Artificial intelligence has accustomed us to the success of its projects. Every day there is a new functionality that machines have learned, their processes are perfected and they are reaching increasingly surprising results. These undeniable advances contribute to the debate about the future of robots and there are those who argue whether they will develop consciousness or overcome humans. But the use of this technology still has many weaknesses who put their feet on the ground to the enthusiasts. An example of this are several projects of facial recognition that have failed, the last one in New York.
In September of last year, the Metropolitan Transportation Authority (MTA) of the US city began to test a face recognition technology that claimed to be able to identify drivers and passengers by collecting images through video cameras located in some bridges and tunnels. . The idea was to compare the images of the drivers taken through the windshields and compare them with a government database to be able to prosecute traffic offenders. Something similar to what China has been doing for a while.
Already then, critics of this practice arose who claimed the risk of the uses of that information. "These systems allow a continuous and real-time exploration of the identities of people and the tracking of their location," said Daniel Schwarz, a technologist at the Civil Liberties Union of New York (NYCLU). "It would really be harmful, let's continually know where people are moving."
But the project has been so inefficient that the citizens of New York do not have to worry. At least, for now. The newspaper The Wall Street Journal He has had access to an email sent by an MTA official to a senior official in the administration of the Governor of New York, Andrew Cuomo, on November 29 of last year. According to him WSJ, the MTA official wrote that "the initial trial period for facial recognition at the Robert F. Kennedy Bridge (connecting the Manhattan, Queens and Bronx neighborhoods) was completed and failed, with no faces detected (0%) within the acceptable parameters ".
Despite this resounding failure of the artificial intelligence they were using, the official also wrote that the MTA was acquiring additional cameras to expand the project to other locations.