With the advent of Artificial Intelligence, it makes me wonder, can such technology have a bias?
There have been some examples of when Artificial Intelligence has failed or even made incorrect decisions. It is due to the datasets that AI use to base their decisions, that are at the base of this problem.
One such example is how Siri or Alexa can not understand certain types of voices. Their recognition rate is actually lower with voices of individuals from color or other ethnic backgrounds. The vast majority of voices used in the dataset used in training Alexa is vastly make and white. This causes the AI system go actually have a bias based on the developers’ own biases.
We are at the advent of AI systems, but we must remain vigilant to ensure that AI systems developed have a complete dataset to learn from and base decisions.
I do believe that systems will get better with time.
The Future is Bright!