Algorithms are panicking companies

The case is striking: Facebook sets up. A video recommendation algorithm at the end of the videos. But here, the algorithm confused black people. With monkeys by asking the video’s readers if they wanted to see “other videos on primates”. The video in question was from the Daily Mail, over a year old, which was titled ‘white man calls cops on black men at marina’. Of course, if this error caused controversy it is because it only showed people and not monkeys. We suspect well of the continuation with a fast reaction of the group which immediately presented its apologies by admitting that it was about an “unacceptable error” which had for origin the system of artificial intelligence according to the spokesperson of Facebook, requested by AFP.

The case of Facebook a striking example

In detail Facebook’s response was. We apologize to anyone who has seen these insulting recommendations. Facebook turned off the recommendation tool as soon as teams “are aware of what was happening in order to investigate the Lebanon Telegram Number Data causes of the issue and prevent it from happening again. If we can say that artificial intelligence is only in its infancy and should progress in the future, it must be noted that it is therefore still far from being developed. If it has often been controversial for moderation since it was mainly used to identify and block messages images upstream, it is clear that several underlying problems are still to be solved. These are not the only areas.

Telegram Number Data

The limits of artificial intelligence in question

Thus a robot from Singapore which was. Used to enforce social distancing experienced the same limits. He was walking around the city to remind people to respect it and made mistakes. Others use it for example when screening CVs. Artificial intelligence sorts CVs according to old applications which have given. Good results and pride themselves on Ew Leads objective sorting, but the question. May arise of the exclusion of certain profiles based on old data, for example. If artificial intelligence is controversial, it is first of all because it is increasingly present in many daily activities. It is thus used in autonomous cars, many connected objects, all intelligent assistants like Siri or even as support in the medical sector.

Leave a comment

Your email address will not be published. Required fields are marked *