Simulated News and Techno-Tribalism

We have reached a point in the early 21st Century at which media cannot be assumed to directly represent reality. This is both true in the general and older sense that we may distrust any narrative, but more importantly it is also now literally true, in a new sense. We now live in an age in which AI and CGI technologies allow images, video, audio, and even social network activity patterns to be rapidly fabricated to a high degree of realism, by anyone wishing to distort our perception and understanding of things. Such simulation technologies are only going to become more advanced and integrated into our societal institutions, so we must think hard (and fast) about how best to operate in this increasingly strange new world.

Unfortunately, we have already seen the emergence of negative reactions to the new technological possibilities and their social effects. There has been a predictable collapse of trust in news media, further complicated by state-sponsored disinformation campaigns, and the co-opting of terms like “fake news” (as a term of criticism) by those who themselves propagate disinformation. As a kind of secondary response, we are also seeing a return to tribalism, whereby people trust information not on the basis of any evidence of its veracity, but on the older (and less rational) model of ancient tribal societies, which is to say based on whether those you identify with have chosen to accept the information or not. Before the age of science and reason began, “truth” was largely a social function, and now in our hyper-modern age that seems to be the case again.

Some may wonder how humanity could have survived and evolved as an irrational species, often eschewing evidence in favour of social utility. The simple answer is that social cohesion has survival value, meaning that individuals tend to survive crises better as part of a group. By optimizing the human tendency to rely on community-groups in times of uncertainty, evolution has therefore baked a kind of “default rationality” into human behaviour, even if that behaviour involves abandoning explicit, short-term rationality for the shelter of the tribe. After all, evolution is a statistical phenomenon which “cares” about entire species rather than individuals, and which certainly has no concern for explicit logic, cultural values, or anything else we humans consider important. In short, evolution has “hardwired” us to trust groups over explicit logic, on the basis that – regardless of the nature of any given crisis – a group is more likely to survive it than is a lone individual.

That simple truth can be a hard pill to swallow for some “rationalists”. For a long time the field of economics was dominated by an idealised model of human decision making, in which humans are expected to always react rationally and optimally, according to objective self-interest. As empirical evidence gathered by cognitive psychologists increasingly showed that humans simply don’t work that way, scientifically-minded economists began to develop the fields of behavioural and neuro-economics, taking the realities of human cognition and behaviour into account. I worked in a related field myself (the psychology of Judgment and Decision Making), and at one conference I heard an opinion expressed which may well hold the key to negotiating this new era of technological uncertainty.

At that conference I attended a meeting of a society which had recently opened their doors after holding closed (invitation only) meetings for over forty years. Older luminaries of this group were complaining that despite decades of ground-breaking research into human biases and irrationalities, no-one was taking their advice on how to improve judgment, making it more rational, optimal, and evidence-based. An engineer visiting the group for the first time then offered an opinion: Don’t try to explicitly convince people to change how they make judgments and decisions, because it won’t work. We humans may wish to change, but evolution has designed us to do certain things in certain ways, and decision making is one of those things. Instead, the engineer suggested, we should develop technologies which embody optimal techniques, and then encourage humans to become dependent upon those technologies.

Such a solution perfectly fits our age of Big-Data-driven Artificial Intelligence. Machines can make more intelligent decisions than us, on our behalf, according to our own designs and rational intent. The trick is to create interfaces for those systems which encourage human trust, which in turn means encouraging the sense of tribal loyalty which evolution has instilled in us. In other words, in an age of big data and fake news, humans can either retreat into the tribalism of trusting ill-informed humans, or evolve toward a “higher tribalism” of trusting Artificial Intelligences.