Sorry you’re getting downvoted. Ideally downvoting would be for unconstructive posts, or posts with good info / good contributions that are presented unconstructively.
You’re just being controversial.
That’s not a strong enough reason to downvote someone.
Thanks for pointing this out. I don't participate in HN discussions like I used to because the HN crowd and I don't agree on much, and down-votes is not an engaging counter-point.
Fwiw I think it’s perfectly fine if people downvote me if they disagree with me. I think that’s an unavoidable effect of having up/down arrows, regardless of what the rules say. If i say something controversial I expect some downvotes. I just hadn't expressed myself clearly enough initially, everybody took me as a “money makes right” capitalist (not a weird assumption, theres plenty of those here on HN) and fortunately could still edit to clarify.
The new world is one where someone can have an LLM assisted insight, post it on their blog for free, have it indexed by every agentic search engine, and it becomes part of the zeitgeist. That’s the new data that’ll feed the new models: a better information diet over time. And
guess what else: models are getting better at identifying - at scale - the high quality info that’s worth using as training data.
There is a huge difference between giving eggs to your neighbours and sending them to a breakfast restaurant in another state. A connection is made between you and the neighbors. A community forms around it.
What would be the point of having a machine write your thoughts so they can shift a model's weights by infinitesimal amounts? How would that compare to building a small following, getting reader mail, and maybe even meeting a few of them?
My website is used as training data. I get nothing from it. The AI twists words I have carefully selected, misleads people, and strips me of the fruit of my labour. If I lost my entire audience, I would stop doing that work.
Moreover, there is a cost to producing high quality information. Research alone takes a long time when you're putting information online for the first time. I wouldn't do it for a vague chance of affecting the Zeitgeist.
Nice work. Aren’t there decent open source alternatives though? What do you think your differentiation will be vs. a customer using an open source solution and hosting the video chunks on a CDN - or even S3?
There aren't any end-to-end open source video host solutions out there from what I can tell. DIY ffmpeg + a CDN is a great way to go. But quickly erodes when you want all the other niceties that are table-stakes today (like storyboards, subtitles, chapters, etc.).
I'll have all the niceties, but I plan to differentiate mainly on performance and quality.
- Higher quality compression (via AV1 encoding)
- Fast load times worldwide (Framerate's custom player is 18kb gzipped versus 200kb+ for vidstack/mux)
- Better publishing experience (bulk editing options, team collaboration, etc.)
Sounds good. I’d suggest niching down and talking to some customers to see who genuinely cares about those things - as they might turn out (tragically I know) to be things only techies care about.
Fair warning.. when we hear talk to customers we are tempted to reply “I talked to customer in my mind and they love my thing!”
But unfortunately customers in our minds don’t have money :)
There’re either customers out there, or if not there's learning about why youtube has almost 100% market share..
That’s too much for a model to carry in its context while it’s trying to do actual work.
Far better is to give that skill.md to a model and have it produce several hundred lines of code with a shebang at the top. Now you haven’t got a skill, you’ve got a script. And it’s a script the model can run any time to check its work, without knowing what the script does, how, or why - it just sees the errors. Now all your principles of web dev can be checked across your codebase in a few hundred milliseconds while burning zero tokens.
TDD is codification too: codifying in executable form the precise way you want your logic to work. Enforce a 10ms timeout on every unit test and as a side effect your model won’t be able to introduce I/O or anything else that prevents parallel, randomized execution of your test suite. It’s awesome to be able to run ALL the tests hundreds of times per day.
Constantly checking your UI matches your design system? Have the model write a script that looks at your frontend codebase and refuses to let the model commit anything that doesn’t match the design system.
Codification is an insanely powerful thing to build into your mindset.
As one often finds with “effective LLM usage” advice, all of those things would help humans on the team as well! As would other advice like keeping the architecture docs up to date, writing down important design decisions with rationale, breaking big features down into steps, etc.
Maybe one should just search for advice from the last 20 years on how to make a human development team more effective, and do that stuff.
It’s funny how this advice has always been around, but we needed to invent this new kind of idiot savant developer to get the human developers to want to do it…
Watch out for Waste of context, whatever can be checked by existing linting/testing tooling and returned as an exact message to the model/agent the better.
We have crossed into a fascinating time in computing: market power lets incumbents make “whatever”, shove it in a forced update with no going back, and still be valued in the trillions.
I honestly think that the ability to shove updates out live is one of the worst things about software these days. Well, the ability to do so isn't bad, but the way companies actually do it makes it bad.
A html forms based radio button is worse than a complex - but standard - shadcn radio button many ways that matter in the real world.
Why does no one do the simpler thing? Because there’s no extra value to it, and it in fact has negative value because then the team has to write and understand it and the rationale for the departure from just using the same component library everywhere.
“Only a few kb of javascript” may as well ZERO javascript, and because of that it’s not even close to the top thing to optimize on your favorite site.
So, you engineered a non-standard radio button that is different to the rest which all use shading?? Why weren’t you building features that you know.. make money?
How's using a custom library any way close to "standard"? How about the actual HTML standards? The whole reason you'd use "shadcn" is that customizing the actual HTML radio button isn't enough for you. Otherwise, if you just want a default-looking button, here you go:
If your team can't understand that, how are they going to understand a few KB of JS? Or maybe they're not supposed to understand it, but how can you then guarantee to your customers there isn't a crypto miner or tracker or something in that? Or perhaps you care more about "making money" than protecting your customers from such things?
You should not look at the button in isolation. The library is likely used to do other things vanilla HTML cannot do, but instead of maintaining multiple code, they just use the library to implement everything.
The library has in essence became an interface for developers to build for.
Not sure why you’re getting downvoted, but that’s exactly what AI is turning out great for.. being able to make something in a weekend that would’ve taken weeks otherwise means other things downstream of it suddenly also become possible.
You’re just being controversial.
That’s not a strong enough reason to downvote someone.
reply