Predicting the Ideal Future for Ourselves and for Others

Our predictions for the future depend on who we are thinking of: self or other.

Posted Sep 05, 2019

Photo by Patrick Heck
The Ideal City, Fra Carnevale (ca. 1484)
Source: Photo by Patrick Heck

Imagining the City of the Future in 1484

Baltimore’s Walters Art Museum houses a five-century old painting that helps explain some of the psychology underlying how we think about the future for ourselves and for others. The Ideal City (pictured left; larger image here), painted by Fra Carnevale (ca. 1484), reveals one artist’s take on what the city of the future might look like. Although the painting is technically beautiful, Carnevale’s prediction wasn't a very good one. Streets and architecture clash in style and scale, and the design and décor serve no apparent function. Maybe others wonder, as I did: where are all the people? Our idiosyncratic vision of the perfect or ideal future often gets it wrong, due in part to an overreliance on the assumption that what everyone's future will look like years from now is what my world looks like today.

Prediction and Projection

Five hundred years later, we continue to paint an image of the future that is rooted in our own experiences and preferences. In other words, people project. Carnevale may have adored coliseums and arches, assumed that one would look good next to the other, and believed that others would share his opinion. While useful in many cases (Dawes, 1989; Krueger, DiDonato, & Freestone, 2012), the tendency to project can give rise to an unavoidable tension between what people want for themselves and what they believe the group or society at large wants for itself. If people vary in their preferences and beliefs, and if people project these preferences onto others or into the future, then many predictions for the future must be wrong due to random variation alone. The psychological processes underlying how we think about ourselves and others (i.e., projection and egocentrism), combined with the uncertainty of the future, can therefore lead to mispredictions and seemingly incoherent beliefs about what is yet to come.

Predicting What Types of Jobs will be Automated in the Future

To illustrate this point, take Pew Research Center’s (2017) recent discovery that a majority of Americans believe that computers and robots will soon fill occupations traditionally held by humans. A majority of those surveyed held this opinion for occupations including fast food work, insurance claims, and software engineering, while fewer than half of participants felt this way for legal work, construction, education, and nursing. Still, 77% of Americans reported that it is “realistic” that computers and robots will someday fill many jobs traditionally worked by humans.

So where is the forecasting error? We won’t know how many of the Pew survey’s respondents are right or wrong until decades have passed. But one supplementary (and psychologically rich) finding from this survey reveals the motivating influence of the self in respondents’ predictions for the future. When these same survey respondents—who had just endorsed the belief that most jobs will someday be automated—were asked whether they believe that robots or computers will perform respondents' own jobs during their lifetime, only 30% agreed. People believe that automation is coming, they just don't believe that it is coming for them. It’s unclear whether this belief is the result of overconfidence, self-enhancement, self-serving bias, motivated reasoning, or any other of a horde of psychological phenomena, but what is clear is that many of these respondents will turn out to be wrong.

Predicting Who (or What) will Replace Humans in the Workforce

A second example of how thinking about the future is unavoidably enmeshed in thinking about the self was published last month in the journal Nature Human Behaviour. Granulo, Fuchs, and Puntoni (2019)* found that research participants tended to endorse the belief that a firm seeking to replace its employees should choose human replacements over robots. Assuming that these firms had already made the decision to cut costs by replacing employees, 67% of university students, 63% of MTurk’s “Master Workers” (who are highly qualified to complete online work), and 60% of typical MTurk workers preferred replacing these employees with humans over robots.

In a clever experimental twist, however, Granulo et al. asked half of their research participants to instead imagine that they were the employee slated to be replaced. Now, the majority of participants (60%, 60%, and 62% of the groups described above) chose robots as the preferred replacement. Just like in the Pew Research survey, participants’ preferences reversed when thinking about their own future instead of someone else’s future. In eight more probing studies, Granulo et al. go on to argue that people would prefer that their own jobs are replaced by automation because this results in less self-threat: that it doesn’t feel as bad to be replaced by a robot as it does to be replaced by a human (who is perhaps more skilled than you).

Conclusion

People often make sweeping, grandiose plans for the future, and yet, little is known about the psychology of futuristic concepts like automation, artificial intelligence, genetically modified foods, and large-scale policy change. Despite this lack of knowledge, however, we continue to make flashy, certain and confident predictions for an ideal future (that a universal basic income will solve poverty in the U.S., for example, or that building a border wall will stop people from entering the country illegally). Yet it is clear that people cannot divorce their own preferences and beliefs from their predictions for the future. Like Carnevale’s bizarre amalgamation of preferences into his image of The Ideal City, occasional incoherence between predictions for our own future and for others’ futures can keep a clear image from appearing before us.

* I commend the authors for making their data and materials freely and openly available to the public. Pew Research makes its data available as well.

References

Dawes, R. M. (1989). Statistical criteria for establishing a truly false consensus effect. Journal of Experimental Social Psychology, 25(1), 1-17.

Granulo, A., Fuchs, C., & Puntoni, S. (2019). Psychological reactions to human versus robotic job replacement. Nature Human Behaviour. doi: 10.1038/s41562-019-0670-y

Krueger, J. I., DiDonato, T. E., & Freestone, D. (2012). Social projection can solve social dilemmas. Psychological Inquiry, 23(1), 1-27.

Pew Research Center (2017). Automation in Everyday Life. Available from https://www.pewinternet.org/2017/10/04/automation-in-everyday-life/