Press enter to see results or esc to cancel.

You cannot be serious?

TARS: I have a cue light I can use to show you when I’m joking, if you like.
Cooper: That might help.
TARS: Yeah, you can use it to find your way back to the ship after I blow you out the airlock.
[cue light flashes]

We’re always on the lookout for ways to make machines understand what people mean rather than just what they say.

And as a generalisation they do pretty well.

With enough data and training they can understand if you said “this”, you may have meant “that”.

In fact it can get spooky at times.

But what they have a preeeeetty big problem with is irony and sarcasm. Really?



But things are set to change, and it’s something we’re keeping a keen eye on…

UC Berkeley’s David Bamman and the University of Washington’s Noah A. Smith, have made some notable strides. Take a look at this:

And for those with such an interest, here’s the paper itself: