kinlan

I think the implication was that an LLM generated the text (which isn't an issue for me even if it was.)

That all being said, I love this project and I love how you're exploring linking in other mediums. I tried to do other linking demos via https://aifoc.us/hypermedia/ - it's something we don't do anywhere near enough

It's hard sometimes because say it's something that I wrote but someone else posted to HN, I've just had a lot of people's opinions foisted on me.

I'm relatively immune to a lot of things, but we're also entering a world where a lot of people can build and might not expect to have potentially millions of people critiquing their work to the level they do.

Slopidly slop slop

I think we still skew back to an insanely high input token ratio when you consider agentic loops. For example, when I see the tools I use do a web fetch or a search or other tool use, it's an incredibly high number of new input tokens.

That's actually useful to know and it aligns with what I see (I wrote the cost post)

I still spend time in Bangor

I don't want to say Yes... but... given all of these tools are mostly built with JS and wrapped in a TUI we could probably go some way to having it run in the browser. There are fewer and fewer Node based APIs that haven't got a way to run in the browser.

I also implemented this in the example site in the post.

The title and content of my post?

Author of the linked post here, years ago there was a thing called "Magic iframes" that would allow you to move an iframe between windows - like a Service Worker before ServiceWorkers. I was always amazed by some of the things you could do, but now it seems we forget about iframes :D

You beat me to it. Thanks for sharing it

Author of the linked post here. This is actually a pretty interesting idea, I'll pass it to the team.

I was Addy's manager when he was on Developer Relations.

He moved to an engineering manager role on Chrome DevTools many years ago and has recently just moved on to a different team. I don't think it's fair at all to say he's not a developer working on a product shipped to users when he led one of our most used developer tools, as well as worked on many of our developer libraries prior to moving to the Engineering manager role.

Here's the OSS repo for Google Opal: https://github.com/breadboard-ai/breadboard

Got it :)

As one of the owners of the GoogleChromeLabs org. Technically anything in this org is not officially supported as it is intended for prototypes and things that might graduate to more fully supported products if there is a fit in the market.

That all being said, I believe this particular change to this particular repo was 5 years ago.

I used it as an example because I felt the data was pretty clear. I also felt that it follows a very human pattern (generative tools need customers, like other tools before, so they go with what the industry is demanding).... but now we seen an acceleration.

Huh - that's actually pretty interesting and I hadn't thought of that as an option.. I know Preact was built as a faster alternative while being broadly compatible, but what you are describing is maybe even blending the technologies as that short circuit. neat.

Fwiw - I'm hoping it can break out too. But one of the biggest challenges is that last bit "asking it to use vanilla JS" - unsee this all the time in developer relations: getting developers to ask for a specific thing or even have it on their mind to think about using it is one of the biggest hurdles.

My actual long term hope is that in the future we won't need to think about frameworks at all: https://paul.kinlan.me/will-we-care-about-frameworks-in-the-...

Just to push back on this a tad. Yes there's growth React, it's popular, but it was consistent up until the introduction of some of the more popular code generation tools where there is a clear acceleration (if you believe builtwith.com data) in the last 9 months or so.

Author here. This is a fair comment. If you have a corpus that can be used as context already it's not like the LLMs will be forcing you in to React, there's probably enough bias (in a good way) to ensure the tool continues to be useful.

What I was trying to get at in the post is that net new experiences is where I see a massive delta