Notes written while the work was still open.

Funny how one can think of a technical problem as easy to solve with these things, but it might not be as easy. But it’s always a good learning.

I had this Hidden Bar app on my menu bar for a while. Lovely little thing to unclutter that area up above. One day it was using tons of memory? Why? And, just as one does, asked Claude what was going on. “Memory leak, open source project, PRs open with proposed solutions for years, nothing merged”. Bummer.

I mean, I could just fork the thing and try the PRs myself. But hey: what if we built the same thing, from scratch, but ✨leakless✨? Should be just a prompt away, right? So I did that and had Claude write a monitor for it. And mine leaked too, I mean not as bad but, still.

But I really want something like this on my screen so, off to iterate I go. Worst case this becomes a cautionary tale or something.

One more of those quick ideas that I don’t want turning into a multi-day side thing but they do.

I was thinking, “hey what if there was one of those curated design sites, but for Temu?” Not my worst idea, I guess. Anyway, one could turn their ugly pictures into a single aesthetic and populate a site with those. Good picks, and make em look good. Slap some referral link to make a few bucks. Bam.

Fable one shotted the thing end to end which impressed me. Then I saw this video of a guy talking about the “Habitus” concept as Bourdieu defined it, which I thought fits well here. Mocked up a logo and slapped it at the top. Not very minimal but hey, makes it look different from other curated design sites. Will try more logos and see if anything works better. Now I also have to see how that referral link stuff works there.

Worked today on mockups for the interface. We had some older web demos but tbh that doesn’t make sense anymore. I’d rather prototype something that looks/works closer to what the final device will be. Idea is to clone that demo to the laptop and use it as if it were a POiO. We’ll see how it goes.

I don’t have it perfectly planned still how I would approach pantry inventory, feels like that will be heavy friction. Might have to do image/audio feedback to solve that? I can imagine I just press a button and say “hey we out of parsley but I got some chives” which should be ok, but want to test the more minimal version for now and see how it works. Mics not planned yet anyway. So it goes.

I was discussing something with Claude, about lavaliers, and got this answer, which feels quite close to the thinking that permeates Branding Oracle.

Me: We used keyboards to talk to computers but now that they’re getting smarter people use voice more and more. One day we’ll be able to translate our thoughts into tokens and this will be remembered as a strange middle point when we spoke out loud to them.

Claude: Notice the verb you reach for: “translate.” It assumes the thought is already there, fully formed, and the medium is just a lossy pipe slowing its arrival. Strip away the pipe and the thought passes through clean.

But consider how often you don’t know what you think until you’ve typed it — or said it aloud and heard the shape it took. The friction may not be delay. It may be where the thought gets made.

If that’s so, then what exactly would go straight to tokens? Is there a finished thought waiting behind the words, or does the speaking call it into being?

Been thinking about this one for a while. Have too much on my plate atm and I have to learn to prioritize work. Feels like it’s one of those things that will drift away and won’t touch it for a while, but in the meantime I just wanted some public record of it. It does have potential, but of course requires quite a bit of time and even investment to get it going. So that’s that.

For the record, here’s another (early) idea of what this might look like. It would be hard to achieve that glow, though?

Decided to put down some raw ideas here, if anything just to keep track of my thoughts around this.

Saw Google announcing that Coral thing just the other day. Makes me think a future with local smart devices here and there around the house ain’t that far off. Exciting and scary.

I mean, imagine a loud squawk from a thing on your kitchen counter mid-breakfast. Gotta keep the future a bit weird, though.

I’ve been going around the idea of squared vs. horizontal screen for this. Was leaning towards squared, but some quick mockups have me preferring horizontal now. Might be more practical. Gotta speed up the hardware prototype. Mocking up 3D renders for the case feels more like me getting artsy than solving the goddamn thing itself.

There are some UI prototypes we mocked up for this, haven’t really had time to go over them. I’d like to put down some thoughts on paper for that as well. Going over recipes should feel better than going over a plain recipe book. This needs a bit of animation, a bit of soul; wonder if the refresh rate would be enough to get playful.

The 3D mockup has me sort-of satisfied. Not fully, but it’s one of those things you gotta sleep on, I guess. See it for a while, check if the weird angles feel less or more weird over time. Still haven’t modeled the buttons, but this screenshot suddenly has me preferring it button-less? Don’t want to go that far for this first version, though.