by Mike Meyer
While evolution has always been presented as a force of nature acting on life forms without their consent, that is no longer the case for us.
Evolution is, of course, not completely our choice. In purely biological terms evolution is a reaction to a changed environment with that change causing selection of different characteristics for survival and reproduction. Sentients, however, are a different situation.
Much of that difference is the presumption of our species that we are in control of evolution. That is a slight improvement over the assumption, held for most of the modern and scientific era, that evolution had stopped for us. This was very much the result of our self perception as the cause of evolution in other species while we would remain the same.
That bit of hubris has come to a sudden and obvious end. The range of genetic changes on other life forms and now on humans with CRISPR/CAS tools has completely disrupted our perceptions of engineered evolution. This combined with the nature of our environment and a growing range of factors needed to ensure our species survival are driving us to change. How fast that change will come may be, to an extent, under our control. If it is we need to get with it.
Complexity is the point that is punctuating our evolutionary equilibrium. Interestingly enough we are a cause at several levels of this problem and also of the potential routes for evolutionary success.
Stated as simply as possible we and our planet are the victims of our own success. Unfortunately that success was very much spontaneous and not well planned even by our older standards. We’ve trashed our planet and our atmosphere while devastating other life forms many of which are closely tied to our survival.
The result of this unplanned success is a very complex, nonlinear set of problems complicated by nearly eight billion human beings. This condition now has a fast timeline that is moving us to a four degree, or more, increase in average temperature by 2100. That will devastate human society and the planet and may well bring human quality of life advancement to a halt.
For an excellent look at what that means check The Uninhabitable Earth by David Wallace-Wells. The devastation is building and we are at only one degree of warming. At this stage all that we can do is slow the rate of increase. It is already far too late to stop it.
Unfortunately rather than working at reducing planetary temperature we are increasing it with the active connivance of very stupid people and great many more who are inactive but willfully ignorant. The climate problem, itself, is at the ragged edge of our ability to manage the data needed to understand it. As a result we have only a range of projected outcomes that are very locally dependent.
We understand this firmly enough now to know what needs to be done but don’t yet have the ability to get enough of our population to the level of understanding necessary to implement planetary change. Of course we have never accomplished a planetary level change in our history, at least not consciously, so that is another major problem. We are in profoundly new territory for human civilization.
As you can see this does not bode well for humanity or many of the other life forms who have to live on the same planet with us. The problems that we face with a ticking clock are both unique in our history and very difficult for us to grasp let alone accomplish a solution.
Not only do we have a range of complex problems, of which climate change is the most deadly, but we have complicated evolution by both genetic engineering, CRISPR again, and the creation of non biological intelligence (AI/ML). This combination is what presents us with the chance or, more accurately, need to choose evolution.
Of course the choice to evolve using completely new and electronically augmented techniques has already been made. We wouldn’t be as well off as we are if we hadn’t already made that choice. Most people are just not aware that is has been decided.
The shock comes when they suddenly begin to understand something about genetic engineering. Most people, particularly in America and certainly in very poor nations, have no where near the educational background to understand this. The fact that this is a criminal level failure in American education while a sad but understandable situation in poorer countries, makes no difference in the end.
The result of this shortcoming is the tragic and completely inappropriate reaction of people with bad information. We see it in the use of NON GMO on food packages designed to reinforce ignorance, instilling fear while making money, and the even more tragic anti-vax movement.
Genetic modification of food is the reason for the survival of Homo sapiens at this population level and that has been achieved for millennia by selective breeding. The anti-vax thing is the result of cold blooded manipulation of people for personal and political gain. In both cases the reasons given are completely false with no scientific validity and force pediatricians countless hours of stress attempting to explain reality to parents for the safety of everyone’s children.
While we understand human generated climate change and have for about thirty years, it took twenty of those years to reach scientific agreement on the danger points. This produced a conservative estimation of how soon we would hit warming that would significantly affect us. Unfortunately that was quickly discovered to be too conservative and we are faced with a maximum of thirty years to achieve changes to prevent escalating to over four degrees of warming.
The problem is, again, that we are in over our heads. We do not have the native ability to either understand the levels of complexity that we are facing but we also have difficulty understanding the ML based data projections that we are getting. The first is a problem of quantity of information and the later is an issue of quality.
We are solving the first by off loading massive information processing to our digital systems. This is now the process of Machine Learning as part of Artificial Intelligence. Obviously this is not any form of sentience or intelligence as we know it as humans but the first steps in that direction. This is, now, the new from of evolution as we struggle to augment our limited biological systems. This is the new evolutionary path whether people realize it or not.
That quality issue is a psychological need that we have based on the existing (previous) paradigmatic structure that defined the modern world. The peak of our capabilities are scientific as they have been for the last three hundred plus years. That paradigm posits a ‘dead’ physical world with only humans as fully aware and with logical laws of nature as the means of understanding reality.
This is much more difficult to change but that is what must happen. It is how we define our understanding and verification of a solution. After millennia of effort we developed both logic and proof of logic based on theory, or laws in the more parochial form, that are the standards against which we determine validity with empirical tests.
Our initial ML systems do not work that way. They do not start from an hypothesis that is valid under existing theory and then process data but go directly to vast quantities of data from which they find patterns in fine detail. This is not a denial of the scientific method but a post-human elaboration on it
The result is much more direct in the case of A versus B to achieve C. In five millions examples B is more successful or more efficient. But B may also be more biased against women or People of Color. That may result from the algorithm used to select the examples. We are learning quickly how easy it is for people to screw up digital expertise with human biases.
But that is another problem. The problem here is that there is no way that people can confirm against logic the ML results gained. To understand the results in a human sense requires that you simultaneously compare a million unique data points. We have a hard time keeping five data points in our minds so, by necessity, we must simplify the data. An hypothesis must be clear, logical and as simple as possible to allow testing with a process that we can understand. But that is now our biggest challenge to confirm ML based results. We can’t handle the needed data points to follow the ML process.
So what does this mean for our next stage of evolution? The road that we have selected for evolution means the merger of electronic and biological. This is commonly identified as cyborg and, interestingly, is commonly now used to frighten children in school. That is an indicator, I think ironically, of the growing awareness of how our evolution is now accelerating. Change is always frightening and totally new change is the most frightening.
While we are beginning to move into a wider range of implanted technology this is still seen, almost exclusively, as replacement for physical organs or lost biological functionality. That has already begun to fade into augmentation rather than replacement. But it will be only a few more years before we become generally augmented. That will be much faster than people now think and that is precisely due to the need for quick human evolution.
Until such time as we have massive digital processing and storage implanted in our bodies or linked to by internal sensory means we will need to learn to trust our external thought and memory systems. In effect we will have to learn to rely on our senses of humanly fair and valid results as we work to build checks into algorithms that are used to run ML based analysis.
This is will be an early form of internalization of the process and will need to be good enough to allow us to have confidence in it. That is the big change but it is happening. As always with change most people will deny it, denounce it, and then do it.
This change is the recognition that we are unable to handle what we have created. This can be improved by strong education but, given America as an example, that will not overcome our limitations on critical analysis of the amounts of information that we must personally process. We must recognize that this requires evolutionary change and the biological process is limited by our very abilities to ameliorate challenges in our environments.
There are strong research results that question previous concepts of instinctive language structures and problem solving skills in humans. It appears that we learn language and social action by imitation but that imitation is not an inherent human skill.
Perhaps this understanding will make it easier for us to accept our digital tools as internalized gadgets that allow us to think better. In the end this will require a kind of faith in our digital internalized tools that process information in ways that we cannot now.
We are not there yet. The question is if we can hurry our new style of evolution quickly enough to avoid killing ourselves with ignorant mistakes.