> if you’re working on novel code, LLMs are absolutely horrible
This is spot on. Current state-of-the-art models are, in my experience, very good at writing boilerplate code or very simple architecture especially in projects or frameworks where there are extremely well-known opinionated patterns (MVC especially).
What they are genuinely impressive at is parsing through large amounts of information to find something (eg: in a codebase, or in stack traces, or in logs). But this hype machine of 'agents creating entire codebases' is surely just smoke and mirrors - at least for now.
20, 15, 10 years ago this sort of stuff (Rails, Laravel, Phoenix) was free, and open source. The good ones are still going strong. I wonder if these sorts of schemes will still be offered in 20, 15 or 10 years
the economic picture is not getting any better. In part because so many people each month are coming off fixed-term mortgage deals bought in the ZIRP era. The more time they give Labour to govern, the more rope they give them to hang themselves with. Or so the theory probably goes. They're already massively, massively behind in the polls and if recent local elections are anything to go on.
If you don't already know a language and its ecosystem, then I'd recommend Rails. It is simply the most mature and batteries-included framework. Ignore the claims that it is dead - if you're a solo developer the specific jobs market doesn't matter.
Ruby on Rails has recently started actively marketing itself as the solo developer framework. I've worked with it for a long time now (among a bunch of other languages and frameworks) and I have to say the recent updates to the frontend 'story' are really compelling for a solo dev.
One of the compelling reasons to write JS on the server was to have the same language and ecosystem in the client and server, and things like Turbo for Rails really go a long way towards delivering on that for Ruby (although of course you still need to understand javascript and browsers.)
That said, if you're already familiar with something else, then pick the most mature batteries-included framework in that language. Languages are just a tool for the job, it's incredibly unlikely as a solo dev that you can pick a "wrong" one.
A srangely loaded question. You’re not going to get a straight and accurate answer to such a broad philosophical topic.
Perhaps it’s worth reflecting on whether, in a pandemic, someone’s personal freedoms are incompatible with the means by which a virus can be prevented from spreading. And then, what is “worth” more?
Just want to say, I love your writing -- it's rare to find someone with such a distinctive voice. I've read a bunch of your articles and posts, keep going!
This is spot on. As people, a sense of community is not something we usually associate with commerce, but these online platforms are _only_ instruments to get a return on someone's investment.
The fact that part of the journey involves mutually beneficial behaviours (like using VC money to acquire more users, build a better UX, and make the community larger) is just an accident, and the incentive to capitalise the whole shebang will always win out.
This sort of thing feels like The Price We Eventually Pay for a massive, free-to-use online community.
got a source for that company? Presumably it'll be semi-public knowledge. I'd be incredibly surprised if a business was already proficient enough to replace software teams. Call centre operatives or content-farmers maybe...
You're among the first people I've encountered with both experience in writing your own software and prompting an LLM to do it. I have a few questions for you (maybe the answers will make you feel better, or maybe they'll make you feel worse)
- when the code the model spat out was wrong, how did you fix it? Did you identify it was wrong before you ran it?
- what level of complexity was there in the code, in terms of "business logic" or complexity of the requirements you fed into it?
I ask these two questions because I am not sure an AI will get to the level of experience you have in the near future in _generating_ complex applications, let alone being able to reflect on why its own creations are wrong, fixing them, deploying some output, and then explaining the changes?
A human is going to be in the loop in these cases for a long time, and I'm assuming part or all of your decades of experience has been spent understanding quite how poor people are at explaining their requirements. What if you thought of generative AI as a tool you can learn to utilise to do your job more effectively?
All the recent furore about Twitter's monetisation and Reddit's API changes reminded me of this article from 1993. Clearly, only one guy's opinion but I think we've strayed pretty far from this vision of the internet, as a protocol and connective tissue for people.
I can't help but feel like the fact the internet began as (or grew into) something free, underground, experimental led us inexorably to this path: where we expect or demand that online products are free at the point of use, which means we become the product.
I don't know that a small, more decentralised, cottage-y internet would be Better. But it's interesting to think about.
I understand what you’re saying and I agree that dismissing it out of hand is obviously the wrong approach - but this does feel like it meets the bar to be a “conspiracy theory”, in that people theorise that the dominant answer for something is wrong, and the true answer is being covered up by powerful groups of people.
This is cool, but I think you lost me when you said “it just works out of the box” about your OS. I mean, sure, but you’ve had to order a new screen, touchpad, battery and RAM. That’s a lot of boxes to work out of
This is spot on. Current state-of-the-art models are, in my experience, very good at writing boilerplate code or very simple architecture especially in projects or frameworks where there are extremely well-known opinionated patterns (MVC especially).
What they are genuinely impressive at is parsing through large amounts of information to find something (eg: in a codebase, or in stack traces, or in logs). But this hype machine of 'agents creating entire codebases' is surely just smoke and mirrors - at least for now.