Hacker News Reader

Comment by ragebol | Hacker News Reader

No. This [0] is 4-5 years old and I don't think much progress has been made in getting a computer to classify that image as 'funny' and explain why. And if/when it could, I doubt we'd call it intelligent. And this is just computer vision, not mentioning other branches of AI.

[0] http://karpathy.github.io/2012/10/22/state-of-computer-visio...

View on HN

Hell, I know a lot of humans who wouldn't understand why that's funny. And some of them have 4-year college degrees.

Many responses in this thread are along the lines of "computers will never be as smart as Carl Sagan", ignoring that most intelligence on this planet could never dream of being half as smart as the genius we're using to define AI. Let's start by getting a computer as smart as a border collie, one of the more intelligent dogs.

Saying a computer isn't smart because it can't laugh at a highly complex visual joke is entirely the wrong way to define AI, and it's no surprise we haven't achieved it. It took humans millions of years to get this smart, and we've only been working on AI for a very, very small amount of time.

If you could encode enough information about the image and have a good training/test set, an AI's ability to determine something as funny probably wouldn't be bad. Anyway, as a subjective concept you might not accept another human's explanation about why something is/isn't funny. This isn't a simple classification problem though, probably suited to something like reinforcement learning.

Reply on news.ycombinator.com