Speedy NEWS

AI ethics are a reflection of human ethics

I often hear the term “AI ethics.”

What follows are some typical problems caused by AI malfunctions.

– Frequent troubles with AI home appliances
Many problems occur close to the home. Such problems include AI speakers placing orders without permission, AI vacuum cleaners messing up and ruining the furniture, or smart appliances leaking privacy information.

– Trouble caused by opacity regarding how AI trained with deep learning make decisions for things like loan screening or stock ordering
The AI algorithm that set the credit limit for Apple Card credit cards (issued by Goldman Sachs) favored men over women. At times, the stock market chaos can ensue from misplaced orders.

– Pedestrian fatality caused by Uber’s AI-equipped self-driving car
In the suburbs of Phoenix, Arizona, a self-driving vehicle equipped with AI failed to realize that an object was human and collided with it.

– AI drones can be used as weapons of mass destruction
An AI may unable to detect the difference between a person holding a gun intending to hunt and a person holding a gun intending to enter combat. AI could potentially misclassify a target and harm citizens unrelated to terrorist activities.

The biggest problem with AI is that you often don’t know why a certain decision was made.
When it comes to AI, the greatest challenge for humans may be the ability to think of “unexpected scenarios.”

Instead of making inferences from available data, humans must take “that which does not exist” to generate “that which is possible.”
To develop the ability to predict things for which we have no data, I think it is important that we interact with nature, which can train human experience and intuition.

◆ Reference
The AI will destroy all humans! Why “Sophia” is attracting attention

https://ainow.ai/2019/10/01/176315/

Movie “West World”on U-NEXT
https://video.unext.jp/play/SID0036898/ED00209438