Listening to another pitch about how AI can empower workers at various jobs across my industry, I was striken by the comparison in the title
3d printing, just like generative models, have it’s actual niche uses, where it’s obvious downsides are irrelevant and they come handy, e.g. prototyping, replacements, small-series production
Where it comes to the top-down AI promotion trend, it feels not unlike the idea of printing the whole product - a car, or a house, from the smallest details - applying the least effective method, doomed to have a worse than average outcome due to technological limitations
And screws, the thing that we nailed down long before, and that is completely incompatible with that mode of production, is a screaming, growling, shrieking example of how helpful tech can be mispurposed in the most stupid way
A note that Aerospace has absolutely leaned into 3D printing metal parts since they can make parts much lighter and be just as strong.
The main use of LLMs, as far as I can see, is to replace the very people pushing it on everyone.
That’s about being willing to eat the nre for one-off special purpose parts that have geometries not attainable by casting or machining.
3d printing is not the default fabrication method now that we’re getting good at it. It just shines in certain applications.
AI is often pitched as being able to do anything, eventually. We even try to use passive fail safes over active ones, a century after electricity became commonly available. Because that will always be a better solution by the nature of the options. AI is the same way. It is a distillation of human English language that is written. Why you would think that that could eventually replace all software developers, or any other field that produces text as its output makes any sense is beyond me. I can’t see how that could be true.
Exactly. It’s not true. Any company that fires all of its developers and sets up some poor intern to prompt-engineer updates to their codebase is going to fail spectacularly.
Source: I’m a software developer and use LLMs regularly. There are certain tasks they are very good at, but anyone who commits unexamined code generated by an LLM gets exactly what they deserve.
Ok, im a hardware dev. They’ve tried to make us do software style project management every time there’s a new fad (agile last time). It usually doesn’t fit.
What do you find them useful for in your role? Like a coding partner, you can ask questions? Or linting? Im at a loss in my role. I need to know the proprietary code base to write a single line of value. We aren’t allowing anyone to train an ai on our code. Thats a huge security problem if anyone does.
So there are a few very specific tasks that LLMs are good at from the perspective of a software developer:
And that’s… pretty much it. I’ve experimented with building applications with “prompt engineering,” and to be blunt, I think the concept is fundamentally flawed. The problem is that once the application exceeds the LLM’s context window size, which is necessarily small, you’re going to see it make a lot more mistakes than it already does, because - just as an example - by the time you’re having it write the frontend for a new API endpoint, it’s already forgotten how that endpoint works.
As the application approaches production size in features and functions, the number of lines of code becomes an insurmountable bottleneck for Copilot. It simply can’t maintain a comprehensive understanding of what’s already there.
I use it to generate unit tests, it’ll get the bulk of the code writing done and does a pretty good job at coverage, usually hitting 100%. All I have to do for the most part is review the tests to make sure they’re doing the right thing, and mock out some stuff that it missed.
Legit. Do you need to feed it your code base at all? How does it know what needs to be tested otherwise?
yeah, it’s the copilot plugin for intellij, basically right click and choose generate tests, it’ll read the file and … well…
Downside to that approach is that it doesn’t know what some function calls do if they’re not part of that file, so it tends to miss places that need to be mocked out.
Occasionally it writes a test that’s “wrong”, and I have to fix the test… very rarely, the “wrong” test is actually “right” based on say a method signature or decision tree, and the method itself needs changing.
You’re right, unit tests are another area where they can be helpful, as long as you’re very careful to check them over.
one other use case where they’re helpful is ‘translation’. Like i have a docker compose file and want a helm chart/kubernetes yaml files for the same thing. It can get you like 80% there, and save you a lot of yaml typing.
Wont work well if it’s mo than like 5 services or if you wanted to translate a whole code base from one language to another. But converting one kind of file to another one with a different language or technology can work ok. Anything to write less yaml…
They are getting faster, having larger context windows, and becoming more accurate. It is only a matter of time until AI simply copy-cats 99.9% of the things humans do.
Actually, there’s growing evidence that beyond a certain point, more context drastically reduces their performance and accuracy.
I’m of the opinion that LLMs will need a drastic rethink before they can reach the point you describe.
We have 100M context AI, we just need better attention mechamisms.
This sounds to me like saying you have enough feathers in the grocery bag you’re holding. All you need now is a beak, and you’ll make yourself a duck.
X doubt
Getting a little theoretical here
With the current state of the technology, 3d printing lags behind some traditional manufacturing techniques like machining and in terms of speed, cost, quality, available materials, etc. except for some relatively niche cases.
However, that gap is closing a bit every day, it may or may not ever catch up completely or surpass the old technique in those aspects
But if it does ever get close, I could very much see 3d printing being a preferred method
Subtractive manufacturing like machining, by design, creates a lot of waste, all of the chips and off cuts that are removed from the stock are either discarded or require additional energy and/or materials to recycle.
And things like injection molding require custom molds that wear out over time, and can be expensive to design and manufacture
And in either case, you’re largely locked into making one thing on an assembly line at a time, and to switch over to a different product you’re probably going to need to switch out a lot of the molds and tooling, recalibrate everything, etc. which can be time consuming.
With 3d printing, you could theoretically use only the amount of material that’s actually in the finished product (if you design it that it doesn’t require any external supports ) you don’t need any custom tooling or mold, just generic, interchangeable nozzles (for FDM, LCD screens or lasers or whatever the equivalent is for other printing technologies) and you could switch production from one item to another by just hitting print on a different file.
Again, we’re not there, may never be there, but it’s a cool thing to think about
There are a lot of things that FDM printing will likely never be better than say injection moulding, and the main thing is speed, as in quantity over time. A single 3d printer might be able to make a plate full of maybe a dozen widgets in a few hours, and in that time, the injection moulding machine will have tens of thousands produced at a higher quality.
On top, 3d printing would require more staff to troubleshoot, clean, re-start prints, remove scaffolding from finished items, sand/polish to remove the layer lines, etc.
What it’s great for in an industrial setting, is prototyping. For example, a case for something can be printed, and the plate can be filled with several variants. If a flaw is found or changes needed, then a new batch can get whipped up on the same printer. Once a design is found that is acceptable, the CAD drawings get sent to have moulds created.
Yeah injection molding blows 3d printing costs out of the water if you reach above 10000 units production. Below that 3d printing can make sense, its just not as efficient time wise. You also have to deal with filament parts having different stelrength in different print orientations.
Yea, should say: FDM 3D printing a screw with PLA vertically in vase mode.
But thats a bit long winded.
Would that not be ironic indeed? Imagine they go for world domination and extermination of their slaves, only for that tech to be useful only for replacing them, and now it’s open-source.
The “unwashed” use it gladly, and our main problem is eliminated.
They have but its only for parts that warrant it by combining multiple pieces into one part, or a part that is highly complex to machine. Because its an expensive and time consuming process to print, wash, sinter, and hope the warp is still in spec, and still requires machining for features and faces that mate to other parts