Hacker News Reader

Comment by ryanSrich | Hacker News Reader

These types of comments are always hilarious. Really? You think we've topped out on 1bn dollar companies in the year 2017? AI, Self Driving Cars, Improvements to Flying, SPACE!? You honestly think there's nothing left to invest in?

View on HN

I think it's more likely that the current hype cycle may be coming to a close. The items you mentioned will all be part of the next one.

Useful AI and SAE 5 cars are multiple years away, and this leg of the business cycle is already 9 years old.

As for the other two categories, only SpaceX is relevant in the spacelaunch world, and I don't think they're going public anytime soon. I don't know of any startups that I would say have a decent chance in commercial aviation.

"Useful AI" is vague enough that you can probably make any sort of statement about it. If you mean things like an Echo that can take "understand" and take actions on the level of a slightly dull personal assistant, you're probably right. (Which is at least correlated with SAE 5.)

Of course, there's all sorts of machine learning and related AI-ish things that are potentially useful in the interim.

Fully agree on the other points.

The Amazon Echo having a 60-item case statement is not how "AI" is typically promoted.

You may have misunderstood my point. I'm saying that being able to interact with a digital assistant (I only used Echo as an example) the way I would with a human admin and have it take appropriate actions is many years away. It would be an interesting application for many of us, but it's not close.

I agree that it is many years away.

I'll believe AI is good for something other than YouTube recommending Top 40 music for me to play next, when Google News stops showing pictures of unrelated celebrities next to political stories (e.g.).

ML/AI is used all the time. Every check deposited via mobile phone uses character recognition. Every letter or package mailed also.

If OCR is now considered to be AI, then that's a perfect example of my point that "AI" is defined downwards in practice.

Reply on news.ycombinator.com