This wasn't obvious a year ago, but today CAD literally reduces to Simon Wilson's pelican test, since CAD is largely a matter of functional CSG, and CSG is really not that different from SVG. It's just one more dimension, which it turns out is not a problem.

LLMs consistently one-shot CSG based video game levels with interesting physics puzzles (citing myself). Given this I'm willing to conclude that the frontier models are good at automated CAD if given the correct harness. But I guess a lot of people don't know this yet.

[1] https://github.com/elalish/manifold

Show me the blog posts where people talk about the results they got "vibe coding" and those arXiv papers look great in comparison!

There is the insidious thing with LLMs is that they can get the general shape of something right but that thing will not be useful if the last 1% is wrong. It might be that the operator sees the problem and fixes it, but it may also be that the LLM hypnotizes the operator into not seeing the errors and gaps.

I know there are many sorts of problems where I've had good experiences with LLMs but I know other people have had bad experiences and some of it might be my skill but some of it is just plain luck.

Neck: https://pastebin.com/Sg3LmmUq Body: https://pastebin.com/FE9nikYB

edit: screenshot, too https://i.imgur.com/FZGyyVO.png

Injecting a natural language layer into the workflow is just not optimal. CAD itself is not a difficult tool to learn and use effectively. There are essentially no layers of abstraction that an LLM can assist in cutting through, and no obfuscated rules or languages to learn.

I think of it this way. If there was someone sitting at my computer, and I had to do all of my CAD design by explaining what I wanted them to do verbally, I'd rip out my hair.

LLMs are doing for programmers what virtual CAD did to the drafter 35+ years ago, optimizing the effort expanded to create the thing already in your brain

Isn't this how a lot of machine shops operate, or how things operate internally in larger manufacturing factories? Customer/person-from-different-team comes in and explains what they want to do. Maybe they have some sketches or pictures of similar parts. Then there is back and forth with the CAD guy to build the thing.

One critical difference is that the CAD guy is usually smart and you have to explain to them things at a more high level, along with some written down hard numbers that need to be obeyed.

For example let's say I need a parametric model of an involute gear with a 14.5 degree pitch angle and a 0.5 inch hub with an 8-32 set screw where the OD, number of teeth, face length, and shaft diameter are all variable. It takes seconds to write the prompt and then adjust the variables to what I want whereas it would take tens of minutes to actually model even for a highly skilled drafstman.

The time savings get more extreme the more the modelling work is looking up information online. For example if I wanted a model of a lightbulb electrical connector, 95% of the work is just going to be googling. Technically you could have an LLM just tell you all the dimensions you need and then model it yourself, but that's definitely still going to be slower and you have to put in the effort to figure out what each dimension refers to. It makes sense just to cut out the middleman.

Add in the fact that CAD is in fact a difficult tool to use and learn effectively. There is a subset of the population that is very good at picturing and orientating objects in 3D space in their minds, and engineering is mostly limited to these folks. For those whose minds work differently, CAD is extremely tricky to learn. An alternate method of interacting with models that does most of the heavy lifting such that a person only needs to tweak a near complete model could be extremely helpful to many people who would like the benefits of CAD design but have struggled to learn the software. It's no different than AI generated art opening digital art to those who are not good with digital artistry software.

Download model, done.

Not attempting to be sassy, sorry, but maybe I can explain my point a bit better. If you are making something complex enough to warrant you yourself designing it (something unique, custom, new), then CAD on its own is the best tool (right now!). I have models with thousands of parametric features linked to hundreds of other parts in an assembly. I would need to write a novel to achieve those same features with an LLM.

If what you want to make is simple enough that an LLM would be able to nearly one-shot it, then that model either: 1. Already exists somewhere (like the gear above) or 2. Is simple enough that a couple hours of training videos will give you all the skills you need to make it yourself

Note how that is not a parametric model with the appropriate variable parameters nor does it have the set screw I specified in my prompt. Even if precisely the model you want has been made by somebody else, it still takes time and effort to hunt it down. By the same logic who needs to generate 2D images when stock photos are a thing?

> If you are making something complex enough to warrant you yourself designing it (something unique, custom, new), then CAD on its own is the best tool (right now!).

I make CAD models every day, professionally. Most are not complex at all. Unique, custom, and new don't imply any level of complexity, and none of the above correspond to difficulty to describe. As a simple example from today, I need a funnel with a specific volume and sizing for the inlet and outlets. No one makes it, there is no model I can download, it is unique, custom and new. The modelling itself is not hard but the tricky bit is doing the math to ensure I get the right volume without compromising the geometry. The CAD software solves the equation but I had to spend time and effort to figure out the equation to type in. This is exactly the sort of task that LLMs excel at.

> I have models with thousands of parametric features linked to hundreds of other parts in an assembly. I would need to write a novel to achieve those same features with an LLM.

Did you hit a button to create that whole assembly and form all those links with a single short action, or did you spend hours if not days meticulously making or modifying models for each of those parts then manually link them each together to create those relationships? You put in the time and effort to write a novel, you just wrote it in a different language.

> If what you want to make is simple enough that an LLM would be able to nearly one-shot it, then that model either: 1. Already exists somewhere (like the gear above) or 2. Is simple enough that a couple hours of training videos will give you all the skills you need to make it yourself

You've got it completely backwards. CAD tools are well set up for doing standard work, where LLMs would excel would be where you're departing from the normal to do more complex stuff. Simple example: let's say you want a 10x10 grid of 1/4-20 threaded holes, most any cad software you can do that in under a minute. Now let's say you want 100 1/4-20 threaded holes arranged in a spiral pattern where the distance between each pair increases by 1% as you move outwards. Is that doable in CAD? Most certainly, but it's gonna be a pain in the butt. Does a model exist on the internet? I doubt it, maybe for some generic flat plate but definitely not the part you need this feature on. On the other hand, typing out a prompt to describe it takes virtually no time at all. It's unique, custom, new but much like my funnel it isn't complex, at least not conceptually.

For your funnel, it seems mainly the problem area is a volumetric calculation, not CAD. A simple funnel would take 1 sketch and 1 revolve. I guess it'd come down to if you'd prefer to type all your variables in a text prompt vs in the CAD dialogue box.

Otherwise, I think mainly the point where we disagree is that it is easier to specify mechanical designs in natural language vs the typical inputs in CAD.

Let's try the spiral pattern example. Which direction is the spacing increasing in distance? Radially? Hole to hole along the spiral pattern? How many instances establish the radiating arms of the spiral? Which direction is the spiral going? Is this a planar pattern? What is the radial pitch of the spiral? How many extra details would I need to type out into my prompt in order to achieve the precise results desired, and how many re-tries?

Vs: Fill out this box of variables. https://i.imgur.com/juDH435.png

I, on the other hand, have used LLM + OpenSCAD to design stuff - while I pulled my hair out everytime I had to sit and write OpenSCAD primitives or use a UI CAD like FreeCAD or Fusion360 and their horrible unintuitive interfaces.

you could sculpt a model à la ILM for star wars or the architecture models, but the only way to have a copy of the object was to make one.

the virtual cad also brought in the ability to do analysis with FEA and get approximately smooth undestandings of the stress and strain on the piece, rather than manually calculating the critical points and stress raisers for doing analysis

Not in an informal way. But from a technical perspective, of course it does: serialize the feature steps to text or to code, job done.

Modifying the prompt and then trying it again did not lead to that self-verification loop and the output was unusable garbage.

https://figure.codes