Our efforts, in part, define us

[dead]

> For myself, in the last 10 years, my work of writing code has largely defined what I do with my working time. Now I experience large swaths of that work being created and done by AI (sometimes amazingly well, sometimes poorly), and I find myself thinking of the photographer above.

Which is ironic, because smartphone photography is astonishingly average to someone who actually understands photography. Phones try to handle every scenario for the user. It has been optimized for bright outdoors, low light conditions, even specific stuff like sunsets. So most people never realize how difficult it really is to shoot in bright daylight with the light source behind the subject. Only when you pick up a traditional camera do you realize that’s one of the worst conditions to take a photo in.

Maybe coding is headed toward the same place. AI coding will smooth out complexity for the majority of programmers, but at the same time, we’ll also see a lot of programmers building things that don’t actually need to be built, or not realizing the limits of their solutions, or just hung in the narrow space of what AI can do for them.

Having done photography with Sony-a6000, I’ve found that it made me more skilled with a smartphone camera. So, I am still optimistic that knowledge gained through deliberate effort has value, even in a world that increasingly prizes “effortlessness” in everything. But only time will tell how much that belief really holds up.

I went through something like that transition the first time I moved into management.

It took me by surprise how much I tied what I did to who I was, and how much it bothered me that I could no longer so easily articulate what I'd done during a day at work. It was no longer I made this, I made that, I added these features, I fixed those bugs. It was all relational. It was all meetings and conversations. It felt like anyone could be doing it. So I felt like I had no value. Like I was a spare wheel. Three or four years and in all that time I never made my peace with it.

So burned out and dejected[0] I went back to software development and performance consulting. I did that for about 3 years.

The second time I went into management was different. Maybe I was just older. Maybe I'd done a bit more reading and thinking about management and leadership. I certainly knew what I liked and disliked, and what I admired in other leaders. Because of the consulting I'd seen a variety of leaders at work in lots of different contexts, with different challenges, different styles, and different personalities. At any rate, I had a different and broader view of what good looked like.

I also had a clear and strong remit and room to operate: build a new product, build and lead a new team, open an office, rearchitect and rebuild our platform, become CTO.

But the critical change was that I now saw my value in enabling the success of others - our team, the teams that relied on our platform and products, our customers and clients - and so I loved my job (for the most part).

This doesn't directly relate to AI but does show that it's possible to go through a transition where you lose a big part - or all - of what you thought gave you value, status, and satisfaction in a role, and grow into something new and valuable as a result of that.

In my case that took for-fucking-ever - I've always been something of a late developer[1] - and many people reading this have and will adapt much more quickly. I'll say this: it's much easier to let go when you have a stronger idea of what "next" looks like[2], so figure out what that is. Software development is changing, for all of us, and even now nobody really knows what that means, but it doesn't mean there aren't plenty of opportunities.

It's also worth bearing in mind that a lot of the more extreme viewpoints published online regarding AI and software development are no better than noise: they're just clickbait designed to generate reaction and engagement. Their real informational value is close to zero.

That is obviously not the case with this piece, which describes the author's personal experience in our new AI enabled world.

And I get their point about the craft. You really do feel that. But what you have to realise is the craft doesn't go away: it moves to a different level of abstraction. What can you (what can we) do now that you couldn't do before? What do the wrinkles and contours of that, where your (our) finesse will be needed, actually look like?

This all reads more like a pile of scattered thoughts than I'd intended, but hopefully there's some value in there for some of you.

[0] Not just for these reasons: there were other things - including one or two other major contributors - going on that I don't want to talk about here, but my lack of comfort with the nature of my role was certainly a somewhat chunky factor. Just don't overread or assume too much from what I'm saying.

[1] No pun intended - nobody should read anything delivery related into this comment.

[2] This touches on some of those issues I didn't want to talk about. Within that organisational context I'd reached a ceiling and no-one, including me, was grown up enough to be real about that. So I kept trying to push on and break through and, well, I guess we all wasted a lot of eachother's time although, from my side, I don't feel too bad about it.

> never looked back

Obviously there are very many Amish people who leave their communities to join the modern world - according to various sources this attrition seems to be between 15-25%. And even amongst those who stay, it's likely that a central reason is the fear of excommunication/shunning, rather than an inherent desire to reject modern technology.

Don't know your area but llms are not taking many software dev jobs around here. Its like saying clueless juniors who sometimes produce good snippets of code and sometimes struggle to distinguish left from right are taking jobs of some seniors... not happening en masse anytime soon. Some other types of jobs sure.