Coded Bias documentary mainly talks about how the algorithm of artificial intelligence evaluates everyone’s value in the social system and its problems. In short, the algorithm will duplicate and even aggravate the long-standing prejudice in the world. For example, the court will refer to a so-called “recidivism rate” algorithm to grant parole, thus drawing the conclusion that the recidivism rate of colored people is higher; Some schools will use algorithms to evaluate the added value of teachers; Even many banks use algorithms to calculate customer credit. However, these algorithms come to the conclusion that women are more likely to be eliminated by AI HR and have worse financial credit than men. Next, I will discuss the summary of my views on this film.
At present, artificial intelligence is widely used in society in two modes: maintaining stability and making profits. Almost no artificial intelligence is an algorithm that exists for freedom and equality. Secondly, the arrival of artificial intelligence is accompanied by unevenness. By uneven, I don’t mean that the rich have technology before the poor, but that the more aggressive technology is often experimented with vulnerable groups first. For example, in new york, face recognition entrance guard was first installed in apartments with low-income colored people by rental companies, and it was made into a fine tool similar to traffic cameras, and these monitored images were used as evidence of violation of regulations for these colored people tenants. I think the meaning of algorithm design is automation and efficiency, but in many cases, humanity means refusing to obey and rejecting the value defined by the system. If we take automation as the end point of all problems, it will lead to very inhuman and even devastating results in many cases.
In a word, this documentary has taught me a lot. I think AI is only the application of algorithm technology and mathematics at present. Emotion, morality and other factors will not be mixed, and we can also think that it is intelligence without emotional temperature. For example, face recognition technology will adopt the same algorithm when facing different skin colors and races. If there is deviation and discrimination in these applications, it must be the cause of technology developers or the intention of technology users. In addition, I think it is unreasonable to judge a person only by coded Bias.
