Bob Mankoff shows off The New Yorker's new crowdsourcing tool
The New Yorker's cartoon caption contest asks people to come up with a caption for one cartoon a week.
It's a popular contest but with about 5000 submissions a week, the magazine wanted to find a faster way to sift through all those jokes.
It turned to computer scientists at the University of Wisconsin who created a way for people to vote on submissions.
We met with cartoon editor, Bob Mankoff, who showed us the new feature.
You know from the New Yorker's point of view, from our point of view, it's more engagement on the site.
And I think it's fun for people to do.
What will happen is that once it's live on the site, thousands of people will be doing this simultaneously.
The algorithm behind all this is working feverishly to look at the rank, how people are judging this, and eventually it's going to bubble up.
Essentially to look at which ones are ranked higher rather than lower, for as more people come to judge it.
And the idea is so that you can as quickly as possible find the rankings.
Okay, so it says please rank the entries to this capturing contest image, then click Done button.
You can rank as many or as few as you like, but five is too few, and 5,000 is way too many [LAUGH].
So here's a Peter Vey drawing, and they're all carrying this big briefcase.
And it says, he wanted to die the way he lived.
So it's a funeral, [UNKNOWN] For the record, I hate our new business cards.
So that, you sorting up to think about that now with all these people on the one business card.
It might be a little bit too long to work as a joke I think it's funny.
They say you have to think it out.
On the other hand it's clever I'm all for flexible working, but this job sharing, okay, see that's a bad caption.
Any caption that ends with but this is ridiculous.
So that's unfunny.
It's a lengthy brief.
Compared to the other ones it's unfunny.
Now you understand, there's thousands of these.
The best way to look at this mass if this is 5,000, or let's say it's 50,000.
Is through crowd sourcing but with people.
People are best at this.