@Marcoarment implied that I do not understand humor by asking whether I had heard of this thing some people like called humor, do you agree?
I wonder if Brooke Gladstone realizes that she's begun to build a dataset for actually teaching a robot to comprehend jokes.
Both Brooke and some comedian.
If you care, message me and I'll fix it, I don't want to lose my flow.
and yea getting this punctuation right nearly made me lose my flow. @Draftsapp fail.
Anyway, now that my echoing memory has been breached, I need to reconstruct my train of thought.
They both were explaining why a joke similar to the one that follows was funny:
a man sees a friend that he has not seen for a very long time walking down the street. He looks happy but he has an enormous orange head. He says
Well I found a genie...
I forget the two first wishes, and this ruins the rhythm of the joke, but…
and what about your head‽
well, third I wished to have an enormous orange head.
iOS just autocorrected me to enormousness. I've never typed that, at least as far as I know. Is this a facebook like joke on me; being studied down to the millisecond I respond to something on this little magic piece of glass attached to some other glass that allows electrons to flow through it so easily (note: silicon oxide is a primary component of both sand and glass, both of which have some kinds of fluid properties if I remember correctly)?
Does anyone have a book to recommend on silicon the element? I know there's one on salt, but I'd prefer to stat with silicon and it's properties.
Anyway the comedian knew the joke right away. She only had to say that
there was a man down the street with a giant orange head Before he responded…
That's my favorite joke.
He then gives the explanation that I would have given, that one has set up an expectation of a joke given a formulaic structure such that I was able to replace out those statements for variables, since the essential feature of humor was invariant to transformations so long as those properties held. This is similar to the concept of the shaggy dog story, or the subject of the movie the aristocrats, though that is far more startling than it is funny.
for bonus credit, the artificially intelligent text and video parser could figure out what individual people would think is funnier based on a previous conversation with the person •the Aristocrats• or •the Sophisticates•.
The essential character is formula built up and then left to collapse in its almost dada…oh. That's a well told shaggy dog story. You need to have heard a lot of jokes to get this joke, as otherwise you won't have built up the expectation.
Brooke though responded with an entirely different tact and I'll go back and listen to her, cause I quote:
I always thought it was… because the answer is that human beings just do stupid [shit]…
Hahahahahaha Things. Just really idiotic choices. Your interpretation is by no means incorrect. . . . I don't think anyone could not have heard any jokes like them and still feel that exact same way about it.
I'll just listen through to get his name… **fine**
they would just say:
Why would you do that? That doesn't seem like a thing to do, what's the rest of the joke?
The joke is that we now have a textualized (& machine readable to make @eseiver happy) account of the exchange and if a text reading program could read my somewhat shoddy and inaccurate transcript, could it provide the answers that they gave? They both have formal characteristics. One requires a notion of what a joke is, and how they are conventionally told. That is, it needs a degree of learned self-reference, to be comprehended. The other just takes the joke as a statement of human stupidity, no self reference required, but past experience is.
And for a bonus, if it can figure out that the second meaning of the joke, as just a statement of human stupidity, but to not comprehend how it is a joke. That is, to not be surprised by that, and to be confused as to how this was a joke. Or in effect whatever could be inferred from @mikecaplan's: "what's the rest of the joke". I didn't catch his twitter handle. I hope that's it. And if not, hopefully someone (*extra bonus for the AI if it can figure this out*) will let him know of my earlier impertinence.
In any case, I've now laid out a research programme that I'm pretty sure would constitute solving the problem of developing a computer system that can comprehend jokes. I can't believe you could have read this far and not cracked a smile a bit, but just in case it helps :) .