3 ms·

That's one of the sentiments I don't quite grasp, though. Why can't they just learn the tools when they're stable? So far it's been sooo many changes in workflo

That's one of the sentiments I don't quite grasp, though. Why can't they just learn the tools when they're stable? So far it's been sooo many changes in workflows, basically relearn the tools every three months. It's maybe a bit more stabilized the last year, but still one could spend an enormous amount of time twiddling with various models or tools, knowledge that someone else probably could learn quicker at a later time. "Being left in the dust" would also mean it's impossible for new people / graduates to ever catch up. I don't think it is. Even though I learned react a few years after it was in vogue (my company bet on the wrong horse), I quickly got up to speed and am just as productive now as someone that started a bit earlier.

duggan 8mo ago
Not the person you asked, but my interpretation of “left in the dust” here (not a phrasing I particularly agree with) would be the same way iOS development took off in the 2010s. There was a land rush to create apps. Basic stuff like the flash light, todo lists, etc, were created and found a huge audience. Development studios were established, people became very successful out of it. I think the same thing will happen here. There is a first mover advantage. The future is not yet evenly distributed. You can still start as an iOS developer today, but the opportunity is different.
- lapcat 8mo ago
  I’m not sure your analogy is applicable here. The introduction of the App Store did not increase developer productivity per se. If anything, it decreased developer productivity, because unless you were already already a Mac developer, you had to learn a programming language you've never used, Objective-C, (now it's largely Swift, but that's still mainly used only on Apple platforms) and a brand new Apple-specific API, so a lot of your previous programming expertise became obsolete on a new platform. What the App Store did that was valuable to developers was open up a new market and bring a bunch of new potential customers, iPhone users, indeed relatively wealthy customers willing to spend money on software. What new market is brought by LLMs? They can produce as much source code as you like, but how exactly do you monetize that massive amount of source code? If anything, the value of source code and software products will drop as more is able to be produced rapidly. The only new market I see is actually the developer tool market for LLM fans, essentially a circular market of LLM developers marketing to other LLM developers. As far as the developer job market is concerned, it's painfully clear that companies are in a mass layoff mood. Whether that's due to LLMs, or whether LLMs are just the cover story, the result is the same. Developer compensation is not on the rise, unless you happen to be recruited by one of the LLM vendors themselves. My impression is that from the developer perspective, LLMs are a scheme to transfer massive amounts of wealth from developers to the LLM vendors. And you can bet the prices for access to LLMs will go up, up, up over time as developers become hooked and demand increases. To me, the whole "OpenClaw" hype looks like a crowd of gamblers at a casino, putting coins in slot machines. One thing is for certain: the house always wins.
  - duggan 8mo ago
    My take is more optimistic. I think it will make prototyping and MVP more accessible to a wider range of people than before. This goes all the way from people who don't know how to code up to people who know very well how to code, but don't have the free time/energy to pursue every idea. Project activation energy decreases. I think this is a net positive, as it allows more and different things to be started. I'm sure some think it's a net negative for the same reasons. If you're a developer selling the same knowledge and capacity you sold ten years ago things will change. But that was always the case. My comparison to iOS was about the market opportunity, and the opportunity for entrepreneurship. It's not magic, not yet anyway. This is the time to go start a company, or build every weird idea that you were never going to get around to. There are so many opportunities to create software and companies, we're not running out of those just because it's faster to generate some of the code.
    - lapcat 8mo ago
      What you just said seems reasonable. However, what the earlier commenter said, which led to this subthread, seems unreasonable: those people unwilling to try the tools "are absolutely going to get left in the dust." Returning to the iOS analogy, though, there was only a short period of time in history when a random developer with a flashlight or fart app could become successful in the App Store. Nowadays, such a new app would flop, if Apple even allowed it, as you admitted: "You can still start as an iOS developer today, but the opportunity is different." The software market in general is not new. There are already a huge number of competitors. Thus, when you say, "This is the time to go start a company, or build every weird idea that you were never going to get around to," it's unclear why this would be the case. Perhaps the barrier to entry for competitors has been lowered, yet the competition is as fierce as ever (unlike in the early App Store). In any case, there's a huge difference between "the barrier to entry has been lowered" and "those who don't use LLMs will be left in the dust". I think the latter is ridiculous. Where are the original flashlight and fart app developers now? Hopefully they made enough money to last a lifetime, otherwise they're back in the same boat as everyone else.
      - 8mo ago

duggan 8mo ago

Not the person you asked, but my interpretation of “left in the dust” here (not a phrasing I particularly agree with) would be the same way iOS development took off in the 2010s. There was a land rush to create apps. Basic stuff like the flash light, todo lists, etc, were created and found a huge audience. Development studios were established, people became very successful out of it. I think the same thing will happen here. There is a first mover advantage. The future is not yet evenly distributed. You can still start as an iOS developer today, but the opportunity is different.

lapcat 8mo ago
I’m not sure your analogy is applicable here. The introduction of the App Store did not increase developer productivity per se. If anything, it decreased developer productivity, because unless you were already already a Mac developer, you had to learn a programming language you've never used, Objective-C, (now it's largely Swift, but that's still mainly used only on Apple platforms) and a brand new Apple-specific API, so a lot of your previous programming expertise became obsolete on a new platform. What the App Store did that was valuable to developers was open up a new market and bring a bunch of new potential customers, iPhone users, indeed relatively wealthy customers willing to spend money on software. What new market is brought by LLMs? They can produce as much source code as you like, but how exactly do you monetize that massive amount of source code? If anything, the value of source code and software products will drop as more is able to be produced rapidly. The only new market I see is actually the developer tool market for LLM fans, essentially a circular market of LLM developers marketing to other LLM developers. As far as the developer job market is concerned, it's painfully clear that companies are in a mass layoff mood. Whether that's due to LLMs, or whether LLMs are just the cover story, the result is the same. Developer compensation is not on the rise, unless you happen to be recruited by one of the LLM vendors themselves. My impression is that from the developer perspective, LLMs are a scheme to transfer massive amounts of wealth from developers to the LLM vendors. And you can bet the prices for access to LLMs will go up, up, up over time as developers become hooked and demand increases. To me, the whole "OpenClaw" hype looks like a crowd of gamblers at a casino, putting coins in slot machines. One thing is for certain: the house always wins.
- duggan 8mo ago
  My take is more optimistic. I think it will make prototyping and MVP more accessible to a wider range of people than before. This goes all the way from people who don't know how to code up to people who know very well how to code, but don't have the free time/energy to pursue every idea. Project activation energy decreases. I think this is a net positive, as it allows more and different things to be started. I'm sure some think it's a net negative for the same reasons. If you're a developer selling the same knowledge and capacity you sold ten years ago things will change. But that was always the case. My comparison to iOS was about the market opportunity, and the opportunity for entrepreneurship. It's not magic, not yet anyway. This is the time to go start a company, or build every weird idea that you were never going to get around to. There are so many opportunities to create software and companies, we're not running out of those just because it's faster to generate some of the code.
  - lapcat 8mo ago
    What you just said seems reasonable. However, what the earlier commenter said, which led to this subthread, seems unreasonable: those people unwilling to try the tools "are absolutely going to get left in the dust." Returning to the iOS analogy, though, there was only a short period of time in history when a random developer with a flashlight or fart app could become successful in the App Store. Nowadays, such a new app would flop, if Apple even allowed it, as you admitted: "You can still start as an iOS developer today, but the opportunity is different." The software market in general is not new. There are already a huge number of competitors. Thus, when you say, "This is the time to go start a company, or build every weird idea that you were never going to get around to," it's unclear why this would be the case. Perhaps the barrier to entry for competitors has been lowered, yet the competition is as fierce as ever (unlike in the early App Store). In any case, there's a huge difference between "the barrier to entry has been lowered" and "those who don't use LLMs will be left in the dust". I think the latter is ridiculous. Where are the original flashlight and fart app developers now? Hopefully they made enough money to last a lifetime, otherwise they're back in the same boat as everyone else.
    - 8mo ago