Comment by gipp
19 hours ago
I see a lot of the same (well thought out) pushback on here whenever these kinds of blind hype articles pop up.
But my biggest objection to this "engineering is over" take is one that I don't see much. Maybe this is just my Big Tech glasses, but I feel like for a large, mature product, if you break down the time and effort required to bring a change to production, the actual writing of code is like... ten, maybe twenty percent of it?
Sure, you can bring "agents" to bear on other parts of the process to some degree or another. But their value to the design and specification process, or to live experiment, analysis, and iteration, is just dramatically less than in the coding process (which is already overstated). And that's without even getting into communication and coordination across the company, which is typically the real limiting factor, and in which heavy LLM usage almost exclusively makes things worse.
Takes like this seem to just have a completely different understanding of what "software development" even means than I do, and I'm not sure how to reconcile it.
To be clear, I think these tools absolutely have a place, and I use them where appropriate and often get value out of them. They're part of the field for good, no question. But this take that it's a replacement for engineering, rather than an engineering power tool, consistently feels like it's coming from a perspective that has never worked on supporting a real product with real users.
I'm not sure you're actually in disagreement with the author of this piece at all.
They didn't say that software engineering is over - they said:
> Software development, as it has been done for decades, is over.
You argue that writing code is 10-20% of the craft. That's the point they are making too! They're framing the rest of it as the "talking", which is now even more important than it was before thanks to the writing-the-code bit being so much cheaper.
When we say generating code is only a small percentage, that does not imply that the rest is just talking. Simon, you were part of a relatively small fast moving project in Django and the news website it powered, with from I understand a pretty small team. Have you worked as part of a team of 10, 20, 100, 1000 engineers? It's different.
"Talking" here doesn't literally mean talking. It means figuring out the scope of the problem, researching solutions, communicating with stakeholders, debating architecture, building exploratory prototypes, breaking down projects - it's all the stuff that isn't writing the code.
I've worked at various sizes of organization. Most notably I joined Eventbrite when they were less than 100 developers and stayed while they grew to around 1,000.
> Software development, as it has been done for decades, is over.
Simon I guess vb-8558's comment inn here is something which is really nice (definitely worth a read) and they mention how much coding has changed from say 1995 to 2005 to 2015 to 2025
Directly copying line from their comment here : For sure, we are going through some big changes, but there is no "as it has been done for decades".
Recently Economic Media made a relevant video about all of this too: How Replacing Developers With AI is Going Horribly Wrong [https://www.youtube.com/watch?v=ts0nH_pSAdM]
My (point?) is that this pure mentality of code is cheap show me the talk is weird/net negative (even if I may talk more than I code) simply because code and coding practices are something that I can learn over my experience and hone in whereas talk itself constitutes to me as non engineers trying to create software and that's all great but not really understanding the limitations (that still exist)
So the point I am trying to make is that I feel as if when the OP mentioned code is 10-20% of the craft, they didn't mean the rest is talk. They meant all the rest are architectural decisions & just everything surrounding the code. Quite frankly, the idea behind Ai/LLM's is to automate that too and convert it into pure text and I feel like the average layman significantly overestimates what AI can and cannot do.
So the whole notion of show me the talk atleast in a more non engineering background as more people try might be net negative not really understanding the tech as is and quite frankly even engineers are having a hard time catching up with all which is happening.
I do feel like that the AI industry just has too many words floating right now. To be honest, I don't want to talk right now, let me use the tool and see how it goes and have a moment of silence. The whole industry is moving faster than the days till average js framework days.
To have a catchy end to my comment: There is just too much talk nowadays. Show me the trust.
I do feel like information has become saturated and we are transitioning from the "information" age to "trust" age. Human connections between businesses and elsewhere matter the most right now more than ever. I wish to support projects which are sustainable and fair driven by passion & then I might be okay with AI use case imo.
Yeah in a lot of ways, my assertion is that @ “Code is cheap” actually means the opposite of what everyone thinks it does. Software Engineer is even more about the practices we’ve been developing over the past 20 or so years, not less
Like Linus’ observation still stands. Show me that the code you provided does exactly what you think it should. It’s easy to prompt a few lines into an LLM, it’s another thing to know exactly the way to safely and effectively change low level code.
Liz Fong-Jones told a story on LinkedIn about this at HoneyComb, she got called out for dropping a bad set of PR’s in a repo, because she didn’t really think about the way the change was presented.
> Takes like this seem to just have a completely different understanding of what "software development" even means than I do, and I'm not sure how to reconcile it.
You're absolutely right about coding being less than 20% of the overall effort. In my experience, 10% is closer to the median. This will get reconciled as companies apply LLMs and track the ROI. Over a single year the argument can be made that "We're still learning how to leverage it." Over multiple years the 100x increase in productivity claims will be busted.
We're still on the upslope of Gartner's hype cycle. I'm curious to see how rapidly we descend into the Trough of Disillusionment.
My recent experience demonstrates this. I had a couple weeks of happily cranking out new code and refactors at high speed with Claude’s help, then a week of what felt like total stagnation, and now I’m back to high velocity again.
What happened in the middle was I didn’t know what I wanted. I hadn’t worked out the right data model for the application yet, so I couldn’t tell Claude what to do. And if you tell it to go ahead and write more code at that point, very bad things will start to happen.
Ive been using LLMs through the web to help with discreet pieces of code and scripts for a while now. I’ve been putting it off (out of fear?) but I finally sat down with Claude Code on the console and an empty directory to see what the fuss was about. Over about a total of 4 hrs and maybe $15 pay as you go it became clear things are drastically different now in web dev. I’m not saying changed for good or bad just things have definitely changed and will never go back.
The book Software Engineering at Google makes a distinction between software engineering and programming. The main difference is that software engineering occurs over a longer time span than programming. In this sense, AI tools can make programming faster, but not necessarily software engineering.
Did you read the article? Author is one of the more thoughtful and least hype guys you'll find when it comes to these things
They're also great for writing design docs, which is another significant time sink for SWEs.