Good point on code but data is actually scarier. A broken function throws an error. A broken segment table just looks fine. Wrong assumptions quietly bake into decisions for months before anyone notices.
I see this problem nearly every day. In my view, engineers using AI and pushing bad code is certainly a problem. It's also a problem for engineers that non-engineers now write code and build apps on their own, increasing the pressure on engineering teams and broader IT groups to review/adapt/etc.
But this article feels a bit reductive. "You will know nothing."
I see lot of folks - engineers and others - performing meaningful work to understand what they're doing with AI and prosecute the answer to "what problem do we have and what do we understand about it"
Yet, if you build the perfect app with engineer expertise, or with AI, or both, and still pipe garbage data through it, you haven't solved much.
Great article. What I think techie people tend to not fully understand or internalize is that no one outside of about 3% of the population gives a shit about how the code works. No one. They want the end result. You’re stuck thinking that your craft matters or that there’s an inherent nobility in understanding and modeling complex systems but until you can show how that knowledge is still necessary to drive the result you will lose out to AI.
Sounds harsh but I guess true, I'd never really thought of it that way. But isn't due process a way to potentiate the end result (even if processes slow us down in the short term)? So the counter-question would be why should this change just because of AI?
To push back on myself I'd say well AI is getting pretty good at putting processes in place too! So maybe in the future they own process and product and we own... nothing?
I'm literally dealing with a bit right now as a junior is tab-accepting claude code without actually knowing the simple fundamentals. When asked:
> "How do you test Swift code on the Mac?"
I looked at him, unsure of whether or not he was trolling me. It made me so uncomfortable that I thought for the briefest of moments that I didn't know this answer myself even though I've been programming in Swift since it was first released, 12 years ago.
> "Uh. Xcode..."
I think I saw death.
This is bad. Like, really, really bad. I'm not sure what to do about it.
Good point on code but data is actually scarier. A broken function throws an error. A broken segment table just looks fine. Wrong assumptions quietly bake into decisions for months before anyone notices.
Underrated view Ahmed.
I see this problem nearly every day. In my view, engineers using AI and pushing bad code is certainly a problem. It's also a problem for engineers that non-engineers now write code and build apps on their own, increasing the pressure on engineering teams and broader IT groups to review/adapt/etc.
But this article feels a bit reductive. "You will know nothing."
I see lot of folks - engineers and others - performing meaningful work to understand what they're doing with AI and prosecute the answer to "what problem do we have and what do we understand about it"
Yet, if you build the perfect app with engineer expertise, or with AI, or both, and still pipe garbage data through it, you haven't solved much.
Great article. What I think techie people tend to not fully understand or internalize is that no one outside of about 3% of the population gives a shit about how the code works. No one. They want the end result. You’re stuck thinking that your craft matters or that there’s an inherent nobility in understanding and modeling complex systems but until you can show how that knowledge is still necessary to drive the result you will lose out to AI.
Sounds harsh but I guess true, I'd never really thought of it that way. But isn't due process a way to potentiate the end result (even if processes slow us down in the short term)? So the counter-question would be why should this change just because of AI?
To push back on myself I'd say well AI is getting pretty good at putting processes in place too! So maybe in the future they own process and product and we own... nothing?
Super interesting future to think about
I'm literally dealing with a bit right now as a junior is tab-accepting claude code without actually knowing the simple fundamentals. When asked:
> "How do you test Swift code on the Mac?"
I looked at him, unsure of whether or not he was trolling me. It made me so uncomfortable that I thought for the briefest of moments that I didn't know this answer myself even though I've been programming in Swift since it was first released, 12 years ago.
> "Uh. Xcode..."
I think I saw death.
This is bad. Like, really, really bad. I'm not sure what to do about it.
"Amazon had to ban junior and mid-level engineers from pushing Al-
generated code without senior approval."
I'm blown away by how reckless Amazon's leadership must have been to ever allow this in the first place.