
Should Humans Be Writing Code Anymore?
My whole career has been about augmenting what people can do, and right now that means figuring out AI. I'm on my own journey with this, including the parts I haven't figured out yet. Here's where I'm at, and I want to know where you are.
My whole career, and especially my time at Agility, has been about augmenting what people can do. What can you do with your job, and how can we help you do more with less?
Right now that means figuring out AI, and I want to be upfront about something. I'm on my own journey with this. I haven't arrived anywhere. I'm somewhere in the middle of it, probably the same as you.
Here's where I'm at.
If you write code
If you're a developer, why are you spending your time typing? Get Claude Code to do that, or another coding agent, while you're having the ideas that only you can have. Then spend the time you get back reviewing the outcome. Is it secure? Is it performant? Is it actually a well rounded solution?
What I find with teams is they get it kind of working, and then they forget about everything else that matters just as much. Does it perform? Does it work in the wild? Or did you just get the minimum thing going? That's usually the difference between junior, intermediate, senior, and a really advanced lead developer.
And here's the part of my journey I have NOT figured out. You can't review code if you don't know how to write it. So when I say don't write code anymore, I still think people need to write it in order to learn it. I don't know how to square those two things yet. It's a tough one and I'm not going to pretend I have the answer.
If you work with content
The other side is the creative side. Content teams, marketing teams. Still technical in its own way, but not coding.
Most of that work starts unstructured. A Google Doc, a Word document, notes from a brainstorm. Then somebody has to get it into the CMS, into the right fields, in the right locales, following the brand guidelines, and click publish. Out of a team of a dozen people involved in the content process, very few of them actually know what's going on in the CMS.
Do we need to burn human hours on that? A lot of people love the creative part, and I think we should be spending more time there. Less time on "gee, what should the slug be based on the title?" AI is really good at figuring that out.
So I've spent a lot of time helping folks build skills in Claude to take unstructured content, Word docs, Google docs, images, whatever, and push it into the CMS automatically. We call it an AI content pipeline, for lack of a better term.
I don't believe that's taking away jobs. I know there are companies out there saying they can pad their margins by cutting staff and bringing on AI. I'm not in favor of that. I'm in favor of asking teams: here's your capability set today, how could you augment it in a way that actually makes sense?
Where I'm at right now
We're building a website for ourselves at the moment, internally, without an outside agency. The current agilitycms.com was built by one of our partners and we paid real money for it. This time we're using our own team, augmented with AI.
Instead of months of work and two or three design revisions, we're getting dozens of revisions in a couple of days. Brainstorm, the agents work, review it the next day. All those things our marketing team has wanted to try for years and never had room for, we finally have room for.
That's my journey so far, unfinished parts and all.
So where are you at in yours? What's working, what isn't, and what do you think this looks like a few years from now? I'd like to hear it. Let's have that conversation.