[ad_1]
When you read These words, there may be dozens of algorithms to predict you. Maybe an algorithm determines that you will come into contact with this article because it predicts that you will read it.Algorithmic prediction can determine whether you get loan or Work or apartment or insurance, And more.
These predictive analyses are conquering more and more areas of life. However, no one has asked your permission to make such predictions. No government agency supervises them. No one tells you the prophecies that determine your destiny. To make matters worse, searches for predictive ethics in the academic literature indicate that this is an underexplored area of knowledge. As a society, we have not considered the ethical implications of predicting people-these people should be injected with initiative and free will.
Overcoming difficulties is the core meaning of mankind. Our greatest heroes are those who are not afraid of hardships: Abraham Lincoln, Mahatma Gandhi, Mary Curie, Helen Keller, Rosa Parks, Nelson Mandela, etc. They all achieved great success beyond expectations. Every school teacher knows that children have achieved more than their cards. In addition to improving everyone’s baseline, we also hope that society will allow and encourage actions that are not afraid of difficulties. However, the more we use artificial intelligence to classify people, predict their future, and treat them accordingly, the more we can reduce human initiative, which in turn exposes us to unknown risks.
Humans have Forecasts have been used before the advent of Delphi’s Oracle. The war was launched based on these predictions. In recent decades, forecasts have been used to provide information for practices such as setting insurance premiums. These predictions are often about large numbers of people—for example, how many out of 100,000 people will crash. Some of these people will be more careful and fortunate than others, but under the assumption that risk sharing allows the higher costs of the less cautious and fortunate people to be offset by the relatively lower costs, the premiums are roughly the same (except for age Groups and other broad categories) expenses of careful and lucky. The larger the pool, the more predictable and stable the premium.
Today, predictions are mainly done through machine learning algorithms, which use statistical data to fill the gaps in the unknown. The text algorithm uses a huge language database to predict the most likely end of a string of words. The game algorithm uses data from past games to predict the best possible next move. Algorithms applied to human behavior use historical data to infer our future: what are we going to buy, whether we plan to change jobs, whether we will get sick, whether we will commit a crime or crash our car. In this model, insurance no longer concentrates the risks of a large number of people. On the contrary, predictions have become personalized, and you are increasingly paying in your own way based on your personal risk score—this raises a series of new ethical issues.
An important feature of predictions is that they do not describe reality. The prediction is about the future, not the present, and the future has not yet become a reality. Forecasting is a kind of guessing, which has built-in various subjective assessments and biases about risk and value. To be sure, the prediction is more or less accurate, but the relationship between probability and reality is much more fragile than some people have assumed, and it is also ethically problematic.
However, organizations today often try to treat forecasts as models of objective reality.Even though AI predictions are only probabilistic, they are often interpreted as deterministic in practice-partly because Humans are not good at understanding probability Part of the reason is that incentives around risk avoidance ultimately strengthened predictions. (For example, if someone is predicted to be a bad employee with a 75% probability, then when their candidate has a low risk score, the company will not be willing to risk hiring them).
[ad_2]
Source link
