Question

Artificial intelligence (AI) and data science are hailed as society’s answer to human bias in decision-making. Why not let the computer algorithms take over, make the decision-making process free of human interaction, and supposedly, free from human bias. Isn’t this rational decision-making? What could go wrong?

• Apple new credit cards (issued by Goldman Sachs) reveal bias against women (lower limits to women, than men)
• 2016 investigation by ProPublica revealed that data-driven AI used by judges to decide whether a convicted criminal will commit future crimes is biased against minorities and people of color.
• Face recognition software make more mistakes in identifying women and people of color
• Automated financial underwriting reveals gender and ethnic biases.

Why is this happening?

What is decision making going to look like in the future?

Answer 1

We should take artificial intelligence as prediction tool. Because it can't understand the full thinking process of human mind and situation. For example a consumer wants to buy something and AI starts its work on his decision making then what if he cancelled his plan due to Money problem.

Computer algorithms are already working on it with thieving our data by pre installed applications and softwares in mobiles and computers. But they don't work well. Suppose that you are searching about world's most expensive jet, then what can AI predict about you that your plan is buying it, highjack it or only search for information.

Human makes decision randomly. He is not an animal whose work is only finding food and then poop. What If a criminal wants to be gentle man and AI predict his next crime, what if a women can fly the president's plane better but AI predict a man can be better pilot.

So human mind is beyond AI's prediction and if we will follow it then it will be harmful for human society.