Hacker News Reader

Comment by nopinsight | Hacker News Reader

This appears to go against Rodney Brooks' (former director of MIT's CSAIL, member of NAE, and a robotics expert) predictions, which was just published last month: https://rodneybrooks.com/my-dated-predictions/

"NET (No Earlier Than) 2022"

"First driverless "taxi" service in a major US city, with dedicated pick up and drop off points, and restrictions on weather and time of day."

"The pick up and drop off points will not be parking spots, but like bus stops they will be marked and restricted for that purpose only"

A possible resolution is in the case that he meant somewhere more ‘major’ than Phoenix, AZ (1.6 million population; 5th largest city in the US), but that is a bit hand-waving.

Any insight or comment for this?

View on HN

Yes. This definition of his:

"My milestone predictions below are not about demonstrations, but about viable sustainable businesses. Without them the deployment of driverless cars will never really take off."

In particular, he's predicting when a company does this profitably, not at a loss as part of research.

(Btw, thanks for linking - that prediction list was a great read.)

Arguably, that milestone hasn't been reached for rideshare apps in general, even ones that use real drivers!

That's just a deliberate business model decision though. If all the ridesharing services decided tomorrow to double their prices and downsize in line with reduced ride volume, they could probably turn a profit fairly quickly, albeit on lower revenues and (probably more relevantly) sharply reduced valuations. People will still use them even if they cost as much or more than taxis. They just wouldn't use them as much.

Two years before their first flight, Wilbur Wright confided to Orville that man would not fly for 50 years.

It took until two years before the first atomic bomb for Enrico Fermi to be convinced that one could be built at all.

Even the foremost experts in a field have a terrible prognostication record, because their predictions are necessarily based on what they do and do not know how to do at the moment.

Within a year or two, a breakthrough could be made by someone else—or by they themselves!—that changes everything.

I have no expertise with self driving cars. I'm not even an engineer. But this is a case where a lot of experts might be wrong. I think there are many paths to a self-driving car, and an expert might only be familiar with their particular research. That is why there are so many vocal voices saying it can't be done. But these voices only know what doesn't work--they don't know what does work

Yep. Was also predicted a computer would not beat top Go player until 2026. Love seeing some of these coming in sooner now. For decades it was the other way.

People tend to put their predictions over the same time-frame. How far something actually is seems to matter very little.

My insight is some guy tried to predict the future and was wrong. He'll join a long line of other prognosticators, it seems.

Oh, Rodney Brooks is hardly "some guy". His wikipedia page is a start, but you can find more information on him in AI - A Modern Approach (the standard AI introductory course textbook).

Reply on news.ycombinator.com