Right, it seems the appropriate analogy is the shift from analog-photograph-developers to digital camera photographers.

I bumped this video to the top of my watchlist because of your comment, thanks for that. My takeaway is that the rocker-bogie is one of the best examples of an elegant passive solution I've ever seen.

I really like this pure-math way of looking at airfoil behavior: https://complex-analysis.com/content/joukowsky_airfoil.html

Considering things like Palantir, and the doge effort running through Musk, it seems inconceivable that this is not already the case.

I think I'm more curious about the possibility of using a special government LLM to implement direct democracy in a way that was previously impossible: collecting the preferences of 100M citizens, and synthesizing them into policy suggestions in a coherent way. I'm not necessarily optimistic about the idea, but it's a nice dream.

Thanks for the comment. Interesting to think about but I am also skeptical of who will be doing the "collecting" and "synthesizing". Both tasks are potentially loaded with political bias. Perhaps it's better than our current system though.

> special government LLM to implement direct democracy

I like your optimism, but I think realistically a special government LLM to implement authoritarianism is much more likely.

In the end, someone has to enforce the things an LLM spits out. Who does that? The people in charge. If you read any history, the most likely scenario will be the people in charge guiding the LLM to secure more power & wealth.

Now maybe it'll work for a while, depending on how good the safeguards are. Every empire only works for a while. It's a fun experiment

Sounds like Helios https://www.youtube.com/watch?v=swbGrpfaaaM

Centralising it is definitely the wrong way to go about it.

It'd be much better to train an agent per citizen, that's in their control, and have it participate in a direct democracy setup.

Indirectly, this is kind of what I was trying to get at in this weekend project https://github.com/stewhsource/GovernmentGPT using the British commons debate history as a starting point to capture divergent views from political affiliation, region and role. Changes over time would be super interesting - but I never had time to dig into that. Tldr; it worked surprisingly well and I know a few students have picked it up to continue on this theme in their research projects

That looks very interesting. Could use a demo or examples for us short attention spanned individuals. Would be cool to feed it into TTS or video generation like Sora.

Real world LLM's cannot even write a proper legal brief without making stuff up, providing fake references and just spouting all sorts of ludicrous nonsense. Expecting them to set policy or even to provide effective suggestions to that effect is a fool's errand.

>Real world politicians cannot even write a proper legal brief without making stuff up, providing fake references and just spouting all sorts of ludicrous nonsense. Expecting them to set policy or even to provide effective suggestions to that effect is a fool's errand.

This has been a more realistic experience of the average American for the past few years.

After all, what is a sandwich but a stack of food?

You’d be surprised in the variety in how people eat food.

I know some people that roll a pizza slice (from crust to center) to eat it. Blasphemous, and inspiring.

I think they're the same? both are built layer by layer but consumed in vertical chunks, right?

> you don't eat a sandwich layer by layer

Some sandwiches naturally want to be eaten from the middle layer out.

It’s vertically integrated.

If I'm working in a dark room, then light mode is eye strain hell. With dark mode, the minimum brightness I can achieve is about 100x lower than with light mode.

Personally, I find treemaps unmatched for disk space analysis. Specifically, I like to use the squarify layout algorithm, to NOT use the "cushion gradient" shading method, to use inset frames to convey depth visually, and to include filenames. This maximizes glanceable information density, for the use case of identifying large objects to delete to recover space.

This is how the old spacemonger app worked, and I liked it so much I had to recreate it for Linux/Mac: https://github.com/alanbernstein/treemonger. My version still needs some work, but it's minimally useable.

Treemaps are also good for profiling (see KCachegrind), they waste a lot less space than flamegraphs and the area-relationship is relatively well maintained.

Why do treemaps suck?

It's not as straightforward to compare area as sorted length.

Look at the example in the link and try to make sense of it.

I do agree that both styles of treemap shown in the article are inadequate for various reasons, but I don't think that applies to treemaps as a whole.

Before reading the article, I interpreted the quotation marks in the headline as addressing this exact issue. The author even describes dark patterns as a product of design.

For an LLM which is fundamentally more of an emergent system, surely there is value in a concept analogous to old fashioned dark patterns, even if they're emergent rather than explicit? What's a better term, Dark Instincts?

I have a lot of experience with openscad. I design a lot of small, simple-to-moderate parts. I want to use something like fusion instead, but the GUI learning curve is a huge blocker for me.

I'm not a CAD professional, I can't seem to find the time to watch hours of video to get the basics down. With code based CAD, the way to start is usually obvious. When I run into a blocker, I search online, find examples, and try them out. Then, crucially, I copy and paste code snippets into my design, and modify them, to solve my problem.

In a GUI cad tool, I find that I spend most of my time hunting for buttons in the UI, often finding UI layout discrepancies between my version and whatever video I found.

In code, I do have to repeatedly solve little trig or geometry problems, and I'm always aware that a constraint based GUI tool would eliminate that completely. But I always know that I can just spend five minutes with pencil and paper and get it done, whereas switching to fusion means adding an hour or more of work to multiple designs.

I really want to design more effectively, with better fillet flexibility. But for my simple tasks, the barrier to becoming productive in a GUI is just too high.

I believe the "command palette" in e.g. VS code solves this well, perhaps a GUI cad with that would be workable for me.

Maybe what would really help me, is a larger more complex project which I can develop over a longer time in fusion while I learn to use it. Too bad I don't have anything that naturally fits that bill.

To me it sounds like it would be worth it for you to learn the basics of Fusion or FreeCAD. You would probably quickly recover the hours spent on learning with increased productivity.

I recommend going through some basic tutorial (written or in video form) to build a simple part. The tutorial should teach view navigation, drawing and constraining a sketch, extruding or rotating it to create a 3D body, modifiying that with chamfers / fillets, creating sketches on top of that to add or cut away parts, add holes, create patterns from features. I don't think you need to learn surface modeling at this point. After that you should be good to go on your own projects. You will still need to look up how to do something (as you do now), but that will improve quickly.

I'm hitting the same problem, getting stuck on simple things in FreeCAD mostly because I'm a novice and don't have hours and hours to watch videos and learn.

What ended up working for you?

I'm generally very much for open-source software, but after my brief experience with FreeCAD and Fusion, I think if I'm going to sink 10+ hours into learning one of them, it will be Fusion. I bookmarked this recently and plan to try it out but have not yet started: https://www.youtube.com/playlist?app=desktop&list=PLrZ2zKOtC...

> In code, I do have to repeatedly solve little trig or geometry problems,

BOSL2!

I adopted bosl2 recently, it's wonderful! Anchoring is a huge improvement, which does eliminate a lot of that work. But it's a very large library and I am slowly expanding my knowledge of it.