Can AI master the art of humor? | Bob Mankoff - TED Talks Daily

Summary

Podcast Summary: TED Talks Daily – "Can AI Master the Art of Humor?" | Bob Mankoff

Episode Information:

Title: Can AI Master the Art of Humor?
Speaker: Bob Mankoff
Hosted by: TED
Release Date: December 24, 2024

Introduction

In this engaging episode of TED Talks Daily, host Elise Hu introduces Bob Mankoff, the former editor of The New Yorker cartoons. Mankoff delves into the intriguing question of whether artificial intelligence (AI) can replicate the nuanced human ability to create and understand humor. Through his extensive experience with the New Yorker Caption Contest, Mankoff explores the intersection of AI and humor, offering insights into both the potential and limitations of machines in mastering comedic art.

The Challenge of AI and Humor

Bob Mankoff begins by addressing the broader capabilities of AI, acknowledging its rapid advancements while questioning its ability to grasp human creativity, particularly in humor. He humorously remarks:

“I don't care if [AI] turns all of us into paperclips, as long as they're funny. Paperclips and the fact that it makes stuff up hallucinates for me. That's not a bug, that's a feature.”
[02:13]

Mankoff emphasizes that humor is deeply rooted in human experiences and cultural contexts, making it a complex area for AI to navigate. He cites Norbert Wiener’s early warnings about thinking machines, humorously adding:

“But not by a thinking machine, but by an unthinking one. He was run over by a bus. That's not true. I made that up. I hallucinated it. Cause it's funny.”
[02:52]

The New Yorker Caption Contest

A significant portion of Mankoff’s discussion centers on the New Yorker Caption Contest, a weekly challenge where readers submit captions for a given cartoon. Initially managed by Mankoff and his team, the contest transitioned to a crowdsourced model in 2016, leveraging over a million online votes to determine the funniest captions. Mankoff shares his skepticism about AI’s ability to handle the subtleties of humor when evaluating thousands of submissions:

“Now, with up to 10,000 captions every week, how do you select that? [...] I wasn't really surprised when Vincent Van Hook, then the chief data scientist for Google's DeepMind, now the head of robotics, sent me this email indicating that winning the caption contest, which was actually somewhat of the sine qua non of human creativity, and... And it turned out that was the case. All of the AI juju circuit 2016 wasn't up to the task.”
[05:45]

Mankoff recounts the initial failures of AI in accurately decoding images and understanding humor, highlighting that even advanced systems like DeepMind struggled with the contest's requirements.

AI’s Progress and Limitations

Despite early setbacks, Mankoff notes that AI has made strides in processing and generating humorous content. He references a collaborative paper aiming to evaluate AI's performance against human creators in tasks related to captioning and humor explanation:

“What it did was create 50 synthetic new cartoons generated from this old data in which there were five options for captions. I picked four of them and I gave them to cartoonist Shannon Wheeler to draw up. Now, Shannon said, well, these are weird. They don't really seem like it's sort of an uncanny valley of cartoons. They're not quite there.”
[09:30]

This experiment demonstrated that while AI could generate plausible captions, the quality and authenticity of humor remained lacking. Mankoff acknowledges AI’s potential as a brainstorming tool for cartoonists but remains cautious about its ability to fully replicate human humor.

The Human Element of Humor

Central to Mankoff’s argument is the intrinsic link between humor and human vulnerability. He eloquently states:

“A human sense of humor is not about making a joke or getting it. It's rooted in our vulnerability. It's the blessing we get for the curse of mortality. Mark Twain said, the true source of humor is not joy, but sorrow.”
[12:10]

Mankoff underscores that humor often arises from complex emotional and existential conditions, which are challenging for AI to comprehend or emulate authentically. He poignantly remarks on the ethical implications of granting AI human-like experiences:

“If we gave AI the thousand natural shocks that flesh is heir to, that would be cruel. If we did that, it might very well want to wipe us out. And if they did, all I ask is that they take Elon first.”
[13:40]

Conclusions

Bob Mankoff concludes that while AI continues to advance and may increasingly assist in creative processes, the essence of human humor—deeply tied to our experiences, emotions, and vulnerabilities—remains uniquely human. He remains cautiously optimistic, viewing AI as a supplementary tool rather than a replacement for human creativity in humor.

Notable Quotes:

“They’re called cartoons. There is no algorithm for humor but now, with the rapid pace of AI I have to wonder, could there be a bot Mankoff?”
[02:13]
“The idea that understanding humor could be a stepping stone to creating it sort of made sense.”
[09:00]
“A human sense of humor is not about making a joke or getting it. It's rooted in our vulnerability.”
[12:10]

Final Thoughts

Bob Mankoff’s talk offers a thoughtful exploration of the capabilities and limitations of AI in the realm of humor. By leveraging his experience with The New Yorker Caption Contest, Mankoff provides a nuanced perspective on the future of AI-assisted creativity, ultimately affirming the irreplaceable nature of human ingenuity and emotional depth in crafting humor.

Additional Information: For more insights and curated TED Talks, visit TED.com.

Transcript

Ryan Reynolds (0:00)

They say opposites attract. That's why the sleep number smart bed is the best bed for couples. You can each choose what's right for you whenever you like. You like a bed that feels firm but they want soft. Sleep number does that. You want to sleep cooler while they like to feel warm. Sleep number does that too. Why choose a sleep number smart bed so you can choose your ideal comfort on either side. And now it's the lowest price of the season on the top selling i8 smart bed your best savings plus special financing limited time shop a sleep number store near you see store or sleep number.com for details hey, I'm Ryan Reynolds.

Bob Mankoff (0:39)

Recently I asked Mint Mobile's legal team if big wireless companies are allowed to raise prices due to inflation. They said yes. And then when I asked if raising prices technically violates those onerous two year contracts, they said, what the are you talking about, you insane Hollywood? So to recap, we're cutting the price of mint unlimited from $30 a month to just $15 a month. Give it a try@mintmobile.com Switch $45 upfront.

C (1:03)

Payment equivalent to $15 per month New.

Warby Parker (1:04)

Customers on first three month plan only taxes and fees Extra Speed slower above.

E (1:07)

40Gb Details the INS and outs of caring for your home. Out Uncertainty Self doubt Stressing about not knowing where to start in Plans and guides that make it easy to get home projects done. Out Word art Sorry. Live Laugh Knowing what to do, when to do it, and who to hire. Start caring for your home with confidence. Download thumbtack Today.

Elise Hu (1:42)

You'Re listening to TED Talks Daily where we bring you new ideas to spark your curiosity every day. I'm your host, Elise Hu. There's a lot that AI is capable of these days, but can it match human creativity when it comes to humor? Former New Yorker cartoons editor Bob Mankoff put the machines to the test when it comes to cartooning and quips. In his 2024 talk, he shares what he learned from trying to get AI to make jokes.

C (2:13)

Whether you're thrilled by what AI can do for us or terrified, but what AI is gonna do to us. Whether it can be funny is probably not top of mind for you. It is for me. I don't care if it turns all of us into paperclips, as long as they're funny. Paperclips and the fact that it makes stuff up hallucinates for me. That's not a bug, that's a feature. My entire career was making stuff up. They're called cartoons. There is no algorithm for humor but now, with the rapid pace of AI I have to wonder, could there be a bot Mankoff? You might think my reflexive answer to this would be, how about never? But while I don't want to be replaced by a bot, I'm not above being helped by it. Steve Jobs famously said that computers are a bicycle for the mind. If that's the case, what's AI A rocket ship. And at my age, you know what? I'd settle for a walker. The fears of machines replacing humans are not new. When the goals of machines and humans go horribly awry, at least for one of the parties, cartoons don't happen in a cultural vacuum. They're part of the zeitgeist. The guy who invented cybernetics, Norbert Wiener, who said thinking machines were putting us on the eve of destruction. Now, sadly and tragically, Norbert Wiener died. But not by a thinking machine, but by an unthinking one. He was run over by a bus. That's not true. I made that up. I hallucinated it. Cause it's funny. So these fears are not new, not novel. But now, in the immortal words of Nigel Tufnell of Spinal Tap, they go to 11. They're cranked to the max. And here is one of the maximum cranksters of all time, Elon Musk, saying, AI Is one of the biggest threats to humanity, but certainly not as big as Elon Musk. People like Elon have a P. Doom number. That's the probability AI Is going to wipe us out. I think P. Doom is pdumb. I'm interested in P Funny, and I've been using the New Yorker Caption contest to look into the probability of that. Every week since 2005, the New Yorker has presented a cartoon without a caption, and chat challenged its readers to come up with the winning caption in the caption contest. And for that, they get the glory of being in the New Yorker magazine, a huge amount of money, a house in the Bahamas that Sam Bankman freed. Actually, it's just the glory. On the page of the New Yorker, there's a contest you enter the finalists from a few weeks before. Three finalists and a winning caption. So it's staggered in that way. Each one of these images are funny. They're incongruous. You certainly think they're humorous, but they're not funny in a way that you get. They're not mentally funny. To make it that, of course, you need the right caption. Okay, but with up to 10,000 captions every week, how do you select that? Now from 2005-20, early 16, that burden fell on me and my assistants, but mainly my assistants to try to cull the good captions from what we uncharitably called the craptions. But then in early 2016, for the benefit of all humanity, but mainly for me and my assistants, we switched to crowdsourcing. So now for every contest, you vote online, and a funniness score from over a million judgments is given for all the captions. Now, overall, I'm against mob rule, but actually, in this case, the mob does a pretty good job. Usually the finalists, almost certainly almost all the time, really. The finalists come from the top 200 captions. Well, this is popular not only with the New Yorker, but it's caught the eye of data scientists, created searchers, cognitive scientists, and AI, of course, and everything adjacent to AI. I wasn't really surprised when Vincent Van Hook, then the chief data scientist for Google's DeepMind, now the head of robotics, sent me this email indicating that winning the caption contest, which was actually somewhat of the sine qua non of human creativity, and. And I was also flattered by that, of course, but I didn't think they had any chance at all of doing it. And it turned out that was the case. All of the AI juju circuit 2016 wasn't up to the task. It really couldn't even decode the image. So for the sine qua non of the human mind, DeepMind was non compos mentis and out of its depth. But time and AI marched on AI marching quadruple time. Vincent gets back to me and says, while human creativity might still be out of reach, we think we have understanding well in hand. He sends me this ridiculous uber nerd example of explaining humor. And I said, you know what? Let me give you a cartoon I did in 1997, this other watershed moment when IBM's DeepMind defeated Garry Kasparov, the world chess champion. And here's the cartoon I did then, and it says, oh, I don't want to play chess. I just want you to read the lasagna. I rate this explanation a solid B minus, but so what if it was an A? Is there ever going to be a beautiful New Yorker cartoon anthology of explanations? I don't think so. But the idea that understanding humor could be a stepping stone to creating it sort of made sense. This paper I was involved in tried to look at, compared to smart humans, what would the best AIs do on three tasks could they, from winning captions from different contests match to the right Image could they between two captions, one that won and the one was pretty good. Pick the right one. And could they explain the humor now for all of them? You know what? Yay, humans. We're still ahead, but AI is closing the gap. The most interesting thing about this paper for me was it showed a pathway for which you could create cartoon humor. And that was how we trained the contest. For 653 contests, the AI was trained fine tuned on these examples, which humans annotated a description of the cartoon explosion of the humor. Okay, if you've used ChatGPT, you sort of get the idea. Now put a number of examples like this, put it in the prompt window, rinse and repeat, and you get new cartoons. Well, Jack Hessel, the chief author of the paper, did something more sophisticated. And what he did was create 50 synthetic new cartoons generated from this old data in which there were five options for captions. I picked four of them and I gave them to cartoonist Shannon Wheeler to draw up. Now, Shannon said, well, these are weird. They don't really seem like it's sort of an uncanny valley of cartoons. They're not quite there. But it is interesting. All of these are new cartoons that never appeared anywhere that are an idea of human, of computer creativity. But I do see this now as a tool for brainstorming for cartoonists in that we played this completely straight. Shannon wasn't able to manipulate the description of the picture or the caption. Had he done that, it could have been better. Also, we could have asked it to make more. We could have put in the rankings for the humor. We could do all this to improve it. So quality comes out of quantity. You can get an awful lot of quantity here. You can have a human being in the loop to do this. But I would not go so far to give AI a true human sense of humor. A human sense of humor is not about making a joke or getting it. It's rooted in our vulnerability. It's the blessing we get for the curse of mortality. Mark Twain said, the true source of humor is not joy, but sorrow. If we gave AI the thousand natural shocks that flesh is heir to, that would be cruel. If we did that, it might very well want to wipe us out. And if they did, all I ask is that they take Elon first. Thank you.