Consciousness is evolution’s solution to dealing with big data
No one (in the rational school of thought) really knows how life on earth came to be. Personally I quite like the panspermia hypothesis, that life came to Earth from space. In fact I really love this crazy idea that long ago a distant dying alien civilisation encoded their history into the DNA of a bacteria and sent rockets out in all directions. One such rocket landed on the newly forming earth, the bacteria colonised our young planet, and eventually evolved into us. Now when we take DMT we can access the historical archives of our alien predecessors encoded in our ‘junk’ DNA. Some amazingly creative imagination there.
But of course what I like doesn't make a difference to what really happened. Hopefully in the next 10–20 years we’ll have more definitive answers on the origins of life. Already potential clues that life may have existed on Mars is very exciting. And if not life, at least we know its building blocks are out there. Amino acids have been found on meteorites, and even more complex organic molecules have been observed in space. Many think it’s very possible that at least proto-life evolved in space billions of years before earth even existed.
Either way, life came to be, either on Earth, or somewhere out in the universe. There have been many hypotheses as to how this happened, and perhaps it’s not fully up to date (first published in 1976, but republished in 2006) but I quite like Richard Dawkins’ explanation of life’s precursor in The Selfish Gene.
(N.B. I’d also like to point out a much more recent hypothesis, Jeremy England’s “Dissipation-Driven Adaptive Organisation”. Apparently life is the most efficient way of dissipating energy, thus carrying out 2nd law of thermodynamics more optimally. To quote England “You start with a random clump of atoms, and if you shine light on it for long enough, it should not be so surprising that you get a plant”. Read about it here, watch a lecture here, and I wrote about it here)
Proto-life and Evolution
(This is my summary and words, not Dawkins’. Please refer to the book for a more detailed explanation) Imagine a molecule, that isn't alive, but is able to replicate itself. A Replicator. That may sound crazy, but actually it’s quite feasible. Imagine a long chain molecule that is made up of smaller building-block molecules (E.g. building-block molecules A, B, C, D and our long chain molecule is A-D-D-C-D-A-B-B… etc. N.B. this is not DNA. Yet). Each building block has an affinity for its own kind (A attracts A, B attracts B, etc). If you have one of these long chain molecules floating around in a body of water filled with the very same building-blocks (A, B, C, D) also floating around loosely, it’s not too difficult to imagine that each building-block in the chain molecule will attract the loose building-blocks, and line them up next to themselves, naturally in the same order as in the original chain molecule. Given the required conditions and energy, these loose building blocks will bond with each other and hey presto we have a second chain molecule with building blocks in the same order as the first one. I.e. they’re identical. Now we have two Replicator molecules floating around in the water.
Or it could be that the building-blocks have an affinity not for their own kind, but to an opposite or negative kind. So building-block A attracts D, B attracts C. Thus the building-blocks in the Replicator attract and pull together a complete chain molecule consisting of the negative of itself — which could be thought of as a template. This negative (or template) then attracts the negative of itself which of course becomes a replica of the original molecule. (This is in fact conceptually similar to how DNA replication works).
For the sake of this argument it doesn't matter if this replication is positive-positive or positive-negative. What matters is that these Replicators are not alive (at least they are not alive by our current definitions, which are also constantly shifting). They are just molecular chains, and it’s very possible that such a completely non-conscious, non-living, automatic molecular replication process is possible in a soup containing the basic building blocks. It’s also quite simple to see that this replication is exponential (1 Replicator becomes 2, 2 becomes 4, 8, 16, 32, 64, 128, 256 etc.). So resources — free floating building blocks — could quickly become scarce. I.e. these non-living Replicators would be competing for the same resources.
This replication process is not going to be perfect. It’s almost certain that during the replication there will be accidental changes to one of the molecule chains. E.g. the 512th molecule chain might be a little bit different to the other 511 chains. In this competitive environment, if this accidental change enables better future replication for whatever reason (e.g. easier, quicker, stronger, more efficient, more robust etc.) then it’s more likely to replicate. I.e. more of the newly replicated molecule chains will contain the new changes, the older molecules will replicate less in proportion (because the resources are limited). So the new beneficial changes will propagate through the population exponentially.
Evolution by natural selection has already begun before the Replicators are even alive. Through this natural process the changes accumulate incrementally and the Replicators evolve into more efficient Replicators. Millions or perhaps even billions of years later (depends whether the Replicators came from space or not), given the right conditions, newly evolved Replicators are capable of basic properties of life. Abiogenesis is not sudden, but gradual. Slowly chemistry becomes biology. It’s quite possible that early life wasn’t based on exactly our current DNA or RNA molecules, but something else, similar, a different Replicator molecule, which eventually evolved into RNA and DNA.
As billions of years and ‘generations’ pass, these Replicators evolve into proto-cells, prokaryotes, eukaryotes and eventually multi-cellular organisms. (N.B. Evolution from uni-cellular to multi-cellular life has been observed in just 2 months in lab experiments under selective pressure favouring multi-cellularity. This demonstrates nature’s ability to organise itself into multi-cellular cooperating and interacting structures without any external help or guidance, all that the researchers did is to create an environment which favoured multi-cellularity, i.e. an environment which is likely to exist in nature).
By the time we have life, the organisms are responding to their environment. Even single celled microbes have the ability to detect and respond to chemicals in their surroundings, light, each other; and wave their little tails, or hairs to tumble and move around. The race to live and replicate has already begun and every organism is evolving its own Evolutionary Survival Strategy. Part of every organism’s evolutionary strategy is to extract some kind of information from the environment, and respond to it in some way.
Models of the environment
Jump forward a few more billion generations and we have highly complex organisms with nervous systems. Some of these organisms — as part of their evolutionary strategy to extract information from and respond to their environment— are able to build internal models of their environment. So they can build an understanding, learn, make predictions and respond appropriately.
Furthermore, when these organisms build models of their environment, it is an evolutionary strategy for some of them to also build internal models of other complex organisms in their environment.
Brief introduction to Consciousness
No one (in the rational school of thought) really knows what consciousness is.
The hard question of consciousness, as David Chalmers puts it, is how and why do we even have a consciousness at all? (N.B. ‘hard’ should not imply ‘impossible’). Why do we feel things, instead of just registering inputs -> processing information -> controlling outputs like ‘mindless’ robots, or zombies? How or why does this sensation come about?
While many people reject even the notion of this question, for the ‘How?’, personally I lie somewhere in-between the John Searle and Daniel Dennett camps. I’m not 100% convinced it must be a computational process (as Dennett seems so sure about). I’m also not convinced it must be a biochemical process, like digestion or photosynthesis (as Searle seems so sure about). At this point they both seem equally plausible to me, and to vigorously commit to only one hypothesis feels like a leap of faith (though perhaps a rational leap of faith, and that’s necessary to make progress).
There is of course Panpsychism. From the heart of ancient eastern philosophies, and as beautiful and poetic as it may sound, I find it hard to believe consciousness is a fundamental property of the universe. Though I really would love to be proven wrong on this. Christof Koch et al (initially in collaboration with Francis Crick, discoverer of DNA and Nobel Laureate, until he passed away in 2004) are making interesting claims that consciousness is an integrated property of all information processing networks. It sounds crazy, but I can’t help but keep a curious eye on their progress.
I particularly like Roger Penrose’s hypothesis of quantum gravity based consciousness. It also sounds equally crazy, but then most radical theories often are to begin with. Even though specific hypotheses are being shot down by the wider scientific community, it’s more Penrose’s approach to quantum gravity (or rather, his criticism of the lack of any widely accepted approach, and the Copenhagen interpretation) that I admire. And other equally crazy theories regarding quantum biology are starting to gain interest.
But again whether I like a hypothesis or not makes no difference to whether it’s true or not. Hopefully in a few years these hypotheses will either be falsified or will gain stronger support. Right now both physicists and neuroscientists alike are ridiculing these hypotheses as pseudo-science, but in the face of radical ideas that is usually the safest strategy as Randall illustrates so well.
But for the sake of this argument it doesn’t matter if consciousness is computational or biological (or even quantum). Like most rational people, I believe it’s physical and natural, i.e. governed by the laws of nature, and it’s not some kind of ‘supernatural substance’ (reasons for that I’ll post separately).
The Evolution of Consciousness
For the ‘Why?’ question of consciousness, I propose:
In complex organisms — as part of their Evolutionary Survival Strategy to extract information from and respond to their environment — it is a successful strategy to build internal models of their environment. This may include modelling other complex organisms, and the most efficient way to do that is to develop a sense of self.
As simple organisms evolved, they started developing more complex sensorimotor systems. Acquiring complex senses correlates with more complex behaviours, to more optimally react to their environment, evade predators, and find food or mates.
Especially when vision evolved around 540 million years ago and as some believe — led to the Cambrian explosion. Perhaps these organisms started needing a higher level of ‘intelligence’ to manage the higher dimensional data streaming in, so they could make better use of the limited bandwidth in their neural pathways, to make more optimal decisions to feed to the relevant parts of their body, and to take optimal actions. This involves being very efficient with sensory input, deciding which data to process, and elevate to higher levels of cognition, and which sensory input data to suppress, and effectively ignore. In higher organisms, this may even include learning to model the environment, so as to make more accurate predictions, and thus be more efficient in processing sensory input data.
In higher organisms, this may even include learning to model the environment, so as to make more accurate predictions, and thus be more efficient in processing sensory input data.
Going even further, to be able to form any kind of social interaction, some complex organisms may need to learn to model each other. The complex organism is indeed very complex. Trillions of cells operating together, complicated multi-cellular interactions, proteins, enzymes, chemical reactions etc. To model the entire system’s behaviour would be immensely complicated. It would require too much data to ingest and too much processing power to simulate. But this becomes more feasible if I can model and interact with you, not as billions of atoms vibrating in a quantum field, not as a huge lump of organic cells moving through space; but as a thinking, feeling individual, whereby your consciousness, is an abstracted high level entity with goals and desires that I can empathise with.
I believe that the self exists as a simplified model of the organism’s overall goals, related thoughts, desires, feelings, actions and plans towards those goals. The conscious mind becomes an abstraction layer, and acts as a simple interface, through which we model and interact with other complex organisms.
In fact this modelling of others’ mind is what we call empathy. Turns out us humans even have a whole dedicated brain region to model other people’s minds, the right temporo-parietal junction. Perhaps this is why we tend to anthropomorphise everything? Not only animals, and plants, but even non-living entities. Perhaps it is how we have evolved to efficiently model the behaviour of and interact with others?
Thus perhaps consciousness is evolution’s solution to dealing with big data.
N.B. Nobel Laureate Gerald Edelman was (☹ he passed away very recently) a champion of consciousness through Darwinian evolution, though from a slightly different (even opposite, but somehow complimentary) angle. Through Neural darwinism, he demonstrated the neuro-biological processes that could allow Darwinian evolution to take place in the brain, and how it could apply to and select neural groups to eventually focus attention. However, he specifically stated that this selection wasn’t necessarily for consciousness per se, but that consciousness (including secondary consciousness) was an emergent property of the complex, hundreds of trillions of connections in the brain’s neural network. You can think of it as he tried to explain the biological how of consciousness evolution as opposed to the why. Whereas I’m coming from the opposite angle, assuming that evolution did select for consciousness, trying to explain why.
N.B. Interesting that now one of the biggest motivations behind the massive push and investment into Artificial Intelligence and Machine Learning is the fact that we’re drowning in data. As we evolve culturally and technologically, we’re generating and extracting so much data from our environment that we don’t know what to do with it. We need ‘intelligent’ machines and algorithms to parse the data and decide what is relevant, what isn't, and what should be done with it. It’s almost as if we need simpler models of the data, and machines that can both create these models, and make sense of them.