AI has a marketing problem

Among my friends and family who don’t work in technology, I mostly see people use AI for two things. They look something up that they’d otherwise have Googled, or they make a funny picture. That’s about it.

Among friends and colleagues who work in technology, coding is the main use I see. When we talk about what else AI might be good for, most sound pretty sceptical.

I wonder how much of that comes down to which stories have cut through. People can tell you what they’re worried about. Their job, the environmental cost, possibly the end of the world. Ask what they’re looking forward to doing with it and it’s a harder question to answer. “This could put me out of a job” is easier to picture than “this could help me do something I haven’t thought of yet”.

I grew up in the 90s, when mobile phones and desktop computers were slowly starting to appear everywhere. I've lived my adult life with a smartphone and laptop. Since leaving school, I've been readily connected to the internet.

But until a few years ago, technological progress had felt glacially slow to me. Sure, phones got smaller, then bigger, and more capable. Cameras and TVs became higher resolution. The internet got faster. Everything was improving incrementally, but little genuine innovation and invention was happening, at least not in consumer markets.

Then I tried ChatGPT. It had been years since I’d got my hands on a piece of technology and genuinely wondered what I could do with it. I’d missed that feeling.

It’s also why I think making funny pictures is a perfectly good place to start. You try something because it amuses you, show someone what happened, then have another go. The interesting bit comes when you start wondering whether it could do something else. You’ll find a few things it sucks at. Then, occasionally, you’ll try something that makes you step back and say, “Wow.”

For me, one of those first things was figuring out where to start in the morning. I used to spend the first fifteen minutes of the working day piecing together where I’d got to yesterday, what I ignored that I can't today, and what had changed overnight. These days, one of my robots (as I like to call them) does that before I sit down. Sometimes I read its take on what matters that morning and disagree. Either way, I can get on with the work.

The funny thing is that the more excited I get about AI, the dumber I want my phone to be. I want more time outdoors with my family, and I’d quite like to stop spending little bits of it checking a screen. I can ignore Slack and email at weekends because I know one of the robots will call me if something genuinely urgent comes up. When I’m out with my family on a Saturday, or relaxing in the evening, I can leave my phone in my pocket and, for once, forget it’s there.