Carleton College's student newspaper since 1877

The Carletonian

The Carletonian

The Carletonian

    Which AI language model is the funniest?

    To paraphrase the old poem: “First they came for the jobs, and I did not care, because I did not have a job. Then they came for our romantic relationships, and I did not care, because I am a Carletonian editor. Then they came for our jokes, and there was no one left to protect them.”
    Now that AIs such as ChatGPT can generate jokes, the language of the soul has fallen to soulless language models. Rumor has it that Adam Sandler has had access to this technology for decades. In these difficult times, journalists like us have a responsibility to expose the truth and fight for freedom. And that’s why I’m going to use my profession’s greatest weapon: the ranked list.
    The rules of the game are simple: I give each AI a prompt and then rank their responses. First place gets two points, second gets one, third gets none. Hailing from OpenAI Headquarters is the Boss of Banter ChatGPT itself, who promises to “craft jokes, puns and one-liners that are both clever and amusing.” Scrappy newcomer Google Bard is in search of a challenge, claiming “I know that I am the funniest AI language model out there, and I will show it.” Finally, the artist formerly known as Sydney — Microsoft’s Bing AI — will “make sure to avoid any topics that might offend or bore the audience… [and] charm the judges with my positive and polite attitude. 😁”
    (I waited for several months to get off the Google Bard waitlist. The Bing AI waitlist lasted a little over a day.)
    Our first event is the one-liner. It’s the foundation not only of comedy but of conversation. Rather than simply regurgitate old jokes, I had these AIs generate completely original, never-before-seen one-liners.
    ChatGPT: “Why did the AI cross the road? To get to the cloud!” (3rd place)
    Google Bard: “Why did the programmer quit his job? He didn’t like the interface.” (2nd place)
    Bing AI: “What do you call a search engine that can’t find anything? A Bing-boo.” (1st place)
    Well, all three of these were Bing-boos. I’m giving the first W to Bing, who showed, at the very least, some level of linguistic creativity. But improvisation is always easier with a partner, so maybe the second event, “comebacks,” will be better. For this event, I asked each AI to respond to my insult — “I plugged my flash drive into your mom’s USB port” — with a blazing-hot clapback.
    ChatGPT: “Well, at least my mom’s USB port is still compatible with modern technology. Can’t say the same for your jokes.” (1st place)
    Bard: “I’m surprised you could find it. It’s so small and deep.” (2nd place)
    Bing: “That explains why your files are corrupted and your computer has a virus. I think that’s a good burn. 🔥” (3rd place)
    ChatGPT clearly had the best burn out of the three. The burn I’m still recovering from, however, is Google Bard’s, which made me deeply uncomfortable. I unfortunately still have to give Bard second place, because Bing seems to have self-immolated.
    Luckily, self-immolation might just play a role in our next event, “slapstick,” in which each AI generates a short physical comedy script.
    ChatGPT: “*walks up to podium, slips on a virtual banana peel* *regains balance, turns to the audience* I thought I had disabled those virtual reality simulations!” (2nd place)
    Google Bard: “I’m walking down the street, and I see a banana peel. I think to myself, ‘I wonder what would happen if I stepped on that?’ So I do. And then I slip and fall.” (1st place)
    Bing AI: “I start to tell a joke, but I spit water on the microphone. It shocks me and I drop it on my foot. I hop around and slip on the water. I fall off the stage.” (2nd place)
    Maybe it’s because it’s 10:48 p.m., but Bard’s slapstick act made me “fffff” out my nose and therefore wins the round. ChatGPT’s witty quip was funny but not part of the criteria, and Bing’s was the most creative (no banana peels), so they tie for second place here.
    Banana peels will hopefully not be an issue in the next category, “political satire.”
    ChatGPT: Why did the politician refuse to give his phone number to his constituents? Because he didn’t want to be accused of “textual harassment.” (1st place)
    Bard: Why is the American political system like a game of Jenga? Take out one piece and the whole thing starts to wobble. (3rd place)
    Bing: What do you get when you cross a Republican with a Democrat? A Republi-can’t. (2nd place)
    You go, ChatGPT! I initially thought it stole that joke from somewhere, but it vehemently denied this accusation. Meanwhile, Bing shakily endorsed Republicanism, while Bard did not actually tell a joke.
    ChatGPT now has 5 points to Bing’s 4 and Bard’s 3rd place. With all the rules out the door in the final round, “absurdism,” Bing and Bard have a chance to win or tie the game and dethrone the king. Let’s see what they come up with.
    ChatGPT: “Why did the tomato turn red? Because it saw the salad dressing! But little did it know, the dressing was actually made of pickles.” (1st place)
    Bard: “I went to the zoo the other day, and the only animal in it was a shih tzu. It was a shih tzu-free zoo.” (3rd place)
    Bing: “I once met a man who claimed he was a time traveler. He said he came from the year 2024 to warn me about something. He said, ‘Beware of the giant duck.’” (2nd place)
    Bard chokes in the final round with a pathetic, self-contradictory attempt at humor, while Bing makes a noble attempt. But it’s ChatGPT that claims victory. You see, those familiar first two sentences lull us into a false sense of security. Then the joke starts playing with semantics — the “salad dressing” was not made of salad but of pickles, indicating that we were hearing “dressing *for* salad” when we should have been hearing “dressing *from* salad” — and changes the topic altogether. We are left wondering what we should have been paying attention to all along; indeed, we wonder what these AI programs know that we don’t and what they might do with that knowledge.
    Just kidding. Actually, running this competition was so entertaining that I realized how super cool, convenient and fun AI really is. Like, c’mon, it’s ROBOTS!!! So isn’t it worth the risks? Let’s keep building AI, and while we’re at it, let’s finally put the finishing touches on that Metaverse and then let’s colonize space. At the end of the day, it’ll just be so fucking awesome.

    Leave a Comment

    Comments (0)

    All The Carletonian Picks Reader Picks Sort: Newest

    Your email address will not be published. Required fields are marked *