One of the big ideas of memetics, is that we can use the exact same approach in cultural evolution - and get most of the same benefits. That's not to say that ontogeny isn't interesting, or that students of cultural evolution can totally ignore developmental processes - just that memes (i.e. bits of heritable cultural information) are a tremendously useful and helpful abstraction in understanding the cultural evolutionary process.
Just as genes are a tremendously useful and helpful abstraction in understanding the organic evolutionary process.
I think that Peter Turchin is just confused about memes. He hasn't found a sympathetic interpretation of them - and so doesn't know what he is talking about. However, why go on in public about something you don't properly understand? In my experience, one sometimes-legitimate reason is to give others the opportunity to straighten you out. Time will tell if this process helps in Peter's case.
[Note: if anyone reads my quoted comment above and thinks that "cultural variant" and "cultural trait" sound like the same thing, please look up how the term "cultural variant" is used by its proponents in academia. It is actually used by them to refer to heritable information - and not to the corresponding traits.]
PLEASE PLEASE avoid “memes.” The meme is not only unscientific, but anti-scientific–Dawkins’ slap in the face to the thousands of anthropologists, linguists and psychologists who have worked on these issues for 200 years and come up with very good understandings of how culture is transmitted. Memes not only do not exist, nothing like them exists, and the word is really seriously distracting attention from the actual data and theory on this.
Were memes really a "slap in the face" to anthropologists? I think one has to agree that they were. Let me explain:
In 1976, anthropology had resisted Darwinism successfully for over 100 years. Cultural anthropology had transformed itself from a science into one of the "humanities". Theories were seen as playing the role of: preconceived notions that might bias your perceptions and distort your observations of other cultures. The focus was on recording the data, without preconceptions. Where had this data-driven approach got them? Anthropologists had simply not "come up with very good understandings of how culture is transmitted". They had persistently failed to grasp one of the key organizing principles of their own discipline: namely Darwinian evolution. Basically their theoretical enterprise had failed to find this important organizing principle. At that stage, a slap in the face from a scientist was entirely appropriate medicine. WAKE UP! SMELL THE DARWINISM!
Some might point to Donald Campbell as an anthropologist who understood Darwinism before 1976. That's fair enough - but Campbell was one person, who came along late in the day, and helped started the process of cleaning up the mess left by all the other anthropologists. In no way was he representative of anthropology, or cultural anthropology in that era.
Indeed, to a large extent, this situation persists to this day. Understanding of evolutionary theory among anthropologists - and among social scientists in general - it still awful. Never mind that it's an important part of the theoretical foundation of these topics. These days a few of these folk are busy educating themselves about the topic. However, for goodness sake, don't try to defend the multiple generations of anthropologists who kept their subject mired in pre-Darwinian darkness for so long. Those folk didn't know the extent of their own ignorance - or the damage they were causing.
Yes, Dawkins gave a "slap in the face" to anthropologists. What else are you supposed to do with a bunch of scientists that persistently refuse to understand the basics of Darwinism? Also: the slap wasn't enough of a wake-up call. Most anthropologists persisted in their pre-Darwinian dream for multiple more decades. For example, look at the contributions of anthropologists to Darwinizing Culture. They are simply embarrassing. The pre-Darwinian era in anthropology should be a source of embarrassment and shame to all those involved. Those were stupid, poorly-educated humans. Now we know better. Or at least, we should do.
Memes and culture traits: There is an absolutely enormous amount of work on this. Nothing remotely like a “meme” exists. Cultural knowledge is not packaged in neat little clumps, does not spread like genes or bugs or viruses from person to person, and does not have a life or identity of its own.
Gene is obviously totally confused about memes. However, I've dutifully added a link to his comment on my meme-denialism page.
Hi. I'm Tim Tyler and this is a video about memetic algorithms.
While social scientists have mostly ignored or misunderstood memetics, the topic has received much warmer welcome from computer scientists. Where social scientists lamented their complex, integrated cultures being atomized into isolated memes, the computer scientists shrugged and went on to use the idea to solve their engineering problems.
Computer scientists had previously employed genetic algorithms as a means of finding solutions to their problems. These involve virtual creatures who are rewarded for producing better solutions to whatever optimization problem their creators are trying to solve. The virtual creatures are iteratively modified using mutation operators, and are maintained in a breeding pool - allowing for recombination to take place.
Cultural evolution appears to have dramatically increased the rate of development of human civilization. If culture evolves faster than DNA does, perhaps we can improve the power of our computer optimization strategies by drawing of some aspects of cultural evolution. This led to the idea of memetic algorithms.
Memetic algorithms are similar to genetic algorithms - but in addition to mutation, recombination and selection, they make use of individual learning, social learning and teaching. Where genetic algorithms were inspired by organic evolution, memetic algorithms draw inspiration from cultural evolution.
Evolution took billions of years to produce creatures sophisticated enough to produce an open-ended type of cultural evolution based on behavioural imitation. This appears to have been because observing the actions of another and then recreating them is not a trivial task. Detailed imitation of behaviour is technically quite a difficult reverse-engineering task, which requires complex cognition to perform.
However, fortunately, we do not have to wait until we have human-level machine intelligence before machines can make use of social learning techniques. We can engineer virtual environments in which even the relatively dumb artificial agents that we can build today can enjoy the benefits of cultural transmission.
Rather than passing on cultural information by a clumsy process based on behavioural imitation, memetic algorithms can work with creatures which have been designed for direct thought transfer. Agents can can record their sense data and replay them for the benefit of other agents. Behaviours can be encoded in a portable format and ported directly between agents. Memories can be directly downloaded from one agent and uploaded into another. Cultural transmission is pretty easy for machines - it doesn't require advanced intelligence.
Memetic algorithms may include a genetic component - and thus may exhibit meme-gene coevolution. Not DNA genes - of course - but rather the same kinds of genes that are used with genetic algorithms. In cases where both memes and genes are involved, there is a tradeoff between the resources expended on each type of evolution.
So far memetic algorithms have become quite a popular technique, with many books and papers on the topic and a string of international conferences dealing with the subject.
Memetic algorithms lie directly on the path which leads towards machine intelligence. Cultural evolution in humans illustrates the process machines will need to master to continue their growth. Machines are masters of mind-to-mind thought transfer, which makes it easy for them to engage in social learning. If you have a partly-intelligent machine, one obvious thing you can do to boost its power is to network it together with other similar machines - and let them cooperate with each other. The dense environment provided inside modern data centers provides an ideal environment for machines to engage in social activity with each other.
Overall, it seems highly likely that memetic algorithms will play a key role in the development of machine intelligence. Memetic algorithms will thus probably prove to be the most important application of memetics - in the long term.
Hi. I'm Tim Tyler, and this is a video about the origin of eusociality - and the role that symbiology might play in that.
Most modern ideas about of the origin of eusociality are fairly simple. They invoke kin selection. The ecological advantages of forming large cooperative groups are obvious - and those groups that have managed it dominate the planet - but it isn't easy for evolution to get individual creatures to cooperate in large groups. Kin selection helps to explains how such cooperative groups can be stable. Almost all cases of close group living involve close relatives. The conventional wisdom is that kin selection did it.
Some have invoked group selection as an alternative idea about the origin of eusociality. However, this explanation has turned out to be a dud - in the sense that it doesn't really add anything to the kin selection explanation - because modern forms of group selection and kin selection have turned out to make equivalent predictions.
There are other forces that promote cooperation and group living. Byproduct mutualism is the idea that individuals enjoy straightforward fitness benefits to living in groups - for example, by sharing defense, foraging and lookout duties with others.
This video is about another explanation how eusociality can be stable - one involving symbiosis. The idea is that living in close groups promotes the spread of symbionts between group members. In turn, the symbionts manipulate their hosts into close living arrangements in order to allow the symbionts to reproduce. These two effects create a synergetic spiral that leads towards ultrasociality and eusociality.
This is not a new explanation for eusociality. Indeed it was the most widely accepted explanation for the origin of eusociality in termites before kin selection came on the scene. In this context, it is known as "the symbiont hypothesis". Termites depend on interactions with other colony members to obtain gut bacteria which are essential of their survival. It was thought that this regular need for bacteria from other colony members forced the termites into living a social life.
Multiple symbonts may be involved in creating pro-social forces. Both mutualists and parasites may be present. Each additional symbiont increases the ecological forces that pull the hosts together.
I think that the symbiont hypothesis is an unfairly neglected explanation for the origin of eusociality. It is interesting to me partly because of its relevance to human culture. There, it suggests that memes manipulate humans into coming into close contact - in order to facilitate their transmission between humans. In turn, humans living in close proximity to each other makes them more vulnerable to horizontal transmission of pathogens and symbionts - including more memes. This leads to a spiral of increased sociality - and to more and more memetic transmission.
Eusocial species are well known for their mutualisms. Termites have their cultivated fungi. Mole rats have their gut bacteria. Bees have their flowering plants while ants have their domatia. The idea of social groups being drawn together by a web of ecological interactions between multiple symbionts makes a lot of sense.
The symbiont explanation differs significantly from the conventional kin selection one. Under kin selection the genes of organisms involved benefit from the cooperation. Under the symbiology explanation, benefit to the host genes is not necessary - the hosts could be being manipulated into engaging in social behaviour - for the benefit of mutualist and parasitic symbionts. You could still classify this as kin selection - kin selection between the symbionts involved and their own offspring, but that's a bit different from the orthodox idea that kin selection between host genes is responsible.
While the conventional kin selection explanation is obviously an important factor, I think that more attention should be given to the symbiont hypothesis of the evolution of eusociality. We need to quantify the effect to see how important it is - relative to kin selection at the level of host genes. The effect seems to be especially clearly applicable to humans - where the cooperating humans involved are often not themselves closely related.
Enjoy,
References
Cleveland, L.R., Hall, S.R., Sanders, E.P. & Collier, J. (1934) The wood-feeding roach Cryptocercus, its protozoa and the symbiosis between protozoa and roach. Memoirs of the American Academy of Arts and Sciences, Series 2 17, 185-342.
Hi. I'm Tim Tyler and this is a video about memetic engineering.
The term "memetic engineering" refers to the deliberative creation of memes using techniques from engineering.
The term is derived from "genetic engineering" - which involves the engineering of genes and genomes.
In the modern world, many memes and memeplexes are engineered - rather than being the product of the deliberate breeding or unconscious selection of memes.
Engineered memes are used in marketing, advertising, entertainment, politics, warfare, religion and education - among other fields.
To give an example, Gangnam Style is a memetically engineered music video. It has over 2 billion views - a testimony to the success of its creators.
Techniques for memetic engineering broadly mirror those used in genetic engineering. Transplanting and recombining existing memes is a common source of new variants. Intelligent design is used to create memes to specifications. Deliberate selection is used to filter the results of these techniques.
With DNA-based creatures, classifying organisms into those that have been engineered and those that have not is a relatively straightforward task. However, with memes, it isn't always easy to distinguish between engineering and selective breeding. The problem is that with DNA, only engineered genes pass through the human mind. By contrast, practically all memes pass through the human mind - and so there is much more scope for changes that might be classified as being engineering to take place.
Like most powerful technologies, memetic engineering is a positive force which also has significant negative potential. Social engineers could use memetic engineering to create a benevolent utopia. However, today, memetically engineered pathogens currently cause a significant quantity of damage. In particular the obesity epidemic, addictive drugs, pornography, movies and computer games represent widespread memetically-engineered plagues.
In some areas, indoctrination targeted at children uses memetically-engineered propaganda to turn kids into soldiers. In other cases, memetic engineering is used to recruit new cult members. Memes can have a dark side - and engineered memes are not excluded from this.
While genetically engineered pathogens are rare and cause relatively little damage, memetically engineered pathogens are common, often largely unregulated, and cause damage in a massive scale. While there are some attempts to restrict more addictive drugs, nicotine, alcohol and caffeine are widely available, many millions are addicted. These are cultural products whose spread is promoted by advertising campaigns which are engineered by large corporations.
Engineered products are sometimes regarded as being inferior to ones that are produced more naturally. This phenomenon does not just involve Luddites: many people often prefer wood to plastic and cotton to polyester. Most of those who would choose ice cream over a banana, recognise that the more natural product is more likely to be better for them.
Broadly in line with this widespread suspicion of artificial products, genetic engineering has acquired a dubious reputation. People campaign against the production and sale of genetically engineered goods. Engineered foods have been dubbed 'frankenfoods' by their detractors. With memetic engineering things are a bit different. In practice, few object to the use of engineered memes simply on the grounds that they are engineered - since so many memes are engineered. However some of the drawbacks associated with engineered products do also apply to memes too.
For example, Esperanto is a memetically engineered language. It was designed - rather than evolving over a long period of time. It has a very regular grammar and has been designed to be easy to learn. However it is widely recognized that Esperanto sucks. Part of the reason is that it didn't evolve over thousands of years - and so is poorly adapted to the human mind. This situation with memetic engineering thus mirrors some of the objections that are associated with genetic engineering.
Despite these kinds of problem, memetic engineering seems to have a bright future. Engineered memes are ubiquitous and dominate the planet. In addition to undirected mutations and selection, they can take advantage of the full toolkit of intelligent design - including interpolation, extrapolation, and evaluation under simulation. As a result they tend to evolve faster and adapt more adeptly.
Despite a three billion year head start, DNA genes are already lagging behind memes in some areas. A more eager embrace of engineering has contributed to the success of memes in the ecosystem. Engineering DNA is challenging - whereas for memes, the human mind is their native environment, and so engineering comes to them more naturally. Unless DNA genes similarly embrace engineering techniques, they risk being rapidly left behind in the dust.
Those
tracking the inheritance of traits have tended historically to
consider what proportion of the variance in a trait is explained by
"genetic" factors - with the rest being explained by "the environment" -
and chance fluctuations.
The figure differs according to what trait is being investigated,
but an average figure of around 50% is often a reasonable guess
for the proportion of the variance in a trait is explained by
human DNA genes.
The concept of "heritability" is used to describe this.
However, a basic question in cultural evolution is what proportion of
the variance in traits is explained by cultural inheritance. Alas, this
question seems to have rarely been addressed.
Ideally, we would like a breakdown for a wide range of traits: human DNA genes, DNA genes of
parasites and symbionts, cultural transmission, other environmental factors and chance.
Part of the problem is the very idea of "heritability" itself. The concept shoehorns the
factors involved in development into 'genes' and 'the environment'. However, in
the context of cultural evolution, this dichotomy is basically a naive and unhelpful
classification scheme. It's the kind of classification that someone totally ignorant
of cultural and environmental inheritance might come up with.
It would be better to split trait variances into inherited and non-inherited components -
and then sub-classify the inherited components according to how the inheritance takes place -
via host genes, symbiont genes, cultural transmission or
environmental inheritance.
It's especially important to know what factors can influence a trait if you want to
influence that trait. We do know some things about this topic - but mostly
the basic science in this area appears to have been neglected.
One of the reasons why I want to know is because I want to quantify how significant cultural inheritance
is - compared to DNA genes and to other forms of environmental inheritance.
One old and persistent fallacy relating to differences between organic and cultural evolution involves the claim that order of exposure to ideas matters - whereas that isn't true for DNA genes.
Take a quick lesson in 
Those
tracking the inheritance of traits have tended historically to
consider what proportion of the variance in a trait is explained by
"genetic" factors - with the rest being explained by "the environment" -
and chance fluctuations.
Posts
