Long. Printers are pretty complicated machines, because they have to work with a natural product that shrinks, expands, folds, rolls itself up and sticks to other pieces of paper. I once heard a printer engineer explain that they use small puffs of air to lift the paper, but because there's also heat involved in the printing process that the paper sometimes rolls itself up or expands which causes jams etc. And I'm sure there's more going on.
Which isn't to say that HP aren't bastards.
If we shoot for a much more primitive printer, we’re pretty close.
Something that uses a pen or quill to draw on an unmoving sheet of paper. Kind of like how CNC routers are set up. The gantry moves along the full length and width of the paper.
After that, you can print everything outside the electronics and the quill. Right?
Everything shrinks and expands as you heat/cool it. It's called physics and it's a fucking mess. The more you learn the weirder it gets 😵💫
God damn physics some much for the immaculate plan. Why did god invent physics anyway?
/s
But wouldn't that mean that you already have a 3d printer that can print a 2d printer? As the 4th "D" is time?
The fourth D is right beside you, even if you can’t see it. It penetrates you everywhere.
There are actually a bunch of theoretical spatial dimensions, including 4th. Check out the HYPERCUBE!
Likely not for a long time.
3D printers aren't a magic make anything tool. They are a versatile and useful addition to our toolboxes.
You could likely print a lot of the parts now. Unfortunately, it would be the "vitamins" that would catch you out. Circuit boards are possible, but difficult. Silicon chips are currently impossible. Print heads would be almost impossible to print too.
Instead, I would expect someone to come up with a more general "auto fabricator". A combination of tools combined with robotics, and a standard set of "vitamin" components. Such a system is perfectly feasible (though not that soon) and could go from raw materials to a functional 2D printer. It could also make a kitchen blender, a new lamp, or whatever you decided you wanted (within reason).
An interesting take is the book series of the bobbyiverse, starting with "We are legion, we are bob". They play around with the limits of 3D printing, and how to go beyond them.
Tl;Dr 3D printers are awesome, but not a "do everything" tool.
If you have enough money, you can technically do so right now. The only thing you wouldn't be able to print assuming money was no object and access to certain tools was readily available are the magnets for the motors. At least I don't think... Can you melt a magnet down and reshape it and it still is magnetic? 🤔
There are metal fab printers and the way circuits are mass produced is also, essentially, 3D printing. You couldn't do a print-in-place type print and make a completed printer. Unless you combined everything into a single unit. Which would be expensive as fuck.
How much of this thing could be reprap'd?
If you're not going to click the link, it's an HP dot matrix inkjet from 1984 that...doesn't look that complicated to make.
Here is an alternative Piped link(s):
Piped is a privacy-respecting open-source alternative frontend to YouTube.
I'm open-source; check me out at GitHub.
It is possible right now. OP didn't say an electrical printer, it is 100% doable on old mechanical printing model.
You can start here: https://hackaday.io/project/176931-hp-printer-cartridge-control-module/details
HP printers are conceptually quite simple devices, the printer just moves the cartridge and the paper. The cartridge does all the actual printing. So you reverse engineer the pinout on the cartridge and you can make your 3d printer do normal printing. That's also how those little handheld cube printers work.
We are not too far from self-replicating machines controlled by AI. They will be built with a simple purpose: find resources and replicate.
Imagine a world with people trying to eliminate such machines, while machines learn new ways to get around any restrictions.
They will be built with a simple purpose: find resources and replicate.
Ah the Hartz-Timor Swarm.
What do you mean by not too far? Because I can see A LOT of issues which cannot be solved easily. Even with maximum wishful thinking and a lot of handwaving that's not something that's possible even by a long shot.
Or do you mean not too far, like within the next 1000 years? Cause yeah I could believe that.
Also AI doesn't mean general intelligence, so not intelligence like human intelligence. Don't be fooled by the PR and hype going around, LLMs aren't general intelligence.