That’s a really good question! So programmable matter is one of the more exotic ideas we got into. It's actually fairly hard to research, as it’s a small group of people who don’t agree on all the technical terms. So we called it “programmable matter.” One of the really useful books we found was a book by Springer Publishing (something that will mean something to a small group of people) called MORPHOGENETIC ENGINEERING. I don’t even remember how we found it! Sometimes it’s called Morphogenetic Engineering, sometimes it’s called “self-assembly,” there’s fields like swarm robotics that are deeply related.
The most plausible case I think we saw with the idea of programmable matter, stuff that can re-shape itself, was one: in the book we talked about a proposal out of Daniela Rus’ lab out of MIT. Apparently something like 3,000 Americans a year get a watch battery lodged in their guts. I assume it is mostly children...but if you have enough people, things happen. And so the problem is most of the time you pass it through, but your body is not designed to deal with metals like that, it can lodge in your gut, irritate the skin, and is obviously dangerous. It would be nice to not have to do surgery to get it out.
The basic idea would be to have a little bot, that folds up nicely into a little pill made of ice. It makes its way to your gut, and the little robot (made of sausage casing) unfolds itself, with a magnet that lodges onto the battery, the idea being that it swims its little origami fins and makes its way out through conventional means. And the robot isn’t too dangerous since its made of sausage casing. It’s solving a narrow problem. But imagine all these origami nanobots to do different tasks inside the body - to receive signals, to grab onto something, to have a compartment with medicine that could be delivered to targeted areas.
So potentially , this is futuristic stuff, but imagine - if you have medicine and it flushes through your body, you might not want to receive it that way, versus if you had nanobots that delivered it to a targeted area to cut down on side effects. For things like cancer therapy, you could limit where it goes to help provide more benefits. There’s a contingent nature to technology - you never know what’s going to provide the breakthrough. It could be something unexpected in software or material science that makes this more feasible tomorrow. Google used to do statistical research methods to do translation, and then in 2010 they added machine learning stuff, and it got better overnight. So predicting is dangerous.
The other thing to think about is: is there a market for improvements? For example, in the beginning of the book we talk about space elevators, a cable which goes to space hanging from an object suspended in space and you climb up the elevator, which would make space travel a whole lot cheaper if it works. The problem is the cable has to be made from really exotic material.
The current top recommendation is carbon nanotubes - think of them as Superman’s hair. Smaller than superman’s hair, but extremely strong. They are both strong and lightweight (weight is another factor with a cable that long). So part of the reason is you’d want one perfect tube of carbon nanotube 100,000 km long - well, several perfect tubes. And part of the problem is economic. The longest carbon nanotube as of 2018 was about 1.5 feet, half a meter, long. Well-shy of 100,000 km. But you can imagine a world where there’s an economic motivator for better carbon nanotube development, and what you might see if that happens would be like what happened with personal computers - starting in the 1940s as ultra-primitive machines, but there’s an exponential growth rate of them getting better year after year.
And so if you have that for nanotubes, without caring about space or space elevators, you can arrive at the right technology. We read a textbook about space elevators from 2013, and they had a graph showing “how good are we at growing good carbon nanotubes over time” and based on that graph’s rate of improvement, by maybe 2040 or so we’ll be able to have ultra-long carbon nanotubes? But so far we haven’t got any additional data points. So if you want to see a space elevator, find a use for ever-longer carbon nanotubes and maybe we’ll get it!
So going back to your question about when we’ll have tiny nanobots, we don’t know the contingent nature of technology - obviously there’s a great market for slight improvements on it, but it looks like an “all or nothing” thing. What you have to find is a market for small improvements. If you’re looking at 2 phones, and one phone has a 5% battery life or a 5% better screen, you’ll buy it. You’ll pay an extra $20 for it. Similar to electric cars. People are willing to pay a LOT of money for a little range. So if you look at battery cost and range, they’ve both gone down drastically in the last 10 years. So for any given technology, you have to identify what the iterative improvement that makes money for a developer is.