Comment by simoncion
6 days ago
> I'm a little ... concerned that this will accelerate the adoption of languages and frameworks based on their popularity and bury away interesting new abstractions and approaches...
I'd argue that the Web development world has been choosing tooling based largely on popularity for like at least a decade now. I can't see how tooling selection could possibly get any worse for that section of the profession.
I disagree. There’s a ton of diversity in web development currently. I don’t think there’s ever been so many language and framework choices to build a web app.
The argument is that we lose this diversity as more people rely on AI and choose what AI prefers.
> There’s a ton of diversity in web development currently.
You misunderstand me. It's not incompatible for a culture to choose options based largely on popularity (rather than other properties that one would expect to be more salient when making a highly-technical choice), and for there to also be many options to choose from.
You raise a valid concern but you presume that we will stay under the OpenAI/Anthropic/etc oligopoly forever. I don't think this is going to be the status quo in the long-term. There is demand for different types of LLMs trained on different data. And there is demand for hardware. For example the new Mac Studio has 512gb VRAM which can run the 600B param Deepseek model locally. So in the future I could see people training their own LLMs to be experts at their language/framework of choice.
Of course you could disagree with my prediction and that these big tech companies are going to build MASSIVE gpu farms the size of the Tesla Gigafactory which can run godlike AI where nobody can compete, but if we get to that point I feel like we will have bigger problems than "AI react code is better than AI solidjs code"
I suspect we’ll plateau at some point and the gigafactories won’t produce a massive advantage. So running your own models could very well be a thing.
2 replies →
In the relatively near future this is going to be like arguing what CPU to buy based on how you like the assembly code. Human readability is going to matter less and less and eventually we will likely standardize on what the LLMs work with best.