How to address job losses? In an interview for the strategy-business.com website, Ng argues for a universal basic income for people, who, however, should become lifelong learners.
Making the world a better place
Tremendous progress is being made thanks to deep learning. Today, programs can predict, for example, whether an e-mail is spam. In the future really efficient image and speech recognition will be possible. Robots able to follow spoken instructions will be developed.
Self-driving cars
By developing deep learning for self-driving vehicles, travel will be much safer. Average life expectancy will increase and the time we spend travelling will be shorter. Mass production of autonomous cars should begin within five years.
It is difficult for a program to recognise hand gestures of workers who are often present in areas where roads are being constructed (gestures easily understood by humans, such as stop, go, slow down). However, once workers have wireless beacons, this problem will be solved. All that is needed are relatively modest changes in infrastructure.
Job displacement: an interim problem
New technology will create new roles for people. At the same time, it threatens the job security of truck drivers, specialists in medical imagery and agricultural workers. People in receipt of a guaranteed basic income would have to keep educating themselves. Thus we should pay the unemployed only if they were acquiring new skills.
-jk-