Modern Science Is Mechanistic, And That Is A Problem

By Gerrit Van Wyk.

Nature versus nurture.

The BBC published an article a few weeks ago about a laboratory in Iceland extracting DNA and then using artificial intelligence (AI) algorithms to match that to detailed information about the people’s lives. It seeks to connect genetics with our personalities and life choices, including relationships, and to possibly inherited diseases. The basic assumption is human bodies are machines, DNA is the code driving the machine, and one may connect the code with everything the machine does. That is the pure machine thinking dominating science and the Western world, and it is a problem.

Let’s for argument’s sake assume for one moment the world we live in is complex, as an alternative perspective, including humans that inhabit it, what would be the consequences?

To begin with, the human brain at birth is a forest of neurons, and there is evidence within the first few months and years of life, the body begins to prune the neurons and create connections between them. It doesn’t do so randomly, which neurons get pruned and how the remainder get connected depends on the environment within which you were born, and the critical part of that environment is your social environment. DNA creates the forest, social complexity the park that remains after it is cleared out. The problem with the BBC article is its focus is on the machine at the expense of the complex social world in which it operates and which significantly impacts on who we become and the life choices we will eventually make. What the algorithms connect to, mechanically, is not DNA, but the combination of DNA and the environment within which that DNA operates.

Secondly, we are born more or less sterile, but from the moment we pass through the birth canal, pick up bacteria, viruses, and fungi, not randomly, but from the people around us, and there is growing evidence those organisms shape us biologically.

We have more bacteria inside and on us than we have cells in our bodies, and many of those remain unknown. We live in symbiosis with them; our bodies feed them, and in return they influence our bodies. One such interaction is the gut microbiome, which we share with our families and probably pick up at birth from our mothers. There is a network of neurons in the gut, likely as large as the brain and in contact with it, which is shaped by the microbiome; it affects the brain and the brain affects it. The microbiome in turn is influenced by what we eat and exposed to, such as antibiotics that may profoundly change it. It turns out there is more to a gut feel than we think.

In the same way, our skins are colonized by contact with the people who hug and kiss us as babies, most skin bacteria happily live with us and we with them, but some can cause skin conditions which, one may say, we acquired socially. Most of these bacteria remain unknown.

Finally, there are multiple viruses living in our respiratory tracts without making us sick, and which may fulfill an, as yet, unknown function as a commensal.

The BBC study suggests genetics may tell us about our exercising habits and preferences, what hobbies we take up, whether we enjoy doing crossword puzzles, mathematical ability, etc. The problem with data trawling by AI algorithms is that, in itself, it is mechanistic. All it does is show up statistical correlations, which, as Russell Ackoff famously showed, can be as weird as a correlation between wearing silk stockings and lung cancer, without being able to explain the correlation. In doing so, it misses the point of the body developing within a complex physical and social environment which shapes it.

Why, then, this quest? The hope, based on a mechanistic perspective of reality, is once we can identify genes doing things, we can use AI for putting people into categories that may be helpful in steering them into the right careers, choosing athletes for sports, job selection, etc., and, the holy grail of gene manipulation is a way to predict and treat certain disease conditions. There are two dangers to this; the first is, this sounds uncomfortably like what, in the wrong hands, can turn into a form of eugenics by stealth, and there will always be wrong hands. And, secondly, it steers the mechanistic turn even further away from the true reality of life, that it is complex, by completely ignoring that complexity and the important role it plays.

There are two important points to what I’ve been writing. The first is the irony that the game I describe takes place in a complex social environment, which the scientists ignore as if it doesn’t exist, except it does, and it influences what they do. Mechanistic thinking rules in the world of science, and that is just a belief, not a factual representation of reality. Choosing to look at reality like that is a social act, not scientific. Secondly, the tools of science being used, DNA extraction, AI algorithms, etc., are artifacts and symbols of a mechanistic social world. And, thirdly, disseminating the results to journals, publications such as the BBC, etc., is a social act driven by social beliefs, needs, norms, etc.

The second is that health care, the most complex biological, social, and technological activity around is infected by the mechanistic turn which has several implications. Diagnoses are disease labels that don’t exist until created. Before psychology and psychiatry, people weren’t depressed, just sad a lot. Once the label was created, depression became a mechanistic entity for analysis by dissection to find a cure. That cure eludes us because via the mechanistic turn we ignore the social nature of the condition. Many years ago, a “new” antibiotic was released to great fanfare. It soon became apparent it was a drug searching for an indication, which scientists retrofitted so the company could turn a profit. That was a social not scientific event. These are two examples of the complexity of health care, yet we now have a generation of professionals believing illness and health are technical issues. A specialist colleague once said to me, to my surprise, an ultrasound report we had a disagreement on was accurate because it was generated by technology and technology is never wrong. This, despite the fact the report was generated by humans interacting with the technology, which makes it very fallible.

Some predict physicians will soon be partially or completely replaced by AI algorithms. A recent publication claimed chat-bots show more empathy than physicians. In other words, a computer is better at crying and grieving with us than another human being. We are fast reaching the stage where some will use computers to replace humans because they are cheaper, more efficient, generate more profit, are unemotional, and can put up with all abuse we can throw at them. The thinking supporting that, troubles me, because machines are not ethical or moral.

I’m not against science and scientific knowledge and a mechanistic perspective of reality. Both can be useful when used under the right circumstances, I’m against both for being exclusive, and not allowing any other perspectives to grow, such as our world is complex, and for being used mindlessly without considering where it comes from, and the moral, ethical, social, and other consequences of creating and using it.