Every ice cream flavour imaginable
There is a correlation between what we’re seeing with AI, specifically generative AI and LLMs, and what is increasingly a consumer era of infinite choice. I’ve already written this year about how the world has enough average – the type of output these technologies generate.
This link between averaging and increased consumer choice isn’t new. Amazon launched ‘Merch on Demand’ a decade ago in 2015 enabling anyone to sell products without upfront costs, or the need to manage and keep stock in the traditional way. This opened up more consumer choice but it has also led to the mass distribution of lower-quality products.
Likewise, with the rise of generative AI, platforms like Medium and LinkedIn are increasingly full of AI-generated content – a big and growing increase in content, but a big and growing decrease in content quality. Books being generated by AI are also being sold on demand by Amazon, sometimes even competing with real authors on the same platform while stealing parts or whole concepts of their work.
Average and the paradox of choice
This was an interesting, yet frivolous, example of generative AI being used last year. As reported by the Times newspaper (behind a paywall): AI is remaking Milan’s ice cream scene and driving a gelato boom.
The big promise here was around AI being able to suggest new ice cream flavours, increasing consumer choice. As the article describes:
“[AI] could help gelaterias become more competitive, both getting the creative juices flowing and ironing out production inefficiencies…”
The paradox of choice – where having fewer options helps us make decisions with less anxiety and greater satisfaction – is getting worse in many aspects of our lives. It’s busyness. With AI you can now have every possible flavour and combination of ice cream all of the time. But most of that choice is probably average.
This reminds me of how when I was 14 years old my mind was blown by my family getting satellite TV and the jump from five terrestrial channels to something like four hundred. The problem is most of that content was average. It was filler. While we do now live in an era of great TV production and content, it’s still surrounded by even more content that’s not. Everything increasingly has to stand out from the average more and more. This is only going to get worse with AI-generated content.
As another example, Coca-Cola, who I’ve written about experimenting with AI advertising using their IP, introduced their first ever AI-created flavour in 2023. You probably never heard of it.
I also read about Starbucks whose big bet on growth had been through customisable drinks. This led to complexity and, ultimately, customer dissatisfaction because of longer wait times in stores. They’re having to rethink a strategy of ‘anything for everyone’ where consumer choice has made their business more expensive and difficult to operate and run. They are also scaling back on a business strategy around automation in stores in favour of hiring more baristas.
My conclusion to this is that it’s going to be increasingly radical to not default to averaging as a business strategy. To have a simple, clear customer proposition. To stick to less ice cream flavours, but also the very best that people will continue to want to queue around the block for. That will be how we recognise quality and craft in an increasingly average world of outputs, products and experiences.
This is my blog where I’ve been writing for 20 years. You can follow all of my posts by subscribing to this RSS feed. You can also find me on Bluesky and LinkedIn.