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

AI is stifling new tech adoption?

vale.rocks

Post reply on HN
Print view View original

311–320 of 442 posts

Re: AI is stifling new tech adoption?

#311

Post by benatkin »

throwup238 wrote: post #285
janalsncm wrote: post #272

I’ve been out of web dev for a while, but maybe the problem is there’s a new framework every 6 months and instead of delivering value to the end user, developers are rewriting their app in whatever the new framework is.

I don’t think that’s really the case anymore. The vast majority are on React, Vue, or Svelte (in order of my perception of their popularity). On the CSS side it seems like Tailwind and PostCSS has taken over. The heavier framework category is covered by Next.js. Other than Next, most of that popularity started solidifying in 2020. There are a bunch of newer frameworks like Astro and HTMX and so on, but it doesn’t see…

I think their combined market share is finally shrinking for leading edge projects. https://news.ycombinator.com/item?id=43008190 https://news.ycombinator.com/item?id=42388665 Also, AI makes a lot of terrible mainstream stuff because the natural bias is towards the mainstream. That's where I count its tendency to default to React if I ask for frontend code without further context.

Re: AI is stifling new tech adoption?

#312

Post by xmprt »

baq wrote: post #305
xmprt wrote: post #298

Earlier quoted context omitted.

If your abstractions match common design patterns then you've solved your problem. It's ridiculous to assume that an LLM will understand 1k LOC of standard library code better than 10 lines of a custom abstraction which uses a common design pattern. It's more prone to hallucinating things if your custom abstraction is not super standard but at least you'd be able to check its mistakes (you're checking the code genera…

LLMs are not human, they see the whole context window at once. On the contrary it’s ridiculous to assume otherwise. I’ll reiterate what I said before: put the whole source of the new library in the context window and tell the LLM to use it. It will, at least if it’s Claude.

Attention works better on smaller contexts since there's less confounding tokens so even if the LLM can see the entire context, it's better to keep the amount of confounding context lower. And at some point the source code will exceed the size of the context window; even the newer ones will millions of tokens of context can't hold the entirety of many large codebases.

Re: AI is stifling new tech adoption?

#313

Post by rhet0rica »

jimmaswell wrote: post #294
mxwsn wrote: post #286

This ought to be called the qwerty effect, for how the qwerty keyboard layout can't be usurped at this point. It was at the right place at the right time, even though arguably its main design choices are no longer relevant, and there are arguably better layouts like dvorak. Python and React may similarly be enshrined for the future, for being at the right place at the right time. English as a language might be anothe…

QWERTY is a poor example. The keyboard layout is not the bottleneck for anyone who does not participate in olympic typing competitions. DVORAK is just as arbitrary as QWERTY to everyone else including professionals, and there's value in backwards compatibility e.g. old keyboards don't become e-waste.

It sounds like you're thinking purely about the speed of a skilled typist. Alternative keyboard layouts offer a tangible ergonomic benefit even at lower WPM counts, and can have a lower hunt-and-peck time for novices by clustering frequently-used letters together. (This last effect is particularly pronounced on small touch screens, where the seek time is non-trivial and the buttons are much too close together for any sort of real touch-typing.)

Re: AI is stifling new tech adoption?

#315

Post by squigz »

janalsncm wrote: post #272

I’ve been out of web dev for a while, but maybe the problem is there’s a new framework every 6 months and instead of delivering value to the end user, developers are rewriting their app in whatever the new framework is.

This would be a symptom of bad management, not bad developers.

And to add to what others have said, this stereotype never really held up in my experience either. Any serious web dev shop is going to have the framework they use and stick with it for both long- and short-term clients. And there are many mature options here.

I don't doubt this happens, a lot, but again, I think it's more about bad management than anything - and bad management will always make bad tech decisions, no matter the topic.

Re: AI is stifling new tech adoption?

#316

Post by hinkley »

milesvp wrote: post #300
crazygringo wrote: post #252

No, AI isn't. Any new tech, or version upgrade, or whatever, takes time for people to become familiar with it. You might as well say "Stack Overflow is stifling new tech adoption" because brand-new stuff doesn't have many Q's and A's yet. But that would be a silly thing to say. I'm not going to adopt a brand-new database regardless of LLM training data cutoff, just because enough people haven't had enough experience…

Strong disagree here. I've been trying to learn Zig, and I'm thwarted enough by chatgpt giving me outdated information on Zig's unstable API that if I didn't have a strong incentive to learn it for it's cross compiler, I'd likely turn my efforts towards another language. This effect can greatly alter the adoption curve of a new tech, which can leave it dying on the vine. You're not wrong though, in that Stack Overflo…

Someone recently put it that it’s the first 10% of adoption that sets the future for a product.

There are people online calling it the “tech adoption cycle” but this is a concept I encountered in a literal Business 101 class. 2.5% of the population are Innovators. 12.5 are early adopters. Then there’s 70% in the middle where most of your cash comes in, and by the time the laggards hit you’re optimizing for cost per unit because you’ve had to drop the price so much due to competition from copycats and from the next new thing.

So by the time 60% of your early adopters are onboard it’s already been decided if you’re on a rocket ship or this is just a burp.

Early adopters have a high tolerance for inconveniences but it’s not infinite. If they bleed enough they will find something else, and then you are DOA.

Re: AI is stifling new tech adoption?

#317

Post by _puk »

milesvp wrote: post #300
crazygringo wrote: post #252

No, AI isn't. Any new tech, or version upgrade, or whatever, takes time for people to become familiar with it. You might as well say "Stack Overflow is stifling new tech adoption" because brand-new stuff doesn't have many Q's and A's yet. But that would be a silly thing to say. I'm not going to adopt a brand-new database regardless of LLM training data cutoff, just because enough people haven't had enough experience…

Strong disagree here. I've been trying to learn Zig, and I'm thwarted enough by chatgpt giving me outdated information on Zig's unstable API that if I didn't have a strong incentive to learn it for it's cross compiler, I'd likely turn my efforts towards another language. This effect can greatly alter the adoption curve of a new tech, which can leave it dying on the vine. You're not wrong though, in that Stack Overflo…

Having spent the past week deep in cursor, it's amazing for building out a basic web app.. but getting it to a point of productivity takes a while..

Command line install for latest svelte.. nope npx install is now deprecated, have to do it another way.. ok, let's go old school and read the docs.

Great, it's up and running, but nope, Svelte has just hit V5 and the LLM isn't aware of the changes.. ok, do I drop back to 4 on a new code write, or spend the time to get a decent .cursorrules in place to make sure it's using the right version. Ok, that's in, but look tailwind is too new too.. ok, fine let's get that into .cursorrules..

Oh look, daisy has a v5 coming in 15 days, luckily I got in there just in time..

I thought svelte, tailwind and daisy were the one true way nowadays!

I now have a rule in my cursorrules that asks for any errors that I spot in the code (related to wrong versions) results in both the fix and a rule for the cursorrules so it doesn't happen again. That works well.

Re: AI is stifling new tech adoption?

#318

Post by hinkley »

crazygringo wrote: post #271
TheTaytay wrote: post #268

Earlier quoted context omitted.

I feel myself taking AI’s base knowledge of a tech stack into account when I work. Otherwise it feels like I am swimming upstream. I can’t be the only one, and this article resonated for me.

Sure, but what I'm saying is that's not where the knowledge bottleneck is. The knowledge bottleneck is in the human population , which is then reflected in Stack Overflow and blogs, which is then reflected in LLM's. LLM's aren't doing anything special to stifle new tech adoption. New tech is harder to adopt because it's new tech, because people in general are less familiar with it . (And there's a little bit of a tra…

It’s generally true that the people who come up with the first interesting idioms in a new space get an outsized influence on the community. It takes a while for people to borrow or steal their ideas and run with them.

On the plus side while the AI can’t write in that language there’s money to be made doing the work organically.

Re: AI is stifling new tech adoption?

#319

Post by DangitBobby »

bilater wrote: post #275

This is precisely why I have said that every new framework/library should have a markdown or text or whatever is the best format for LLM models endpoint that has all the docs and examples in one single page so you can easily copy it over to a models context. You want to make it as easy as possible for LLMs to be aware of how your software works. The fancy nested navigation guide walkthrough thing is cool for users bu…

This is something I'd like to have for pretty much any framework/library anyway.

Re: AI is stifling new tech adoption?

#320

Post by baq »

xmprt wrote: post #312
baq wrote: post #305

Earlier quoted context omitted.

LLMs are not human, they see the whole context window at once. On the contrary it’s ridiculous to assume otherwise. I’ll reiterate what I said before: put the whole source of the new library in the context window and tell the LLM to use it. It will, at least if it’s Claude.

Attention works better on smaller contexts since there's less confounding tokens so even if the LLM can see the entire context, it's better to keep the amount of confounding context lower. And at some point the source code will exceed the size of the context window; even the newer ones will millions of tokens of context can't hold the entirety of many large codebases.

Of course, but OP’s 1kloc is nowhere near close to any contemporary limit. Not using the tool for what it’s designed because it isn’t designed for a harder problem is… unwise.
Post reply on HN