Artificial Intelligence can help our day to day life. However giving computers the ability to sense, think and act gives rise to enormous risk to the human kind unless we raise ethical computers.

Good program vs bad program

A number of programs help us to do things - there are also bad programs like viruses. These however, reflect the ethics of the people who wrote the program and in general would be accountable for creators' own actions.

Google self driving car

What if Google^TM's self driving car is at fault in an accident. Yes, again, Google may be held responsible for the mistake. Risk may also be passed on to owners or backup operators.

Taking it to the next level

A truly self learning system takes this to the next level. You create the system and it learns on its own. After completing the training, it can do bad or good but the system is responsible for its actions. You are no more responsible than you are for your adult children's actions. That is why it becomes important to teach ethics as part of the training program.