A.I. As I See It
We're fighting the same feeling that got us into development in the first place. And the same pitfalls we've gone through ourselves. And dealing with a paradoxical situation.
Do you remember when you were just starting out as a developer? Remember that feeling you had when you FINALLY got some code to do The Thing ™️ you wanted it to do? That "drunk with power" moment that had you believing you could build anything? That feeling that made you think "this is what I want to do for a living"?
That's the exact same feeling A.I. is giving to a lot of non-coders right now.
But There is a Difference
When we had that feeling, it's (likely) because we finally understood the code that we wrote. It's a necessity, most of the time, in order to get code working.
"My code doesn't work and I don't know why - my code does work and I don't know why" jokes aside.
When a non-coder gets ChatGPT or whatever to spit out some code, I would venture to say that, most of the time, they don't understand it. It was just handed to them with (presumably) a set of instructions on how to run it. They got the baby without the labor. In my experience, a lot of the time they even require a dev to run it for them (if they're trying to integrate it into existing code), fix it if it doesn't actually work, deploy it and all of that.
But, when it does result in something that "works". It's still that same feeling.
What It Could Mean
In my opinion, it means that, just like is wont to happen with new "real" developers, the Dunning–Kruger effect comes into play. And I swear, I don't mean that as a slight at all, to anyone. It's simply a common consequence a lot of us go through as we learn new things, myself included.
So that sense of empowerment, combined with (perhaps) an overestimation of confidence, also combined with the notorious hype cycles that tech goes through, can lead to some bad opinions and decisions.
More and more "simple" code could start being produced by the non-coders of an organization as a means to not "bother" an actual on-staff developer. Before long, the question could start being asked "do we need this many developers if A.I. can do so much of it for us?"
How We Earn Our Keep
Because we, as developers, understand code, we should be able to build some wild shit with A.I. We should be able to write code that helps train and tune models to do things casual users (and probably even "power users") of A.I. simply can't.
Everyone wants a chatbot, apparently. This, in my opinion, is the most boring and generic use of the technology. But they can be quick to spin up and seen as a somewhat easy win. They are also (again, imo) a consequence of a lack of understanding of the technology, a lack of imagination, or some combination of the two. And, as I've said before:
Engineers have a unique perspective on products because we actually build them. We know the technical limitations. We know the caveats. We know what's actually possible.
The way I see developers protecting themselves from A.I. "taking their job" or whatever, is to understand the technology and show people the kind of stuff it could/"should" be used for. Simple development tasks are quickly becoming table stakes for A.I. so the way we show our value is by doing the "Holy shit!"-style stuff with it.
The Caveat
The problem with more deeply understanding the technology is knowing that, in it's current (and maybe forever) state, it is inherently an unethical technology. Bias, accountability, copyright, environmental issues. The list goes on and on.
And a seeming theme in tech is to ignore these issues. Whether it's ignorance (willful or not), an expectation that "it's not my problem to solve", or a ride on yet-another hype train in tech, people are rushing to include (and advocate for) A.I. solutions and features in pretty much every product they produce lately, mostly in search of a buck.
I Don't Know What to Do
It's a number of different catch-22s for developers. We're "fighting" with people experiencing the same feeling of empowerment that got us into this industry to begin with. For self-preservation, we have to understand and demonstrate how the technology can be used "in the right hands". But, for a socially and morally conscious developer, are there really any "right hands"?
It's an interesting time. "Interesting" isn't the right word, but it's the first word that comes to mind.
Webmentions
Likes
Reposts
Replies
Mentions