Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Domain expertise has always been the real moat

brethorsting.com

Post reply on HN
Print view View original

561–570 of 592 posts

Re: Domain expertise has always been the real moat

#561

Post by steve_adams_86 »

AndrewKemendo wrote: post #556
steve_adams_86 wrote: post #552

Earlier quoted context omitted.

You’re right. I have a friend who works in landscape architecture and he’s getting up to really cool things involving parametric modelling tools. Stuff he couldn’t easily do with off-the-shelf software is now attainable and he’s going wild with it. You’ll never hear about it here (though it totally belongs here)

I’ve posted about it a lot, it’s posted all the time. It’s pretty much the main topic The issue isn’t that it’s not happening, it’s that people on HN have their identities tied to typing code and can’t deal with the fact that amateurs are trying to compete

I think the competition will be incredible for bringing new idea and new challenges into the field. I’m all for it. I’ve never had to think harder about what’s worth doing and what’s possible, and I love it.

Re: Domain expertise has always been the real moat

#562

Post by emodendroket »

ElProlactin wrote: post #548
emodendroket wrote: post #508

Earlier quoted context omitted.

I bought a book however many years ago with no previous development experience and delivered a Web app people paid for and eventually honed that as an actual career, so I’m just not really seeing what’s a difference in kind here. I also disagree with the “democratization” frame because now developers are spending like $1000 per month on tokens at their jobs, which does the opposite of making things more accessible.

> I bought a book however many years ago with no previous development experience and delivered a Web app people paid for and eventually honed that as an actual career, so I’m just not really seeing what’s a difference in kind here. So you taught yourself how to program. Do you not see that there's a difference between teaching yourself how to program so that you could build a web application and using a new technolog…

> So you taught yourself how to program. Do you not see that there's a difference between teaching yourself how to program so that you could build a web application and using a new technology that allows you to build a web application without having to learn how to program?

Well, I guess where we're not lining up is I don't see that you are going to deliver a serious application that isn't the equivalent of table knocked together in a home workshop without, in practice, "learning how to program." Like, what are you not understanding here? I keep saying this and you keep alternating between accusing me of being deliberately obtuse and accusing me of motivated reasoning, but you aren't actually addressing the argument I am making.

> I'm curious: if you're already working as a developer and your employer values your skills, why do you have to spend anything on tokens? Are you not able to do your job satisfactorily without AI?

Sure, but I can do more stuff faster with AI, which they also value enough to pay for the tools. Is that a serious question? You will find few professional software developers whose employers aren't encouraging AI tool use today.

> The "democratization" has nothing to do with people working as developers. The "democratization" refers to the ability of a growing number of smart, motivated people who don't know how to program to create working software using AI.

Yeah, I understand that, and I reject the framing because 1) it's made the way professional developers work much more capital-intensive and different from the ways hobbyists work in a way that's unattainable for the average hobbyist 2) I don't really agree with the idea that all the barriers are gone and everyone's ready to deliver commercial-grade software without understanding what they're doing. If you think that's the case, then you'll have to conclude most businesses are acting highly irrationally by continuing to pay high wages to employ people with specialized knowledge of software development to operate AI tools rather than just handing it over to your new breed of semiskilled laborers who don't need to know how to program.

Re: Domain expertise has always been the real moat

#563

Post by tevli »

steve_adams_86 wrote: post #36

Such a good example of this I encountered recently: I was on a fishing trip. I asked the charter if he’d want to check out a free app I work on ( https://oceanconnect.ca ) in case it might be useful for his work. I don’t know how people on the ocean use ocean data. I don’t really know what they want to know, or why. I wasn’t totally prepared for the incredible onslaught of questions and information pertaining to how…

Related thinking: https://www.lesswrong.com/w/the-map-is-not-the-territory

Re: Domain expertise has always been the real moat

#564

Post by ElProlactin »

emodendroket wrote: post #562
ElProlactin wrote: post #548

Earlier quoted context omitted.

> I bought a book however many years ago with no previous development experience and delivered a Web app people paid for and eventually honed that as an actual career, so I’m just not really seeing what’s a difference in kind here. So you taught yourself how to program. Do you not see that there's a difference between teaching yourself how to program so that you could build a web application and using a new technolog…

> So you taught yourself how to program. Do you not see that there's a difference between teaching yourself how to program so that you could build a web application and using a new technology that allows you to build a web application without having to learn how to program? Well, I guess where we're not lining up is I don't see that you are going to deliver a serious application that isn't the equivalent of table kno…

> Well, I guess where we're not lining up is I don't see that you are going to deliver a serious application that isn't the equivalent of table knocked together in a home workshop without, in practice, "learning how to program."

My argument is that AI is now "good enough" that there are real people who are smart and tech-savvy but who don't know how to program who are building real applications, shipping them and using them commercially.

By any reasonable standard, these people haven't "learned how to program." They've learned how to use a tool that can program for them, troubleshoot for them, give them clear, step-by-step instructions on how to deploy to numerous services that have made it possible for non-developers to deploy, etc.

> Sure, but I can do more stuff faster with AI, which they also value enough to pay for the tools. Is that a serious question? You will find few professional software developers whose employers aren't encouraging AI tool use today.

But you stated previously:

> I also disagree with the “democratization” frame because now developers are spending like $1000 per month on tokens at their jobs, which does the opposite of making things more accessible.

So who is paying for the tokens? You or your employer? If your employer is paying for them, what's the problem?

> I don't really agree with the idea that all the barriers are gone and everyone's ready to deliver commercial-grade software without understanding what they're doing. If you think that's the case, then you'll have to conclude most businesses are acting highly irrationally by continuing to pay high wages to employ people with specialized knowledge of software development to operate AI tools rather than just handing it over to your new breed of semiskilled laborers who don't need to know how to program.

I never argued that all barriers are gone and that every idiot can deliver "commercial-grade" software. What I've argued, again, is that AI has for a growing number of smart non-developers improved to the point where it offers a third path separate from learning-to-program or hiring a developer.

As for what businesses are doing, the general trends speak for themselves. Companies are citing AI in layoffs. It's absolutely brutal right now for new grads and juniors who a decade ago were inundated with 6-figure offers. Lots of freelancers/contractors/agencies who could easily sell 5 and 6-figure projects or command $xxx/hour rates just a few years ago are finding it much harder to do so.

The market for highly-paid developers isn't going to 0 overnight but anyone who thinks it isn't going in a certain direction is in my opinion in denial.

Re: Domain expertise has always been the real moat

#565

Post by rcxdude »

hn_throwaway_99 wrote: post #337
vmurthy wrote: post #324

Earlier quoted context omitted.

>One of the things he told me, and that I also observed, was that the vast majority of financial experts (basically, the people in the accounting department of companies) had an extremely difficult time just telling him what the rules of any particular transaction should be We have internalized more knowledge than we can explain sounds like the textbook definition of Polanyi's paradox:"Polanyi's paradox, named in hon…

Thanks so much! I hadn't heard of Polanyi's paradox before but it is exactly what I was talking about. The Wikipedia entry even highlights the problem for AI-driven development: > Polanyi's paradox has been widely considered to identify a major obstacle in the fields of AI and automation, since programming an automated task or system is difficult unless a complete and fully specific description of the procedure is av…

I'd say the opposite in some ways: it's arguably the reason for the bitter lesson and ML techniques tending to beat other approaches, because they work by giving examples as opposed to trying to articulate the rules in an explicit manner.

Re: Domain expertise has always been the real moat

#566

Post by maccard »

ryan_n wrote: post #145
maccard wrote: post #104

Earlier quoted context omitted.

But a modern drill absolutely 100% removes the need for a brace and bit. An LLM doesn't replace any existing tools.

From what I've heard for many devs it replaced an IDE... I still use one myself, but I've a lot of people don't anymore.

I've heard that. I hope that the people who are IDE-free are just better AI-wranglers than I am because my experience has been that if I can get an agent to one shot something, it's fine but if I can't, the agents tend to make an absolute mess of spaghetti that doesn't actually do what it was asked to do IME.

Re: Domain expertise has always been the real moat

#567

Post by lpln3452 »

threethirtytwo wrote: post #7

“The hard part of writing software has never been the writing.” I’m tired of these endless articles on HN about software engineers trying to reinvent their identity while trying not to lose touch with reality. One way of dealing with LLMs is to deny the skill level of LLMs. Claim they can’t code as well as you. This excuse works to a certain extent but it also fails because not only are their multitudes of cases wher…

It is premature to conclude that someone is wrong and will ultimately be replaced, as the final verdict on LLMs is not yet in. In fact, recent trends show rather negative outcomes. Multiple analyses continuously indicate that recent layoffs are driven by the normalization of the post-pandemic bubble and economic recession due to wars, not LLMs. Companies are merely using LLMs as a convenient excuse to frame hiring freezes as innovation. Furthermore, many companies are reconsidering LLM adoption due to cost and efficiency issues.

You claim to be rational and logical, but your argument completely lacks substance and is just full of highly subjective claims. Your 'prediction' is closer to astrology than actual forecasting. Sure, prophecies hit the mark by chance every now and then, but that doesn’t prove the person has any predictive ability. That is precisely what your argument looks like: completely confident without a single piece of evidence. To top it off, you totally abandoned reason and logic in your last sentence. Saying that anyone who disagrees with you is just deluding themselves and that you already know what people will say to snap back is exactly the kind of stuff a cult leader would say.

You're completely giving up based on some strange delusion. I don't even blame you for that. But it's genuinely ridiculous how you use it as an excuse to attack others and act all smug while pushing your defeatist arguments.

Re: Domain expertise has always been the real moat

#568

Post by detourdog »

bigfatkitten wrote: post #543
detourdog wrote: post #409

Earlier quoted context omitted.

Verizon in NYC was trying to make ISDN happen in the home in the mid 90's. I had it. The hard part was getting an ISP that supported SLIP.

In Australia ISDN was available pretty much everywhere to any residential customer who wanted to pay for it, but it was not particularly popular. It was widely used by businesses, however. The main reason was that the line rental was significantly higher, POTS lines had untimed flat rate local calls, whereas ISDN didn’t, and ISPs charged more for ISDN plans. Towards the end of the 90s that changed for voice calls, an…

My favorite part was due to it being a digital line there was no modem negotiation.

Re: Domain expertise has always been the real moat

#569

Post by jillesvangurp »

wiseowise wrote: post #366
jillesvangurp wrote: post #355

Agentic coding favors senior generalists with a broad experience in many things rather than narrow experience in one or two things. I no longer think of myself as a backed engineer. That's what I was. I can do it all now. I've been building products and doing CTO jobs for a few decades now. I'm not a specialist in all layers of the software but I know enough about all layers of the stack to be able to do things with…

> Agentic coding favors senior generalists with a broad experience in many things rather than narrow experience in one or two things. Is it? Agents are coming for generalists first.

I would say agents will come for specialists equally hard. It takes a long time for a specialist to build up expertise in their domains. But given enough background material, an AI agent can absorb a lot of the essentials and do productive things in the same domain. And of course a lot of general purpose frontier models have been trained on an insane amount of research and documentation. The only difference between training data for generalists and specialists is that there's a lot more material for generalists out there.

If you are a generalist with a lot of experience, it's not that hard to "specialize" to different domains that you have no direct experience with the help of AI. Effectively that's what generalists do anyway even without AI support. If this includes proprietary stuff, the job is basically making sense of a lot of specs, documentation, etc. LLMs are pretty good at picking apart stuff like that.

A lot of projects I've been doing in the last 25 years have in common that they always require me to wrap my head around a new business domain, new technology, or framework. That makes me a generalist. I got good at figuring out new things and filtering out things not relevant to the job at hand. That's what makes me a senior (that and my age).

I've on purpose stepped out of my comfort zone and targeted a few tech stacks lately that I've not used before with AI. That works amazingly well. My experience as a generalist is helpful, a lot of concepts are not really that tech stack specific and port well across stacks. I don't need to be a Go specialist to be productive with Go anymore. Same with typescript, rust, and python. I've barely touched Kotlin (my go to tech stack until last year) in the last months.

Re: Domain expertise has always been the real moat

#570

Post by RugnirViking »

aleph_minus_one wrote: post #558
DanielHB wrote: post #451

Earlier quoted context omitted.

I think you are underestimating how hard it is for average joe to learn programming basics. I remember a fellow in high school that just could not accept that = in programming is assignment not an equation (like in high school math)

> I remember a fellow in high school that just could not accept that = in programming is assignment not an equation (like in high school math) The English noun "mist" has a different meaning than the German noun "[der] Mist" (dung, crap). Or the French "[la] infusion" has a different meaning (herbal tea) than the English "infusion" and the German "[die] Infusion". So, you are very used to the situation that in differ…

not to mention the other english word, "missed", which sounds exactly the same.

However, people really do have a great deal of trouble with this exact concept. Its a common test to give to people to define a function with a misleading name, like multiplyByTwo(int x, int y), and then show them the code inside (return x/y) and ask what this results in. Surprisingly many non programmers have a great deal of trouble with the idea that the name can be different to the behavior, especially in cases where its subtly different not entirely like this.

Other fields have similar issues, where people have trouble with this sort of abstract thinking, divorcing abstract knowledge about a thing from its appearence. iirc its a stage in child psychology and development, and shows up in teenage years after puberty. Many adult people never develop it fully, and struggle with this to a greater or lesser extent

Post reply on HN