Software engineering may no longer be a lifetime career

> But lifting heavy objects puts long-term wear on your back and joints, making you less effective over time.

Odd argument considering you could also say lifting heavy objects makes you more effective over time, as it makes you stronger.

Also not a fan of the assumption that software is a lucrative career to the degree implied here, similar to pro atheletes. Perhaps in certain parts of the US, but not elsewhere, where it's just a regular decently paying job. This also changes the replacement economics depending on where the true cost of LLM's end up landing. There's plenty of jobs that could be replaced today if we had infinite access to capital, but are still at no risk.

Perhaps it wouldn't be the worst thing in the world if US engineer salaries deflated to be more in line with the actual difficulty of the work relative to other professions.

I wonder if the author looked at Civil Engineers and put himself in their shoes.

This is like a Civil Engineer looking at the highways & bridges and saying - yeah no more work for me.

'Engineering' will always be a career - specially in a bit driven world. Just cause LLMs can spit words doesn't mean the LLMs are doing engineering.

> On the one hand, moving from assembly language to C made programmers less effective in some ways and more effective in others. On the other hand, the transition from writing code by hand to using AI is arguably a bigger shift,

I think that's a bad comparison, mainly for two reasons:

We still have people dealing with assembly for building the compilers. We don't quite have that for AI - which became very clear with the claude code leak: We just have people trying to get the correct behaviour, but without understanding why.

The second reason has more practical implications: I generally rely on my compiler to be deterministic, and not introduce security issues (which they sometimes do, by optimising away safeguards placed by the programmer - but that is relatively rare, and we can trace and fix that). But generally as a developer I can rely on my compiler to produce machine or byte code I don't have to think about.

The same is not true for AI - the regularly produce insecure code. This can partially be mitigated by having another AI review it - but it's not a proper solution. Until we get proper AI which can actually understand things we need somebody in the loop who can understand the code.

i think using AI is going to transform our definition of programming into something more abstract, but it's still going to require human.

We're going to start thinking in terms of program general properties ( termination conditions, invariants, symetries, etc), and code quality indicators, to drive our work into certain directions, depending on the project.

Then in 20 years we're going to consider the task of actually typing code something of really low value and uninteresting.

I can feel it's already happening to me after just a few months of vibe coding.