James Bond can be the Dread Pirate Roberts.

Oh, it took me a while to parse what the screenshot was doing, but I guess it'd be useful in a ML-lang--since stuff is functions, and you often want to know what's being passed inbetween.

At first glance, also not sure if colors help inline. What surprised you so far about having the traces?

This something else is near-term customer need that fits within the constraints of unit economics. Most innovator builders never address this because they don't care about it. Hence, early majority builders make up their own solutions. Then when it's adopted, innovator builders lament.

Funnily, I think I'm writing something related, but on the other end. I think early movement rallying cries like "decentralization" works on innovator builders, but fails to resonate with early majority builders. They then can take the movement's ideas piecemeal, in service of something else.

Which parts are missing? Noticeably the read/write aspect. But are you referring to other parts, or specific aspects?

Looks cool: 1) what are you trying to show?

2) how did you decide on the colors?

I should have known better than to ask.

Part of making good decisions is to spot viable options. AI currently has a hard time with this, so a) you need to spot when it happens and push the AI to look where it doesn't usually. b) notice when it happens and lean into struggling.

I think the trick is to know the level of question you're asking. If you're asking a beginner question, then the options are likely pretty good. But if you're pushing boundaries, then its options are likely only under the lamplight. There's probably a better one in the bushes.

The worst trio AI gives are: Option A, Option B, and Hybrid Option A/B. And then it recommends the Hybrid. That's when you should really push it to go back to the drawing board.

a) It's hard generating good options. Hence, in moments of struggle, you turn to easy. b) AI options feels like a slot machine: you want to pull the lever and see what comes out.

Have you debugged the source of the difficulty?

Are you just pointing out it happens? Or you have a stance that it shouldn't happen, or that it's inevitable? Or something inbetween that it happens with caveats?

Does the paper go over the mechanism by which the regularizationalist bring technological innovations pushed by the reconstitutionalists back into being tools of regularizationalists? I have the paper open in a tab, just haven't read it yet.

That inevitably disappoints the original advocates. They didn’t have these other constraints. They tend to be more future looking or care more about an ideal in the near future than how to bring it about today.

And sometimes, that means giving up some of the original ideology in order to scale up/make the economics work/help people make progress toward goals. The builders that bring it to scale are pragmatists first, and are willing to cast off anything that won’t work today.

What kind of liberation? My limited view is that these movements, in order to be adopted more broadly, needs to be adopted by builders. And builders have to work within unit economics. They may have to partially adopt or morph the original movement in order to get it to work more broadly.

I hope you leave stuff up in the future. I was researching React/Solid arguments for and against, and it was hard doing archeology on the topic since I kept running into your deleted posts.