According to the researcher of Psychology and computer sciences from the Massachusetts Institute of Technology and Cardiff University robots which have artificial intelligence are learning biases from one another. Hence, now robots like humans and animals can also hold biases against one another.
Robot named, AI is observed to have a complicated relation with sexism and racial discrimination as it was previously observed in a Microsoft Chatbots a few years before. But the difference is this robot, AI do it all by itself. The researcher told us that a certain pack of free machines can hold prejudices by noticing, watching and learning this from one another.
The researcher made up a task in which each AI robot has to give or donate to anyone from his own family or out of the group, which must be according to the robot reputation and donation policies. But it was observed that they hold biases against those from another group.
By encouraging, robots learn from any one from their own group or from across their populace. According to the results of the study, robots copy the policies which give them benefits in a short term means which shows that they do not required high mental capabilities to hold prejudices.
Professor Roger Whitaker from the Cardiff’s university tells us that to hold a prejudices against each other is a natural phenomenon and through development it can also become incentivized in computer-generated population to damage widespread connectivity with others. By protection from these groups leads to form another prejudices groups which will make fractured populace. Hence, this wider prejudice is difficult to inverse.
For those who question that why they all cannot live together. There are some groups which have less prejudices level in them. Whitaker says that all the non-prejudice groups can make unions with each other without any mistreatment. This will lead to lead us to not think that prejudice taking hold. But this needs certain situations where robots are ready to interact outside their groups.
While giving testimony to Congress. Twitter CEO Jack Dorsey talked about the problem AI developers have in decreasing the accidental bias. He says that let’s just hope these robots don’t start disliking human faces before we fix this problem as they are becoming more and more free.