Perils of Utopian Science for Societal Design
One of the earliest books that could be called utopian was Edward Bellamy’s Looking Backward. It described a future scientific society, devoid of competition, in which everything was run scientifically. It was all to be rational and orderly.
As a 16 year old, I was captivated by his vision. I later realized, with hindsight, that Mr. Bellamy probably never spent much time with actual scientists doing actual science.
Science is messy. It is an unending quest for knowledge that the scientists know can never be called certain in the way that religious people[1] crave. They are trying to take snapshots of a complex reality, discerning from careful tests what are generally small truths. Yet they know that whatever truths they may uncover could later be superseded.
(To understand HOW messy science can be, http://www.theatlantic.com/science/archive/2016/03/the-quest-to-make-synthetic-cells-shows-how-little-we-know-about-life/475053/ and http://fivethirtyeight.com/features/failure-is-moving-science-forward/)
In this lack of certainty, science differs from revealed knowledge. It also differs by virtue of being testable and repeatable. Scientists pursue their quest anyway because knowing—even if only partial and conditional—is far more satisfying to them than ignorance.
Even “settled” science such as the Big Bang Theory could still be superseded by more complete or deeper understandings. As described in my book, a highly respected physicist recently developed a theory that appears to explain the origins and behavior of the cosmos as well as does the Big Bang, but with a cyclic universe. Likewise, scientists discovered evidence that life may possibly only arise in dry conditions and not within water, as had long been assumed necessary.
Other such examples exist and will continue to be found. All of science consists of developing testable, falsifiable hypotheses. The experimenters, if good scientists, do everything they can to reduce the experiment to a single variable. But sometimes it’s the unexamined assumptions that may do them in.
This isn’t hypothetical. Dr. Bruce Alexander developed his famous Rat Park experiments after examining such an unspoken assumption. He observed that nearly all of the addiction research studies used Norwegian white rats. The rats eagerly consumed all manner of drugs, and displayed addictive behavior.
While others took this as evidence that the substances are themselves addictive, Alexander asked a different question. Given that these rats are sociable, intelligent, playful creatures, might their lab conditions—essentially, solitary confinement for life—constitute a kind of torture, from which the rats would desperately seek escape?
He and his team built Rat Park as a kind of utopia for rats. They ran the experiment for many years, with one key finding: they could not induce the rats to consume drugs, even when mixed with favored sugar water. Their conclusion was that the previous experimenters had failed to examine a crucial assumption.
Again, science is an evolving body of knowledge. That’s what makes it exciting to scientists, and to those of us who admire their work and revel in their discoveries. There is, however, a lot of popular confusion about science. For example, some hear scientists use the word “theory” and misunderstand:
“… the two words, theory and law, have very different common meanings. But in science, their meanings are very similar. A theory is an explanation which is backed by “a considerable body of evidence,” while a law is a set of regularities expressed in a “mathematical statement.” … A scientific law is not “better” or “more accurate” than a scientific theory. (http://evolutionfaq.com/faq/why-isnt-evolution-considered-law)
What has this to do with societal design? Mr. Bellamy, and more recent advocates of a scientific society such as Jacques Fresco (The Venus Project) share what I regard as a serious misunderstanding. They believe that, if only we were to scrupulously follow a scientific approach to societal design, everything would be fine.
When Mr. Fresco speaks of a proper system of governance for his ideal society, he proposes to turn over all important decisions to machines. The machines will make decisions that allocate everything rationally, he believes.(Said his associate, Roxanne Meadows, in a February 2016 Atlantic interview, “We would use scientific scales of performance for measurement and allocation of resources so that human biases are left out of the equation.”)
Once systems are in place, they’re hard to change and harder to replace. Even more seriously, the machines will need to have guiding values. If the values selected are interpreted by the AI in a way that’s inimical to human interests, the consequences may be severe. For example, if an AI with power to rule us decides to minimize pain, it may permanently immobilize all humans in order to protect us.
This is why a Celebrationist system of governance will be by people, but people who coexist within a very different kind of system than we have now. Celebrationism will encourage cooperation, mutual respect, evidence-based decision making, mutual service, and the cherishing of differences. Further, as I envision it, none of the people in government will have strong individual powers. Likewise, the AI’s will be partners who advise us.
I advocate a scientific, technological basis for society, but do so in order to create a neutral vessel within which all manner of beliefs and lifestyles, including faith-based practices and revealed knowledge, can flourish.
Finally, while some such as Mr. Fresco believe that a single model of a scientific society can be perfected, I have no such expectations, and true sustainability requires significant diversity. As I see it, those of us who come together to simulate a Celebration Society will not seek to make a single perfect model, but instead compete in teams to see who can devise the first viable and practical societal design—knowing that it still includes flaws, which will come to light over time. Then we’ll test that design in the real world, with each implementation being locally appropriate, and refine it further.
When others, inspired by our successes, want to build their own Celebration Societies, we’ll support them and allow them to share the “brand’ if they agree to certain principles. But beyond that, they may well take very different paths to expressing how their society(ies) serve those principles.
We’ll all learn from each other, because we’re all imperfect beings. So, too, will superhuman AI’s be imperfect, should such emerge. We will ever more closely approach truth and perfection, never achieving either in this relative world. But the journey will be delightful and awe-inspiring.
[1] I am using the word “religious” more broadly than most do