Domain expertise has always been the real moat • Hacker News Board

Re: Domain expertise has always been the real moat

#271

Post by svara » Sun, May 31, 2026, 5:48 AM UTC

simonw wrote: post #128

toastmaster11 wrote: post #62

How much pontificating needs to be done before people acknowledge nobody has any idea what to do with AI on an individual level? First being good developer and learning how to use AI was sufficient, next it was being able to design architecture, then it was “taste” that made all the difference and now being an expert in the domain is the only thing that matters really. Until AI is basically in a stable, predictable,…

An idea that's beginning to solidify for me is that AI tools make software development harder . It's harder because they dramatically raise the bar for what's possible to do. An individual developer can take on significantly more challenging projects now, because the ultimate constraint has always been time and AI can help you get more done in the time available. But the stuff you can get done with that time is a who…

I had the same thought recently, I've had it happen to myself.

I've been working on something relatively large and greenfield recently.

A big chunk of my time is spent thinking about the hard parts. The raw information processing rate needed to keep up with the state of the project is high.

It feels almost like mental athleticism, whereas coding used to be a rather chill activity.

Re: Domain expertise has always been the real moat

#272

Post by baxtr » Sun, May 31, 2026, 5:51 AM UTC

hibikir wrote: post #232

simonw wrote: post #128

Earlier quoted context omitted.

An idea that's beginning to solidify for me is that AI tools make software development harder . It's harder because they dramatically raise the bar for what's possible to do. An individual developer can take on significantly more challenging projects now, because the ultimate constraint has always been time and AI can help you get more done in the time available. But the stuff you can get done with that time is a who…

From that perspective, development has always been harder since I started. I left college with a copy of K&R and remembering courses that applied to real life immediately, because data structures and such were just what we had. In my first job, I ended up writing a code generator to help serialize a large number of data structures, straight from a compiler design class.. which right now you don't need to know a thing…

So in essence: it got harder because the easier parts have been solved for?

Re: Domain expertise has always been the real moat

#273

Post by ElProlactin » Sun, May 31, 2026, 5:52 AM UTC

simonw wrote: post #266

ElProlactin wrote: post #263

Earlier quoted context omitted.

> Try asking a non-developer in your life what their dream software would be for their work, or their hobby. If they don't have what Nilay Patel calls "software brain" I'd be surprised if they came up with something actionable. I've posted this in numerous comments because I think it bears repeating: there are tech-savvy non-developers who are actually building and shipping stuff with AI. I personally know a few who…

> there are tech-savvy non-developers who are actually building and shipping stuff with AI I absolutely believe that. I think those are people with "software brain" who are on their way to becoming real developers. By the point they can write apps that are secure and scale... they'll have learned enough about software development to be employable as software developers. They'll be part of a new breed of developer who…

> I absolutely believe that. I think those are people with "software brain" who are on their way to becoming real developers.

In my opinion, this is a software developer-centric way of thinking that reminds me of the saying, "if all you have is a hammer, everything is a nail."

Here's an alternative perspective:

For billions of people, technology products are an integral part of daily life. As a result, lots of people have an interest in building technology products, particularly software. Thanks to AI, you no longer need to be a "real developer" to build software. You can learn enough to build things that are commercially viable without seeking to be employed as a developer.

> If we want to stay employed, we need to be notably better at building software than they are.

While I don't believe that the market for developers will shrink to 0, unfortunately, I think this type of comment reflects the fear, existential angst and denial that has overtaken many people in this industry.

The reality is that developers are no different than all the displaced workers who came before them. One day you had a job that seemed secure and capable of providing for a comfortable life and the next you were facing the prospect of diminished wages and unemployment because the world simply needs fewer people with your skills and there's no way around the secular trend.

The sad irony is that when software was eating the world and new CompSci grads could take their pick of $150,000+ job offers before ever writing a line of production code, a lot of people in the industry had a smug "tough luck" attitude towards all the workers being displaced by the tech boom. Now it's their turn.

Re: Domain expertise has always been the real moat

#274

Post by necovek » Sun, May 31, 2026, 5:53 AM UTC

zarzavat wrote: post #240

simonw wrote: post #128

Earlier quoted context omitted.

An idea that's beginning to solidify for me is that AI tools make software development harder . It's harder because they dramatically raise the bar for what's possible to do. An individual developer can take on significantly more challenging projects now, because the ultimate constraint has always been time and AI can help you get more done in the time available. But the stuff you can get done with that time is a who…

I worry this is looking at where the ball is now instead of where it's going. The recent disproof of an Erdos conjecture should put to rest the idea that LLMs will reach a skill ceiling before they reach superintelligence. I believe we are headed for a world of superintelligent AI where LLMs are much better at logical thinking than humans, the same way that chess engines are much better at chess than humans. In that…

If building software (and even programming, as the basis for it) was just an expression of logical thinking, we would have cornered it long time ago IMO.

But then again, logic is really a lot more discrete and well defined and easily expressed with traditional computing than LLMs are (which are probabilistic systems instead and as such require large knowledge bases).

We can observe that at a couple hundred billion parameters they behave similarly to a point (in the sense that they can produce similar results), but the challenge is really in understanding the problem's multifaceted structure and competing needs and priorities.

Re: Domain expertise has always been the real moat

#275

Post by emodendroket » Sun, May 31, 2026, 5:56 AM UTC

ElProlactin wrote: post #263

simonw wrote: post #249

Earlier quoted context omitted.

I don't buy that at all. It doesn't matter how great the LLMs get, the act of creating software using them will still require a great deal of skill. Most people just don't think in terms of software. Try asking a non-developer in your life what their dream software would be for their work, or their hobby. If they don't have what Nilay Patel calls "software brain" I'd be surprised if they came up with something action…

> Try asking a non-developer in your life what their dream software would be for their work, or their hobby. If they don't have what Nilay Patel calls "software brain" I'd be surprised if they came up with something actionable. I've posted this in numerous comments because I think it bears repeating: there are tech-savvy non-developers who are actually building and shipping stuff with AI. I personally know a few who…

> I've posted this in numerous comments because I think it bears repeating: there are tech-savvy non-developers who are actually building and shipping stuff with AI. I personally know a few who have been successful in acquiring initial customers.

I mean, sure. But there have always been people teaching themselves to program too. In the end it's a pretty small population.

Re: Domain expertise has always been the real moat

#276

Post by modelhub » Sun, May 31, 2026, 5:56 AM UTC

simonw wrote: post #227

dogcomplex wrote: post #215

Earlier quoted context omitted.

That's a roundabout way of saying it makes software development easier. Perhaps even a 180. But yes - once it's that easy you have to step up your ambitions.

Maybe it makes software development easier, but a career as a professional software engineer harder.

[dead]

Re: Domain expertise has always been the real moat

#277

Post by toastmaster11 » Sun, May 31, 2026, 5:57 AM UTC

locknitpicker wrote: post #260

toastmaster11 wrote: post #239

Earlier quoted context omitted.

If the vibe coders are right I would give at least 90% of the credit to all the well made libraries, rules, and best practices that developers have built up over the decades. That’s what is embedded into LLMs and what might be the saving grace of slop code bases in the future.

> (...) rules, and best practices that developers have built up over the decades. It seems you are not understanding that the reason all these "rules, best practices" had to be created in the first place was the fact that your average old times developer was churning out shit code and weaving spaghetti just as hard as today's vibecoders. Those "rules, best practices" spawned from the same evolutionary pressure as tod…

All that stuff you mentioned is derived from a core set of principles established by decades of software best practices applied to a new means of generating code. Like quite literally those instruction files/skills essentially just reiterate the practices themselves.

To your last paragraph, I never say that nor do I imply it. I find that as a pretty disingenuous interpretation of what I said actually. The practices I mentioned were derived from hard learned lessons and designed as a means of mitigating the human tendency to write bad code.

Re: Domain expertise has always been the real moat

#278

Post by emodendroket » Sun, May 31, 2026, 5:58 AM UTC

ElProlactin wrote: post #273

simonw wrote: post #266

Earlier quoted context omitted.

> there are tech-savvy non-developers who are actually building and shipping stuff with AI I absolutely believe that. I think those are people with "software brain" who are on their way to becoming real developers. By the point they can write apps that are secure and scale... they'll have learned enough about software development to be employable as software developers. They'll be part of a new breed of developer who…

> I absolutely believe that. I think those are people with "software brain" who are on their way to becoming real developers. In my opinion, this is a software developer-centric way of thinking that reminds me of the saying, "if all you have is a hammer, everything is a nail." Here's an alternative perspective: For billions of people, technology products are an integral part of daily life. As a result, lots of people…

Maybe the tools are going to get to a point where this isn't true but today even with Claude Code at whatever at hand you're going to have to learn enough about software to basically be a developer in the traditional sense to deliver a multi-tenant application that has to deal with high TPS or whatever. At least at present you're positing there's no need for carpenters because the home gamer can knock together a table or birdhouse at home.

Re: Domain expertise has always been the real moat

#279

Post by svara » Sun, May 31, 2026, 6:00 AM UTC

zarzavat wrote: post #240

simonw wrote: post #128

Earlier quoted context omitted.

An idea that's beginning to solidify for me is that AI tools make software development harder . It's harder because they dramatically raise the bar for what's possible to do. An individual developer can take on significantly more challenging projects now, because the ultimate constraint has always been time and AI can help you get more done in the time available. But the stuff you can get done with that time is a who…

I worry this is looking at where the ball is now instead of where it's going. The recent disproof of an Erdos conjecture should put to rest the idea that LLMs will reach a skill ceiling before they reach superintelligence. I believe we are headed for a world of superintelligent AI where LLMs are much better at logical thinking than humans, the same way that chess engines are much better at chess than humans. In that…

Chess and proofs only work as comparisons to the extent that you can find parts of your job that share their key property: A solution is sought to a problem that can be stated with relatively little information.

What prompt would someone have used to get a superhuman coding agent to output the Linux kernel or GTA5?

Before you accuse me of moving the goalposts, that's not my point: The examples are there to help think about what humans would still need to do to build complex projects even if the coding itself was perfectly reliable.

Both the Linux kernel and GTA5 contain a large amount of incompressible information; humans thought long and hard about how to design them, i.e. about what that thing they were building was even supposed to be.

Re: Domain expertise has always been the real moat

#280

Post by fragmede » Sun, May 31, 2026, 6:02 AM UTC

simonw wrote: post #266

ElProlactin wrote: post #263

Earlier quoted context omitted.

> Try asking a non-developer in your life what their dream software would be for their work, or their hobby. If they don't have what Nilay Patel calls "software brain" I'd be surprised if they came up with something actionable. I've posted this in numerous comments because I think it bears repeating: there are tech-savvy non-developers who are actually building and shipping stuff with AI. I personally know a few who…

> there are tech-savvy non-developers who are actually building and shipping stuff with AI I absolutely believe that. I think those are people with "software brain" who are on their way to becoming real developers. By the point they can write apps that are secure and scale... they'll have learned enough about software development to be employable as software developers. They'll be part of a new breed of developer who…

Nevermind syntax, what's a variable? function? class? What's the difference between int/float/boolean string? Nevermind more advanced concepts like O(1) vs O(n). But when the vibe coder just needs to prompt "the page loads really slowly. plz fix" and the LLM can go in, add an index to the right SQL table, add a limit and pagination, so what if I can tell you the difference between PostgreSQL's dialect of SQL vs MySQL, and what the difference is in row types supported. I can describe what happens when you type Google.com into your webbrowser to an inane level of detail off the top of my head, but when the LLM can do an even more through version, I mean, I can pat myself on the back and be smug that I know most of that innately, but what is it really worth?

About a decade back, we, as an industry were collectively learning how to make apps webscale, and oh the blog posts about not using a database as a queue. But the LLMs have ingested all of them. I've only read the ones I came across, and of course my professional experience being part of teams implementing that at various companies. So I've got that going for me, but when the Vibe-platform-dev just has to tell the LLM "hey, when the user hits the send message button, it's slow. /goal make messages fast", and the LLM grinds for hours overnight switching the entire system over to a pub sub event driven architecture and the vibe-platform-dev doesn't even know what pubsub stands for or that they're using one unless they go back and read the transcript. I don't think there's as much of a domain expertise moat for as long as we're hoping.