December 2022:
I started hearing about something called ChatGPT…
June 2023:
A brainstorm for a client. Mid-session, a participant looked up from his laptop and said, “I’ve just typed a few things into ChatGPT. This is what I’ve found…”
December 2023:
At the end of each year, Benedict Evans, the tech consultant (below), produces ‘a big presentation exploring macro and strategic trends in the tech industry’…
…This was the one he did shortly before Christmas:
…And here was his proof of the hype: that red line indicates interest in AI.
…In short: something was goin’ on.
My subject is humour and brand communications — in advertising, pitches and, increasingly, conversational AI. So my big questions are: what will the impact of AI be on humour? Will it have an impact, if at all? Will a machine be able to respond to, understand and create humour?
To try to answer these, I’m doing a three-parter. Next week I’m going to give a little overview of AI and humour; the week after, I’m going to do a Q&A about the subject.
As for today? I thought I’d take the bull by the horns and go to ChatGPT itself. After all, you’ve got to start somewhere.
But I want to keep all this light, and not too techy. I don’t want it too sci-fi-dystopian, like Black Mirror, you know? So I’m going to include some New Yorker cartoons, such as the classic above and this one by the wonderful Robert Leighton:
So! Let’s go.
***
Here’s what I asked it and here’s what it answered:
Admittedly, asking ChatGPT ‘Can humour and AI work together?’ is a bit like asking a barber if you need a haircut…
…Also, don’t let the chirpy tone of ChatGPT fool you. Read the text above closely and you’ll see that puppy-dog enthusiasm of “Absolutely!” at the start is followed up by a number of caveats, such as:
‘It’s worth noting that humor can be subjective, and what one person finds funny, another might not.’
There’s more than a little bias here. And in the words of MIT Technology Review, ‘Bias has become a byword for AI-related harms, for good reason. Real-world data, especially text and images scraped from the internet, is riddled with it, from gender stereotypes to racial discrimination.’
What’s more, I hear the ‘Humour is subjective’ line a lot in my work and, as much as I respect this point-of-view, it’s become one of those lazy “ah, we don’t use humour because it’s subjective” get-out-of-jail free cards.
Why do I object to it?
First, it’s not really true — we have a very good knowledge of the conditions in which humour thrives, its mechanisms and, increasingly, how it works with certain demographics. Moreover, we have a very good knowledge of how to minimise potential offence. Indeed, this is the bedrock of my business.
And second, it’s a bit like saying to a wine-grower, “all wine is subjective”. Well, yes. And? Personally, I’ve always enjoyed a glass of Malbec. But if a friend of mine orders a Shiraz I’m not going to throw a wobbly.
So I read this text above and some of it, frankly, verges on cliché; some verges on misinformation. Our good friend Benedict Evans says in his presentation:
…And I have to agree. It reads well, but it’s an intern’s contribution: aiming to please, saying bland things in meetings, clumsy at the coffee-machine.
What it lacks, in a word, is rigour.
***
Next week? A little overview of AI and humour. And if, by then, AI has taken over the world and we’re living under the stairs on tinned rice and lentils, then it’s been great knowing you all. But all being well, I think we should be a-OK.
Many thanks for reading,
Paddy
Book a free half-hour meeting with me here.
ncG1vNJzZmiaopa7pb%2FAp5uhrZ2kwrN60q6ZrKyRmLhvr86mZqlnn6N6tbTEZqCnrJWnu6bAjKemm6eUrnqsus6wqmaxn6q%2Fpg%3D%3D