Which AI prediction from the past has aged the worst?

[TheReal_Aaron96]

Predictions about artificial intelligence have a long history of being wrong in both directions. In the 1950s and 1960s, researchers expected human level AI within a generation. In later decades, sceptics said computers would never beat humans at chess, Go or creative tasks, and they were proved wrong too. More recently, confident timelines for self driving cars and job automation have come and gone without the promised results

Some predictions were too optimistic. Fully self driving cars were promised by several companies for around 2020, and most of us are still driving ourselves. A famous suggestion from 2016 that we should stop training radiologists because AI would replace them aged badly, since demand for radiologists has grown. Early hype around IBM's Watson in healthcare also faded after the results disappointed

Other predictions were too pessimistic. Many experts thought AI could not write convincing essays, generate realistic images or hold natural conversations for decades. Then large language models and image generators arrived much faster than expected. Some people who said AI could never be creative have quietly changed their minds

I find this topic interesting because it shows how hard it is to predict technology, even for experts. It also makes me more cautious about today's predictions, whether they claim AI will transform everything in five years or that it is all hype

Which AI prediction from the past has aged the worst? And which current prediction do you think will look silly in ten years?