Is it just me or is Apple's new Liquid Glass not very readable?
Is it just me or is Apple's new Liquid Glass not very readable?
The question remains the same as itβs always been: how much value can you provide to the customer.
With high quality token prices coming down, the value that apps can provide only goes up.
What I fail to see is how this can be a negative for AI *applications* ie how most users interact with LLMs
OpenAIβs entire business model is predicated on these models being difficult to run on cheap hardware.
Sure, weβre not at the point where we can run an o1-level model locally easily yet (at least not without culling it), but this is a signal that we could get there one day.
Lots of buzz about the ai bubble bursting these days.
Deepseek R1 certainly bodes poorly for the likes of Nvidia and OpenAI.
1/x
Good luck my friend
Did you know it's possible to run CommonJS directly in the browser? No bundlers.
It takes quite a few tricks to get it running.
Blog post coming soon.
Did you know it's possible to run CommonJS directly in the browser? No bundlers.
It takes quite a few tricks to get it running.
Blog post coming soon.
I love a good deep dive.
This past week I've learned so so so much about Node.js internals, wasm, wasi, wasix, and quickjs.
I feel like I'm on the verge of a breakthrough.
Like what? Svelte/Vue? Or backend included? Or something else completely?
What would you prefer?
Hey! I'm super late to this post, but I'm working on building Easel to close this gap and make design more accessible.
If you give it a try let me know what you think :)
tryeasel.dev/blog/introdu...
Yeah I've noticed the same thing. Originally I was pretty stoked about this platform.
But after taking a closer look at the people following me in the past few weeks, I'm convinced that most of them are either bots or have no interest/relevance to the content I post.
Curious about how Easel can save YOU time?
Easel simplifies UI development by turning your ideas into production-ready components in no time.
Start building faster today: tryeasel.dev/blog/introdu...
#buildinpublic #llm #uxdesign
The new Cline version is making waves, so I decided to put it to the test against Easel. Spoiler: Easel completes the same task in half the time β and Iβve got the video to prove it.
Check it out: youtu.be/qm5eep22Xjg
#buildinpublic #llm
@npmjs.bsky.social seems to be having issues.
`npm ci` is failing for me repeatedly.
Consequently, I can't push my release to production.
Anyone have any tips other than to just keep trying and crossing my fingers that it goes through this time?
#buildinpublic
bsky.app/profile/sean...
bsky.app/profile/sean...
@figma.com is having issues now too. Could this be related to the NPM service disruption?
bsky.app/profile/sean...
The active incident report:
status.npmjs.org/incidents/jl...
@npmjs.bsky.social acknowledges the issue. Worth noting that this has been going on *all day*
#npm #buildinpublic
bsky.app/profile/sean...
Success! I don't know if NPM fixed the issue, or if I finally got lucky, but we're in the clear.
or rather... it WILL be SOON... if you want to try the new features...
Once we can get past this NPM issue
By the way, this is a pretty substantial release I'm working on pushing out for tryeasel.dev. If you haven't given it a try yet, now's the time.
@npmjs.bsky.social seems to be having issues.
`npm ci` is failing for me repeatedly.
Consequently, I can't push my release to production.
Anyone have any tips other than to just keep trying and crossing my fingers that it goes through this time?
#buildinpublic
Do you envision that it will be usable with other languages as well?
@crowdhailer.bsky.social just read your post on building a structural code editor. Looks very interesting.
Have you considered any potential use cases of this kind of tool for accessibility and/or education yet?
Iβll add that the changelog was designed and implemented using Easel itself π
Screenshot of the new Easel changelog detailing the stated changes.
Two big improvements for Easel today:
1. Adjusted our model so that code generation is faster and more reliable.
2. Implemented a changelog so that our users can stay updated on all the new features we'll be releasing over the next few weeks.
#buildinpublic
Interesting that OpenAI is teasing o3 right now. The full o1 model was only released on December 5, around two weeks ago.
Intrigued to see how it compares, especially on pricing.