Well that sounds like an exciting idea — imagine a car tree that grows on its own, and we need not do anything other than pick the ready-made cars off it. Perhaps we can breed faster models, or cross-breed to make new ones — in all seems like a lot less effort than having hundreds of engineers and designers spend hours planning out every detail. Ok, but what exactly do we mean when we think of growing cars, and how is it different from how they are currently built at manufacturing plants? (I f**ing love that “plant” is already the word we use for factories!!! I think there is some deep insight there)

To dig a bit more into the distinction here, imagine this scenario: some aliens come to Earth that are so different from us that they cannot even identify humans as clearly distinct from the matter of our environment. Imagine that they don’t see us as distinct living organisms, let alone as individuals with unique personalities and interests, or anything like “consciousness.” They merely see lots of biochemical processes that lead to more biochemical processes, all intertwined, interacting, and producing some plastics. Some of these plastics combine with bits of metal into cars, as catalyzed by some more biochemical processes (which we call humans, but they don’t). From this perspective, a manufacturing plant will look to them precisely like a tree that grows cars, all on its own! Furthermore, if they tried to control what sort of cars it produces, they wouldn’t think to talk to workers and change design plans — instead they might try to cross-pollinate: bring two different manufacturing plants close together and see if they learn to produce some hybrid model of the cars.

Ok, I can see a few holes you can poke in this line of reasoning — please do so in the comments! Even so, I think the overall point holds: we can construct a reasonable “outside perspective” from which manufacturing plants will fit all the criteria above for “growing cars.” Do you buy this? What other distinctions do you see between “building” and “growing” cars that I missed in the above table? Finally, if we still want to invent a way to “grow cars” (from our perspective) what would be the main way we would know that we succeeded (our KPI)? 
I’ll continue this discussion next week, with some brainstorming (read: BS) of how we might actually achieve this in the future ;P

New to LessWrong?

New Comment
4 comments, sorted by Click to highlight new comments since: Today at 3:02 AM

Isn't this horses?

It sounds like you want a way to set up a system which autonomously creates a product that fills some certain parameters that you set.  In this case you want self building/designing cars.  

I feel like the alien perspective already answers your question of how.  Car manufacturers are already these self modifying car building systems, with humans as an integral part.

I don't think we can get to "growing cars" without humans as the substrate until something else is able to fill that role such as AI, or in the case of horses, natural selection.

Building implies giant metal presses. Growing implies that if you watch it, you see a car shape form with no obvious outside forces. 

So Imagine we develop nanotech. The cars are still designed by humans. You can have a "seed" in your pocket that, when planted will absorb surrounding material and grow into a car. 

Now lets bring in some AI. Humans design the rough criteria of the car, how big it should be, what power sources it should accept and so on. AI fills in all the technical engineering details. 

This is quite like growing the car from our perspective.

EDIT: I now see you research these questions and so want to add a disclaimer that I have not thought about these things nearly as deeply as you probably have...

Epistemic status: very speculative.

Cool post, I've long been fond of the, likely less difficult, thought experiment of whether we can grow a house using synthetic biology.

At first, I was thinking growing vs. building was just about the amount of labor involved to go from raw materials to final product. Then I realized this doesn't work because under this definition a fully automated robot factory would qualify as growing a car.

My next best guess is that "growing" is related to:

  • the system containing its own description,
  • the system maintaining itself given "raw" materials, and
  • an aesthetic component that connects growing to things that look and feel biological.

In terms of KPIs, the first things that come to mind are metrics like:

  • How small a seed can the system bootstrap itself from given raw materials and otherwise little to no outside intervention?
  • Can the system repair itself when damaged?

Thanks for all the great comments! - I feel like the follow-up post I just published gets at some of them: https://www.lesswrong.com/posts/WNjKyFxNbhonBvhwz/building-cars-we-don-t-understand