Hitting the Pounds: How Much This Insurance Watch Discount Could Really Cost You

For years, machine learning systems have outperformed their human counterparts in everything from Go to Jeopardy! drug discovery and cancer detection. With all the advancements being made in the field, it's not uncommon for people to be wary of the robots that will replace them in tomorrow's workforce. These concerns are misplaced, argues Gerd Gigerenzer in his new book How to Stay Smart in a Smart World, if for no other reason than uncertainty itself. AIs are phenomenally capable machines, but only if they have enough data to act. Introduce the wildly unstable precariousness of human nature into their algorithms and watch their predictive accuracy plummet - otherwise we'd never need to swipe left. In the excerpt below, Gigerenzer discusses the hidden privacy costs of sharing your vehicle's telematics with the insurance company.

MIT Press

Excerpt from How to Stay Smart in a Smart World by Gerd Gigerenzer. Published by MIT Press. Copyright © 2021 by Gerd Gigerenzer. All rights reserved.

If self-driving cars aren't happening, an alternative seems to be to train humans to use AI as a support system, but stay alert and in control when they fail, which is called augmented intelligence. This equates to partial automation, i.e. sophisticated versions of level 2 or 3. Yet augmented intelligence involves more than just adding useful features to your car and may well lead us into a future different, where AI is used to both support and monitor us. This possible future depends more on the insurance companies and the police than on the car manufacturers. Its seeds are in telematics.

Young drivers are reckless, overconfident and an insurance risk, according to the stereotype. Some indeed are, but many are not. Nevertheless, insurers often treat them as one group and charge a high premium. Telematics insurance can change that by offering better rates for safe drivers. The idea is to calculate the premium from a person's actual driving behavior rather than that of the average driver. For this, a black box that connects to the insurer is installed in the car (using a smartphone is possible and cheaper but less reliable). The black box records driver behavior and calculates a safety score. Figure 4.6 shows the rating system of one of the first telematics insurers. It observes four characteristics and assigns them different weights.

MIT Press

Rapid acceleration or hard braking are assigned the most weight, followed by driving beyond the speed limit. Each pilot starts with a monthly budget of 100 points for each of the four features. An “event” causes points to be subtracted, for example 20 points for the first rapid acceleration or for exceeding the speed limit. At the end of the month, the remaining points are weighted as shown and added up to get a total security score. Although telematics is often called black box insurance, the algorithm is not a black box at all like most love algorithms. It is explained in detail on the insurer's website, and everyone can understand and check the score obtained.

Personalized pricing is advertised as promoting fairness. To do this, they take into account the individual driving style. But they also create new sources of discrimination when driving at night and in town is sanctioned. Hospital staff, for example, may have little choice but to avoid working at night and in cities. So some features are under the driver's control, but not all. It's worth noting that one driver-controlled feature is missing from virtually all personalized fares: texting while driving.

And the black box that a...