The Era of Blind Faith in Big Data Must End
In the film, In the Age of AI: Promise and Perils of AI, Cathy O’Neill explains that the use of AI fueled by big data should not be dependent upon too much. She states that the main issue with AI is that algorithms are not accurate or fair since they blindly repeat people’s patterns. The other problems include lack of integrity and accuracy. No single algorithm is perfect; therefore companies, organizations or medical professionals should rely entirely on the results or outcomes of AI. They need to consider how often algorithm errors occur and for whom the model may fail. In addition, the cost of the failures must be considered as well as the long term implications of the faulty algorithms. The kind of biases that could result from the use of AI include a racially segregated society where the police are only send to minority residences to look for crime. This would lead to racially biased policing and criminal justice system.
The film also illuminates the issue of facial recognition. China has embraced AI surveillance technology to govern around 1.4 billion citizens. The country is estimated to have around 300 million surveillance cameras already installed. Likewise, the police department spends around $30 billion on the AI-based technology. Thus, the tool is serving as a new way from managing the economy and the society. On the other hand, the use of facial recognition in the US has faced a serious backlash from public protests. For instance, Amazon’s Rekognition facial-recognition program was recently withdrawn from the market following shareholders and employees’ protests about potential abuse. In essence, the main issue with facial recognition is “surveillance capitalism,” which refers to the commoditization of personal data.