People are more likely to follow the advice of an algorithm

People's opinions are given less attention than machine advice

When it comes to personal advice, many people now trust algorithmic calculations more than the opinion of a friend or acquaintance. Even when judging a person's attractiveness, machine calculations are more likely to be believed.

This is the result of a recent study by researchers at Harvard Business School .

Machine security

In the study, test subjects were asked to make judgments about different facts: For example, “How popular is this song?” and “How much does this person weigh?” They were then offered help in reaching a decision and then given the opportunity to re-examine their decision. The additional advice was basically always the same, but some subjects were told that it came from either a human or an algorithm.

If the test subjects judged a person's weight and were asked to take another opinion into account, their estimate was 45 percent closer to the estimate of the advice if they were told that it came from a computer. Study participants only changed their estimate by 30 percent if they believed the advice came from a human.

Experts work differently

The researchers originally assumed that people would value further human advice more than an algorithm. But when asked about a person's attractiveness, the participants were also more likely to pay attention to the computer system's opinion. However, the results differed in one aspect.

“Experienced professionals who regularly make forecasts rely less on algorithmic advice than non-experts - but this affects their accuracy. The results shed light on the important question of when people rely on algorithmic advice rather than listening to advice from humans. This has implications for the use of 'big data' and algorithmic advice,"

said the scientists.

To the paper “ Algorithm Appreciation: People Prefer Algorithmic To Human Judgment ”