FROM PUNCHLINES TO PROMPTS: HOW SOCIAL MEDIA AND GENERATIVE AI ARE REWRITING HUMOUR

Social media has transformed humour from private conversation into a rapid, visual and participatory language. Memes, reels and reaction posts have shortened attention spans while making irony, remixing and cultural references central to comedy. Generative AI is now extending this shift, allowing ordinary users to turn jokes into increasingly sophisticated visual experiences.

Humour was once largely a human-to-human experience: a joke told across a table, a cartoon in a newspaper, a comedy routine on stage or a witty remark exchanged among friends. Social media has changed that equation. Today, humour is faster, more visual, collaborative and deliberately designed for sharing. A joke is no longer simply something people hear and laugh at; it can become an image, meme, Reel, GIF, reaction, parody, deepfake or AI-generated visual within minutes.

The smartphone has effectively turned everyone into a comedian, editor and publisher. Platforms such as Instagram, YouTube, X and WhatsApp have created an environment where humour competes not only for laughter but also for attention, comments, screenshots and shares. The funniest content is often that which can be understood almost instantly. Context is compressed into a facial expression, a familiar film scene, a piece of dialogue, a trending song or a recognisable cultural reference.

India offers an especially interesting laboratory for this transformation because humour here is multilingual and deeply rooted in popular culture. Hinglish, regional languages, Bollywood, cricket, politics, television, food and everyday frustrations constantly feed the country’s meme economy. Research into Indian meme marketing has found that Hinglish, observational humour and self-deprecation are particularly prominent, while memes connected to familiar Indian cultural moments can outperform generic formats.

The evolution can be seen in the social-media behaviour of brands. Amul’s topical cartoons represent an older form of visual humour adapted successfully to the digital age. Zomato and Swiggy took the next step by making humour part of their social-media personalities. Their pandemic-era meme activity was significant enough to become the subject of academic research examining how memes could function as advertising and audience-engagement tools.

Zomato, in particular, demonstrated that a brand could behave almost like a witty participant in an ongoing internet conversation. Its social-media voice has relied on topical references, wordplay, relatable situations and rapid responses. Swiggy has similarly experimented with repeatable meme formats and participatory ideas. The result is a fundamental change in the relationship between advertiser and audience: people do not merely consume the joke; they remix it and sometimes improve it.

Britannia’s recent ‘Prashant’ episode illustrates the same phenomenon. When a creator humorously mispronounced ‘croissant’ as ‘Prashant’, the brand did not treat the mistake as something to correct. It embraced the joke and turned the accidental pronunciation into a social-media character and cultural moment. The campaign reportedly generated extensive earned media, views and engagement.

This is where Generative AI introduces another major shift. Earlier meme culture depended heavily on finding an existing photograph, movie frame or template and adding new words. Generative AI removes much of that limitation. A user can now describe an imaginary situation and ask an AI system to visualise it: a politician as an 1980s film hero, a celebrity in an absurd historical setting, an ordinary office meeting as an epic battle, or a familiar Indian family transformed into a Bollywood poster.

The technology is therefore changing not only how humour is distributed but how jokes are conceived. The joke can begin as a sentence and end as a fully constructed visual world.

A striking Indian example came during the Delhi Assembly election campaign in 2025, when political parties used humorous and satirical posters and AI-generated videos to attack their opponents. The election became what one report described as a digital ‘memefest’, demonstrating how generative tools could turn political arguments into highly visual satire.

That trend has continued. An AI-driven nostalgia trend involving photographs transformed into 1980s-style portraits recently evolved into a political meme exchange involving India’s major political parties. What began as an entertaining visual experiment became a vehicle for satire and political mockery.

Generative AI also allows brands to make audiences the protagonists of the joke. Prime Video’s campaign for Mirzapur used Generative AI and WhatsApp to place fans inside personalised elements of the show’s fictional world, including memes and AI-generated video sequences. The campaign reportedly reached 310 million people and generated 3.5 lakh shares.

This points towards a more participatory future for humour. Instead of watching a comedian make a joke about a fictional character, audiences can increasingly ask AI to place themselves inside the joke. Instead of forwarding a meme, they can generate a personalised version for a friend.

But there is a catch. Social media has also accelerated the distance between humour and context. A joke that works brilliantly within one community may appear offensive, misleading or incomprehensible outside it. AI intensifies the problem because convincing visuals can now be manufactured at extraordinary speed. The same technology that makes satire more imaginative can make misinformation, impersonation and political manipulation more believable.

The evolution of humour therefore reflects a broader transformation in human communication. We are moving from spoken jokes to visual jokes, from consuming jokes to remixing them, and now from remixing existing images to generating entirely new realities.

Generative AI is not replacing the human sense of humour. At its best, it is giving that humour a new canvas. The important question is no longer whether a machine can make something funny. It is whether humans can continue to understand why something is funny, when it is appropriate, whom it affects and what happens after the laughter.

In the age of social media, humour has become a language of participation. In the age of Generative AI, it is becoming a language of visual imagination.