Why some physicists think we're building computers wrong for A.I.

There might be a better way.
Derya Ozdemir

Have you ever considered the possibility that analog computers might actually be better suited for the tasks we're facing in the 21st century than digital ones? As odd as it sounds, this video by the YouTube channel Veritasium's Derek Muller makes an excellent case for why that might be.

Although digital computers have served us well for decades, he argues that with the emergence of artificial intelligence, a completely different type of computer, analog, is required. An analog computer is a computer that is utilized to process analog data, and it stores data in a continuous form of physical quantities and performs calculations with the help of measures. Compared to the digital computer, which uses symbolic numbers to represent outcomes, the analog computer does not employ symbolic numbers.

Most Popular

As artificial intelligence and deep learning applications become more widespread, the requirements for improved performance, greater deep neural network (DNN) model capacity, and lower power consumption are becoming increasingly vital to businesses. If you want to learn how analog computers can help and why they are making a comeback, make sure you watch the video embedded above, and, as always, happy viewing.

message circleSHOW COMMENT (1)chevron