Scientists Create AI That Predicts Future Crimes With Eerie Accuracy
Imagine a world where crimes can be predicted long before they happen. A world where AI-powered algorithms can calculate where and when the next lawbreaking will happen. Recent research from the University of Chicago suggests that such a future may be closer than we think.
AI is indeed advancing at an incredible rate, permeating every aspect of our lives, from writing to planning travel. Even in dating, technology is starting to play a role. But what if AI is so advanced that it can predict crimes before they happen?
It sounds like science fiction, but reality is already approaching this scenario.
A study conducted by the University of Chicago found that the AI they created was able to predict crime rates with about 90% accuracy. The analysis used historical crime data from eight major US cities, including Los Angeles, Chicago, and Philadelphia. The predictions were made at the 1,000-square-foot level.
Project leader Ishanu Chattopadhyay compares the system’s operation to watching dark clouds predicting rain:
“We identify patterns in events and use them to calculate the likelihood of future occurrences in certain locations… only here the patterns are much more subtle and difficult to perceive.”
The main purpose of this development is not to identify specific criminals or crimes, but to analyze data on past incidents to identify high-risk areas. Chattopadhyay emphasizes that the system is not intended to prosecute individuals or punish them for crimes they have not yet committed.
The ethical issue is deeply controversial. Can a person be justly accused of a crime he or she may potentially commit? This dilemma is at the center of Steven Spielberg’s film Minority Report, based on a short story by Philip K. Dick, and remains a subject of debate.
However, researchers see potential in using AI to allocate police resources. Based on predictions, law enforcement agencies can better prepare for and prevent potential incidents. However, there are concerns about possible bias in the data.
A recent Reddit poll found that 60% of respondents would support implementing a similar system in their city if it was 100% accurate. Comments from users reflect mixed feelings about the project, with some expressing concerns about the intrusion on personal freedom, while others questioned the fairness of the system.
