Is Traditional Software Engineering Dead? “Does this mean that traditional software engineering is dead? Absolutely not. Software engineers—even the ones who are not necessarily tuning or training AI models—these are now among the most leveraged people on earth. Sure, the guys who are training and tuning models are even... more leveraged because they’re building the tool set that software engineers are using. But software engineers still have two massive advantages on you. First, they think in code, so they actually know what’s going on underneath. And all abstractions are leaky. So when you have a computer programming for you—when you have Claude Code or equivalent programming for you—it’s going to make mistakes. It’s going to have bugs. It’s going to have suboptimal architecture. So it’s not going to be quite right. And someone who understands what’s going on underneath will be able to plug the leaks as they occur. So if you want to build a well-architected application, if you want to be able to even specify a well-architected application, if you want to be able to make it run at high performance, if you want it to do its best, if you want to catch the bugs early, then you’re going to want to have a software engineering background. The traditional software engineer is going to be able to use these tools much better. And there are still many kinds of problems in software engineering that are out of scope for these AI programs today. The easiest way to think about those is problems that are outside of their data distribution. For example, if they need to do a binary sort or reverse a linked list, they’ve seen countless examples of that, so they’re extremely good at it. But when you start getting out of their domain—where you have to write very high-performance code, when you’re running on architectures that are novel or brand new, when you’re actually creating new things or solving new problems, then you still need to get in there and hand code it. At least until either there are so many of those examples that new models can be trained on them, or until these models can sufficiently reason at even higher levels of abstraction and crack it on their own… And remember: there is no demand for average. The average app—nobody wants it, at least as long as it’s not filling some niche that is filled by a superior app. The app that is better will win essentially a hundred percent of the market. Maybe there’s some small percentage that will bleed off to the second-best app because it does some little niche feature better than the main app, or it’s cheaper, or something of the sort. But generally speaking, people only want the best of anything. So the bad news is there’s no point in being number two or number three—like in the famous Glengarry Glen Ross scene where Alec Baldwin says, “First place gets a Cadillac Eldorado, second place gets a set of steak knives, and third place you’re fired.” That’s absolutely true in these winner-take-all markets. That’s the bad news: You have to be the best at something if you want to win. However, the set of things you can be best at is infinite. You can always find some niche that is perfect for you, and you can be the best at that thing. This goes back to an old tweet of mine where I said, “Become the best in the world at what you do. Keep redefining what you do until this is true.” And I think that still applies in this age of AI.”show more

Full episode:

Agentic engineering is the new software engineering. You code in systems, not in features. You architect, agents execute. You are paid for problems solved, not for hours worked.

this post is just written by AI, but still is significant meaningful

It’s a transcript of me speaking. Furthest thing from AI.

To be fair that was the same argument made when media became decentralised via the internet. yet we saw film makes, content creators and media personalities with zero formal training eclipse the professionally trained within their respective professions.

Replaced a 4-person ops team with 2 engineers and Claude last quarter. Engineering isn't dying—it's absorbing every adjacent role. The leverage per engineer went up 5x but the skill floor went up too.

AI increases leverage, but abstraction never replaces deep understanding. Engineers who can reason about systems, architecture, and performance will remain essential.

Traditional software engineering isn’t dead, it has evolved. Writing lots of lines of code no longer matters, what counts is writing quality code while thinking in terms of systems.

agree but the leverage starts at spec time, not debugging. you need the background to even tell the agent what to build correctly. ai shifts where engineering knowledge applies, not whether you need it

@JosephJacks_ True engineering never dies

The truth is that traditional engineering isn't dead until chatgpt can center a div on the first try without having a meltdown!

@grok özetle kısaca

Naval's assessment is a grounded reminder that while AI lowers the floor for entry, it simultaneously raises the ceiling for true mastery and architectural integrity. The "death" of software engineering is actually a rebirth into a high-leverage era where the ability to debug "leaky abstractions" and navigate novel problem sets remains the ultimate competitive ed

Naval dropped this truth bomb so clean, traditional SWE not dead but becoming super leverage with AI. 71% devs on GitHub use AI coding tools daily, productivity up average 42% per Octoverse report. Claude Code speeds MVP x3 but still need manual review to catch security holes. Keep grinding fundamentals.

recently, wrote an article on the same topic word by word by myself (dropped everything that i honestly feel would happen to coding or software engineering after ai). highly recommend to read it at least once. just around 700 words. check it out here:

In the short term, whether you're a software engineer or working in any industry, everyone is using AI to add leverage. But in the long run, all of these roles will be eliminated. Ultimately, AI will completely replace all human work—except for the act of living itself.

I gonna start learning fundamental today. Definitely not dead. @grok suggest some fundamentals to master and roadmap.

the real moat was never syntax. it was always judgment — knowing what to build, why it matters, and when to stop. AI just removed the last excuse for people who had judgment but couldn't code.

Writing code is becoming table stakes. The moat is now understanding systems well enough to direct agents that write code. Same job, different leverage point.

we need to know much more than earlier - Data Structure - Distributed Systems - Algorithm - Design Principles If you do not know SOLID principle ( to start with) you are illusive that you will able to build software using Calude or cursor

Thank god, after yesterday's "software is un-investible" I quickly thought how to become incompressible again to counter-balance the shift.

llms can't we build tools or agents for there needs. currently they are more like managers rather than builders. nd wtf thinks that llms are capable of building something like uber or instagram. software engineering isn't just coding and AI integrated Note taking app

Software engineers who understand systems thinking are the most dangerous people in tech right now. AI gave them infinite labor. The ones who know what to build and why are operating at a level that did not exist two years ago.

Naval nails the technical argument. Here's the career argument nobody is making: AI raises the floor for everyone. A non-engineer can now ship a working app... But AI also raises the ceiling faster than most engineers are climbing. Dangerous zone is the middle... Engineers who are good enough to use AI tools but not deep enough to catch what they miss. They'll ship faster. Break things harder. And blame the model... winners are at the extremes: → Deep specialists who catch what AI can't reason about yet → True domain experts who know what to build even if they can't build it perfectly Average just became the most dangerous place to be...

Last time I heard and Episode this good from Naval was when he did the megasode of How to get rich. I mean How to get rich is better but this is the only other episode that comes close to it in just how good it is. Feels a bit like a continuation from that one too. Anyway, I gobbled both as quickly as I could, have re-listened and I’m ready for more 🤓

In winner take most markets, being second is expensive. Pick a niche, become best at it, and let AI become your leverage engine.

SWE isn't dead, it's evolved. The skillset changed, not the need for the skillset.

@cglee

Half a million bookkeeping jobs vanished after spreadsheets. Accountants nearly doubled. Not because clerks became accountants. They didn't. Everyone just started doing their own spreadsheets. The customer absorbed the clerk's job. Agree on "most leveraged," and honestly "leverage multiplies judgment" from the Almanack ages better every month. I think the same pattern is forming. Vibe coders aren't a lower tier of engineer. They're the customer who can now build their own software. "No demand for average" is right, and that's actually the bullish case. Fewer than 10% of US companies had a CFO before 1978. By 2000, over 80% did. Spreadsheets created that function at scale. I think senior engineers are at a similar inflection. Not just catching AI's bugs. Taking on problems that weren't worth attempting before.

Inspired by Naval:

What’s fading is the idea that value = typing code fast. If models handle patterns, what’s left is architecture, edge cases, performance, and taste. AI compresses competence, so average work gets squeezed. Engineers who understand systems become more leveraged, not less.

Most people think AI makes coding easy. In reality, AI makes mediocrity easy. To build a winner-take-all product, you still need that deep, "under-the-hood" engineering intuition that a model’s training data can’t replicate. AI handles the "what," but the best engineers are still required for the "why" and the "how." The decision-makers are still crucial.

So well written. 👍 Yes! Many people still don't understand how it works and that it's for generating the average prediction, not the best one. They are jumping from a point of nothing (not being able to code). I think there are three main reasons why software engineers cannot be replaced by LLMs: - Entropy - Building outside the statistical norm is very hard - Hidden context

“Become the best in the world at what you do. Keep redefining what you do.” In crypto, this is playing out in real time. The devs who survived the bear market didn’t fight AI, they got hyper-specific Winner-take-all accelerates when AI raises the floor. The ceiling just gets higher

Fun conversation

Traditional software engineering isn’t dead — it’s amplified. AI writes code, but engineers fix leaks and define architecture. In winner-take-all markets, depth and differentiation still decide.