Extremism Bias
Extremism bias occurs when a foresight practitioner goes to extremes in their forecasts, scenarios, and thinking, either in positive or negative directions. There are many varieties of it, and we’ll just look at a few.
This is an unfortunately a rather common bias in the transhumanist community, of which I’ve been a member since 1999. Many transhumanists will take any particular technology or scientific advance and push it to the extremes in their visions, usually without recognizing all the social, ethical, economic, scientific, or technical constraints that will make their extreme future very unlikely to emerge. They do this because they have a strong and entirely natural desire to escape (transcend) the limits of their biology. But when they let this natural cognitive-emotional desire to suspend their critical faculties, they are often much more extreme in their expectations than reality will admit. Extreme in expectations of positive futures, and of negative ones alike.
As veteran forecaster J. Scott Armstrong says, a Golden Rule of Forecasting is that all forecasts should be more conservative than the forecaster typically expects. Both conservative in terms of the upside (optimism, protopias) and the downside (pessimism, dystopias).
As veteran futurist Paul Saffo reminds us, prediction analysis shows that in most domains, we futurists and foresighters tend to “mistake a clear view for a short distance ahead.” The late Roy Amara observed that for many big STEEPS changes and trends, we overestimate their effects in the short run and underestimate them in the long run. We see new things that are coming, but don’t recognize all the social factors that will actively slow down and oppose that change in the short run, including large business and political interests, moral conflict, and plain old inertia.
For example, Clay Christensen’s The Innovator’s Dilemma (2011) reminds us that big companies are counterinnovative. They try to get to the future first, patent it, sit on it, and sell us their old products or services until they are forced to introduce a new innovation by the marketplace, typically some small to midsized company they were not able to buy out, sue, or otherwise thwart. They only get serious about innovation (bringing new inventions to market) when another large competitor in their oligopoly is threatening them (rare, due to the anticompetitive nature of oligopoly) or when a small to midsized company is finally starting to gain some real market share.
Change is often bad for the shareholder, and conservativism usually wins in the boardroom of large companies and the legislature of large political institutions. The best organizational action is often to make sure there is no action. All of this slows many types of societal change. So too does wealth, and we’re becoming very wealthy societies today, with ever greater numbers of restrictions on our social, political, and environmental activities, the so-called Nanny State. Seeing all this inertia can help us become appropriately conservative in our forecasts.
All rules have exceptions of course. The exception to getting too extreme in our future claims is certain types of technological change, which are becoming more human-independent with each passing year. To adapt a maxim of Jim Dator, in the domain of accelerating science and technology, and almost exclusively in that domain, some of our longer-term claims about the future should “appear to be ridiculous.” In this domain only, we can agree with Dator’s maxim. In most every other, increasing conservatism is usually a more prudent forecast.
Let’s look now at several other common biases, several of which are closely related to extremism bias: hype bias (inflated expectations and positive emotions), drama bias (inflated fears and negative emotions), clear view bias (too-early expectations), and elitism bias (inflated expectations that the general public will identify with the views, policies, or timetables of the person making the forecast).
Each are commonly found in certain individuals and contexts, not only in transhumanists but in our wider foresight practice community.
