AI Can Eat Your Lunch, Just Not Mine
I've noticed a shift over the last year in how people are talking about AI. After the Canterbury Tech Summit the topic has been rattling around my head and refusing to leave.
I'm finding the conversation has moved on from "AI is going to replace us all" and is well and truly in the "it's here, now we need to figure out how to get this thing to play into our strengths" phase. I don't really know yet whether that's a good thing, but for the tech world it seems to be where we've landed. However, if I only stayed focused on what the tech world was thinking, I'd be missing the point entirely.
Tomo, one of our programme leads here at Smudge, wrote recently about how he and his wife use AI completely differently, and how what you put in makes the difference between a nice meal and a burnt one. I've been chewing on this since reading it.
Not long after reading it I started to see a cognitive dissonance around when people think AI is the bees knees, and when it's gone too far.
A bit of your soul goes into it
When you put craft, attention, care and effort into making something, you tend to leave a bit of yourself in it. Consciously or not. Because a bit of your essence has gone into the thing, it's worth more, to you, to the people who care about you, and sometimes to your customers. It's that bit of you that makes it valuable.
Art is the obvious example.
Something I've observed recently, from artists close to me and artists I follow, is the tendency to say AI doesn't belong in art, AI can't make art, and potentially that AI is toxic to art. Essentially the argument is that for something to be art, it has to be made by an artist and not a machine.
What's made me chuckle is that these same people, often after telling me how bad AI is, are the first in line to open one of these tools and ask it a question, run some numbers, or ask it to draft up a contract so they don't have to pay a lawyer.
Basically, "AI is great, as long as it's not eating my lunch!"If it's eating someone else's, well, that's business baby.
Is this hypocrisy?
I sat with this observation for a while, feeling reasonably smug for having spotted it and wondering whether I'd missed my calling as a comedian.
Unfortunately, whilst stuck in traffic after making a wrong turn driving home, I realised I was of course, completely and utterly wrong. The artist wasn't being a hypocrite, I was. The artist was of course being human in a very specific and very forgivable way.
They know there's a bit of themselves in the painting because they've stood in front of a canvas at 11pm fighting the urge to throw the whole thing out because they think it's terrible. They know exactly how much blood, sweat and tears went in, because it’s theirs.
They have, however, likely never sat across a table watching a deal go bad.
So when they look at the contract, they see words on a page, and the tool can read words on a page faster than any human alive. What they don't see is the twenty years of knowing which clause will matter in three years' time, or worded in a way that puts them at risk.
To the untrained eye, everything looks standard in a contract and it's hard to know when someone is trying it on, and what should be pushed back on without torching the relationship arguing about clauses that were never going to harm you.
The craft, and the real value in it, is visible from the outside. The artist values a good contract, but only the finished thing, not the judgement behind it. So outsourcing it to AI doesn't feel like a loss at all.
Embodiment is blindingly obvious from the inside and almost impossible to see from the outside. Our outrage at AI splashing around in our space isn't an inconsistency, more of a distribution problem. We defend the crafts we've inside of, and we're relaxed about the ones we've only ever bought.
Guilty as charged
Last year I went to my fiancée in a crisis. I had experimented with ChatGPT and reached the conclusion that it might be time for me to retrain as an electrician; AI could write user stories and I felt I was going to be obsolete by Christmas.
She listened patiently, then asked me what I thought my value actually was. It was a good question, and one I hadn’t stopped to think about in a while. Naturally we started talking about it and it was of course the value I add goes far beyond creating user stories. She wisely suggested that, for the fire safety of our house and everybody else's in the wider Canterbury area, I should take a breath.
I suspect we've all had a version of this in the last 18 months. The part of our work that is so closely tied to our identity feels like it is under threat from automation. That bit of work we feel we've toiled over is the exact thing people are looking at and saying AI will now solve for them.
That judgement comes from people
Here at Smudge, we build custom software. We're proud of what we do. Because we know how much judgement we contribute as people, passionate about our craft that goes into the parts that never see the light of day. What not to build, when to stop, which of the three things the client asked for is the one they actually need right now. There's a lot of us there. And the only reason we know any of this is that we went and sat with the people doing the work.
Right now, I suspect someone out there could be looking at something we’ve done and thinking that with a weekend and a good prompt, they could do it themselves.
Annoyingly, despite how I feel, they are not being stupid. They're doing to us precisely what the artist did to their lawyer, and what I very nearly did to myself.
They can't see the craft because they've never been inside it. From the outside it just looks like typing and unnecessary tech jargon, and Claude seems pretty good at doing that.
Just like the artist, I can’t sit here and claim that our work has soul in it and the lawyer's doesn't.
You can't get there on your own
What I can say is AI hasn't made my work any more enjoyable or more meaningful. The tools have simply made it easier for me to execute more quickly,and sometimes that means moving more quickly in the wrong direction.
There's an old Latin phrase, festina lente: make haste slowly. The key is to balance urgency with careful consideration, so you can move quickly without making mistakes. I love this as an idea, but I have to acknowledge that I am also pretty bad at practicing on my own. It's been made worse by these new tools making it easier for me to start the next thing before I've really finished thinking about the first one.
What I’ve learnt so far is that people make the difference between what looks like a good outcome, and what is actually good. Talking about what I'd done, why I did it, who it was for, and what I got wrong. That's what I see as honing my craft, and as grounding the value I add in reality.
Nobody talked me out of retraining as an electrician with a well-reasoned argument either. Instead, my fiancée asked me a question I hadn't asked myself, what do people actually come to you for?".
I couldn't see my own blind spot from the inside, just like you can't always see the craft in someone else's field of expertise from the outside. I think both are resolved the same way, and it isn't by thinking harder on your own - it’s through community, discussion, and asking for someone else's perspective, and input
So whose lunch are you relaxed about being eaten by AI? Go and ask them what the hard part of their job is. It takes about ten minutes, they will tell you and likely humble you, deeply.