[austin-cheney]

Quite the gap.

The legends the article talks about are legends because they either started a project that blew up in popularity and/or solved a demanded problem with original code.

For most people writing software for a living that is gone. Its been gone since I started writing software 20 years ago. The goal post has moved. Its no longer about solving any problem. Its about hiring.

The distinction is massive. Most of the people doing this work will never encounter an important problem to solve or write original code. Instead they will use tools and modify templates. There is still some troubleshooting there, but no originality. Its like being a plumber. Plumbers still make good money, but they aren't engineers. Now, with AI being pushed on everybody even becoming something like a plumber is becoming a distant gap for the next generation.

The most clear exception are hobbyists, which has always been there as an exception through my years writing software. The only real distinction between most of these hobbyists and the legends is obscurity. The very real distinction between the hobbyists and less original professional is time spent practicing.

[suncemoje]

Reminds me of when Jensen Huang recently compared Linux to OpenClaw and showed this ridiculous GitHub star comparison. To me these projects are incomparable.

[embedding-shape]

It's worth remembering that when people like Jensen talk publicly, they most of time are addressing investors in various indirect and direct ways. Comparing Linux with OpenClaw is obviously bullshit and irrelevant, for almost everyone except clueless investors who like that sort of stuff. He's saying those things for the people who neither understand Linux nor OpenClaw, but have lots of money regardless.

[chrishill89]

Does that show Jensen in a better light?

[jlundberg]

Yeah, GitHub stars is becoming a vanity metric and not indication of quality.

I have been contemplating a rating system for open source software with a mandatory tag for each star. Allows you to filter out perspectives you don’t care about.

[embedding-shape]

> GitHub stars is becoming a vanity metric and not indication of quality

Becoming? :D Since day one, me and other's have called it a vanity metric, and trying to push back on hiring decisions being made over what developers have the most starred repositories/followers (no joke, one place I worked at almost hired one developer over another because of their "total star count" :'( ).

Stars been around for as long as GitHub been around, and people actively shouting to get people to stop caring so much about stars been around for the same time yet.

[seba_dos1]

Have these stars ever been useful at all? For me they've been just a cute noise ever since they were introduced. A rough proxy for project's visibility in a certain specific context, nothing more.

[jdw64]

I agree with you. There's a saying in Korea and East Asia: 'Open source is a moat.' This might sound difficult, but it means that if you're trying to sell a product, its quality or UX/UI basically needs to be better than what's already publicly available in open source — and that's not easy.

As the era becomes increasingly advanced, the cognitive cost of making a single project public keeps rising. But if you try to use an LLM to share or assist, there are many people who say LLMs are bad.

It's a difficult problem

[keeda]

> Now, with AI being pushed on everybody even becoming something like a plumber is becoming a distant gap for the next generation.

Yes, but this also has a silver lining: nobody will need to do the rote, toily, plumbingeque, but still economically valuable work that most software engineers spent most of their time on, which means they will be freed up to solve drastically more novel and/or higher level problems.

The obvious issues with this are 1) not everybody is a Carmack or a Bellard; 2) not everybody may want to be, either; and 3) you most likely won't get to be them by outsourcing all your coding to an LLM (meaning we would need to rethink upskilling.)

The short term will be painful, but in the long term, Economics theory suggests we would experience a golden age of productivity and progress once we build the pipeline of novel / higher-level problems to solve and upskill the hobbyists and erstwhile plumbers to tackle them.

[austin-cheney]

The problem moving forward is a concept known as lost knowledge. The ancient Egyptians had steam powered robots, full industrialization, incredible libraries, and paved roads. Europe had none of that until 200 years ago even though they were in Egypt at the time.

If people cannot refactor the code they own the only option is to start from scratch every single time. Any perceived economic benefits assumed of AI cannot be realized if the common knowledge underlying it dies.

[keeda]

Hmm, I doubt that would be an issue for a couple of reasons:

1. We have infinitely more durable means of storing, indexing and retrieving knowledge (and now potentially even means of programmatically reasoning about that knowledge!) We can re-derive things from our records using first principles, as long as we have critical reasoning skills -- about the only skill we will need in the future, and ironically the one most at risk of becoming rarer as people increasingly outsource all thinking to LLMs.

2. We will always need and have engineers of the caliber in TFA. They derive their capabilities, in no small part, by having in-depth knowledge of the entire technology stack they work with. I'd say most of them can operate at the abstract architecture and algorithmic level down to low-level hardware bit-twiddling. If most engineers in the future are of that caliber (which I'd argue is now easier with LLM assistance for those so inclined) there is no chance we'd lose that knowledge.

[austin-cheney]

That missed the point in past history and misses the point now. The technology to better preserve human knowledge has been available for thousands of years. It’s an irrelevant compensator for human behavior. If people, in large enough numbers, are not willing to learn, and extend, their craft it will fade into obscurity.

There is no cheat sheet or magic tool for human behavior.

[keeda]

Hmm, I would say the technology to durably preserve human knowledge is actually very recent. Until just a few decades ago, other methods have been extremely lossy. Even today, with triple-redundant, geographically-distributed data replicas we see cases of catastrophic data loss. Previously, all it took was an errant spark in a library.

You're right that the main driver for loss of knowledge is human behavior, but I would posit the underlying reason is economics. Craft that reduces in economic value simply disappears because there is no incentive to preserve it.

As such, there is a cheat sheet for human behavior, it's called economics! ;-) And the already high economic value of code is only going to keep increasing sharply, so even if LLMs do most of the work there will be strong incentives for people to manualy craft code whenever LLMs, like all systems, inevitably fall short.

[lelanthran]

In a previous post I lamented that my I should no longer call myself a software developer - after all, I write my own code!

Maybe there should be a distinction between software creators and programmers.

[brookst]

In the same way there’s a distinction between writers and typists, and you really don’t hear anything about typists these days.

[lelanthran]

> In the same way there’s a distinction between writers and typists, and you really don’t hear anything about typists these days.

Not sure that analogy applies. I'd compare it to the difference between an engineer building a bridge and a politician ring-fencing the funds for it.

[brookst]

Again, these tools let you operate in that model as well.

But if you want to design the aesthetic of the bridge, ensure that it is ergonomic for cars, bicycles, pedestrians, make sure it is designed to last and withstand rare weather events, and fully modeled with FIE… you can do that too.

AI is like having a thousand specialists in every field, waiting for you to ask. You can have them research, check each other’s work, build simulations, etc… or you can let them go unused, or you can ask them to tell you how smart you are.

[chrishill89]

It took me many years to even think of problems that I wanted to solve. Or were missing. But eventually it happened. And it seems that the more time passes the more that list of ideas just expands since no one can follow up on all of those ideas.

Someone skilled enough should have many original enough problems to work on. But such persons would have to speak on that topic.

[laughing_man]

As my brother said once with exasperation: "I got into this business to write code, and now I'm just an integrator."

[amelius]

integrator -> plumber

[raffael_de]

personally I don't find this particularly unappealing and have often referred to myself as some sort of plumber. plumbing is all about connecting standardized interfaces (threads) and then some improvisation. in IT the amount of improvisation is higher due to less standardized interfaces and interfaces are more complex. but the analogy works and I enjoy thinking about how to make interfaces exchange information efficiently.

[brookst]

If you choose that direction, yes. But you can also choose integrator -> CPO.

[xtiansimon]

> “… will never encounter an important problem to solve or write original code.”

I’m sure you’re right. Though, let me add, there are a lot of minuscule problems in the small business space. Not fame and fortune level, but gratifying nevertheless.

[tim333]

>Most of the people doing this work will never encounter an important problem to solve

AGI is an interesting one.