Ten memes explained by GPT-4V

GPT-4 with Vision has a sense of humor?

By Philip Winston | Thursday, October 05, 2023

With its new GPT-4 Vision model, ChatGPT can now answer questions about images. Below I asked ChatGPT to explain ten memes. I’ve found explaining humor is a good way to judge how well an AI understands the input.

It did amazingly well. It nailed the first eight memes and got partial credit on the last two. I’m really impressed. Something about an AI which can understand images seems more human than one which can only read text.

ChatGPT’s multi-modal capabilities will improve over time, likely expanding into sound, music, video and 3d models and environments. In the future AIs will ingest and produce every medium that humans care about. The fact that AI was ever text-only will be a historical footnote.

Results

Below are the ten memes, each followed by ChatGPT’s verbatim explanation.

#1 Rock Paper Scissors

The image humorously depicts a fictional history of the “Rock, Paper, Scissors” game.

Rocks: At the very left, there’s an image of a rock. Since rocks have been in existence since the beginning of the Earth, the implication is that for a long period, only rocks existed in the game.

Invention of Scissors (~3000 BC): The next image shows scissors with a date of “-3000,” suggesting scissors were invented around 3000 BC. With only rocks and scissors in the game, rocks would always win because, in the traditional game rules, rock crushes scissors. Hence, the “Rock advantage” period.

Invention of Paper (~179 BC): The rightmost image illustrates the invention of paper around 179 BC. With the introduction of paper, the game becomes “balanced,” as each option (rock, paper, or scissors) has one that it can beat and one that it can be defeated by. The blue timeline labeled “All ties” humorously suggests that when only rocks existed, every game was a tie since both players could only choose a rock.

Overall, the image is a playful take on the evolution of the “Rock, Paper, Scissors” game, suggesting that its dynamics changed with the invention of each component.

#2 Sponge Bob

This is a meme that uses imagery from the popular animated TV show “SpongeBob SquarePants.” The characters shown are Patrick Star (the pink starfish), SpongeBob SquarePants (the yellow sponge), and two other characters in the background.

The caption humorously comments on a trope in storytelling where writers often include a romantic subplot between a male and female protagonist, even when the story might not necessarily need or benefit from it. The text “Writers trying not to make the girl protagonist fall in love with the boy protagonist” paired with the strained and exaggerated expressions on SpongeBob and Patrick suggests that avoiding this trope is a difficult task for writers, almost as if they’re battling their own instincts or popular expectations.

In essence, the meme pokes fun at how prevalent and sometimes forced romantic storylines can be in media.

#3 AirBnB

This meme humorously contrasts the experiences one might encounter when staying at an AirBnB versus a traditional hotel, both priced at $225 per night.

AirBnB Experience: The top half depicts the potential negatives of staying at an AirBnB:

The skull image, a common meme symbol for “feeling dead inside” or being exasperated, suggests a sense of overwhelm or frustration. The texts show common complaints or issues guests might face: strict entry times, detailed instructions, potential damages or mishaps leading to the host’s frustrations, and the possibility of losing a significant cleaning deposit.

Hotel Experience: The bottom half illustrates the more straightforward, albeit potentially bland, experience at a hotel:

A calm couple simply asks for a room. The hotel staff (represented by the “feelsgood” meme face) calmly directs them to their room. The overall ambiance in this section is devoid of any potential conflicts or issues.

The meme highlights the idea that while AirBnBs might offer unique, homely stays, they can sometimes come with unexpected complications or strict rules. In contrast, hotels, while potentially less personalized, offer a standardized, hassle-free experience.

#4 Darth Vader

This image is humorous due to the juxtaposition and blending of two vastly different contexts.

On one hand, we have a scene that appears to depict a formal or serious event, possibly related to sports or a ceremonial gathering, given the attire and expressions of the majority of the characters. These individuals seem to represent athletes or officials from an event, perhaps reminiscent of the Olympics or a similar international competition.

On the other hand, amidst this group stands Darth Vader, a character from the “Star Wars” universe, who is entirely out of place in this context. His presence here is unexpected and incongruous. The humor comes from the absurdity of seeing a fictional and iconic dark lord from a space opera saga inserted into a real-world setting. It plays on the element of surprise and the blending of fictional and real contexts.

#5 Tough Job

The humor in this image comes from subverting expectations. Here’s a breakdown:

Setup Expectation: The bold text at the top, “You really think YOUR job is tough?” paired with the rugged appearance of the man, sets up an expectation for a sarcastic or condescending punchline. It looks like it’s leading to a comparison where the pictured man’s job is far tougher than the viewer’s.

Subversion of Expectation: Instead of delivering a punchy or mocking retort, the text below offers a genuine, understanding, and supportive statement: “I’m sure you’re right. Most jobs have difficult aspects to them. Kudos to you for sticking with it.” This unexpected wholesomeness is the twist, and it’s humorous because it’s not what the viewer anticipates based on the initial setup.

Contrast: The rough and tough appearance of the man in the image contrasts with the gentle and affirming message, further amplifying the comedic surprise.

#6 College

The joke is a humorous reflection on the transition from the excitement of being accepted into college to the immediate reality of academic responsibility. The top image shows a person looking joyful and captioned with “Me after finally getting into college,” while the bottom image, featuring the same person with a more stressed or disappointed expression, is captioned with “Me after getting my first assignment.” The humor comes from the quick shift in emotion, emphasizing the juxtaposition between initial elation and the realization of forthcoming challenges.

#7 WTF

The joke plays on a relatable experience many people have: walking into a room and suddenly forgetting why they entered. The image features a cute character looking puzzled in a kitchen setting, with the text “WTF DID I COME IN HERE FOR” amplifying the character’s confused expression. The humor arises from the juxtaposition of the adorable character with the relatable forgetful moment, expressed in a slightly irreverent way with the use of “WTF.”

#8 Bartender

The joke is a play on the meme format of “Nobody:” followed by “Literally not one person:” to emphasize that no one asked for or prompted a specific action. In this case, the action is a depiction of a mechanic pouring a liquid from a height into a car’s engine, likened to bartenders who sometimes pour drinks from a height to create flair or showmanship. The humor comes from the exaggerated comparison, suggesting bartenders pour drinks in an overly dramatic fashion when there’s no actual need to do so.

#9 Stairs

ChatGPT painfully doesn’t understand this one…

The humor in this image stems from several elements:

Literal Interpretation: The child’s question, “What’s upstairs?” is a straightforward inquiry about what is located on the floor above. However, the man’s response, “Unfortunately the stairs don’t talk,” is a literal interpretation of the question, humorously suggesting that stairs would need to communicate in order to answer the child’s query.

Play on Words: The joke plays with the multiple meanings of the word “what’s”. Instead of answering what items or rooms might be located upstairs, the man jokes about the inability of stairs to speak.

Unexpected Response: The humor is heightened by the man’s deadpan and unexpected response to a simple question. His answer is clearly not what the child or the viewer expects.

Contrast in Seriousness: The man’s attire and demeanor appear formal and business-like, juxtaposing the silliness of his answer and amplifying the comedic effect.

You’re missing that “what’s upstairs” sounds the same as “what’s up, stairs?”

You’re right! The phrase “What’s upstairs?” phonetically sounds like “What’s up, stairs?” This adds another layer of wordplay to the joke. When the man responds with “Unfortunately the stairs don’t talk,” he’s humorously treating the question as if the child is addressing the stairs directly and expecting a reply from them. This play on homophonic words and phrases further contributes to the comedic element by introducing ambiguity and multiple interpretations to a seemingly simple question.

#10 But my name is on it…

Mostly right but misses an important visual detail.

The comic humorously addresses a common occurrence on the internet: content theft or lack of attribution. The sequence of the panels is as follows:

A character claims to have created something and shows it off, asserting ownership with the phrase “I made this.”
Another character, labeled “Internet,” quickly dismisses the creator’s claim, saying “No you didn’t,” and takes the creation.
The original creator attempts to assert ownership again by pointing out that their name is on the creation.
The “Internet” character, now holding the creation, denies this fact with the statement “No it isn’t,” suggesting that the internet often takes content without giving proper credit or acknowledging the original creator.

The humor stems from the blunt and casual manner in which the “Internet” character disregards the original creator’s claims, reflecting the often impersonal and appropriative nature of online content sharing. It’s a satirical take on how original content can be easily taken and rebranded on the internet without giving credit to its original creators.

He’s cutting off the name with scissors!

You’re right; upon closer inspection, the “Internet” character is indeed using scissors to remove the name, which further emphasizes the joke about content appropriation on the internet. The act of cutting off the name showcases the lengths to which some online entities go to claim content as their own. My apologies for overlooking that detail initially.