You meet the end of a long day in your apartment in the early 2040s. You did a good job and decide to take a break. “Movie time!” You say. Home responds to your urges. The table splits into hundreds of tiny pieces that crawl under you and take the shape of a chair. The computer screen you were working on spreads over the wall and turns into a flat projection. You relax in an armchair and in a few seconds you are already watching a movie in your home theater, all within the same four walls. Who needs more than one room?
This is the dream of those working on "programmable matter".
In his latest book on artificial intelligence, Max Tegmark distinguishes between three levels of computational complexity for organisms. Life 1.0 are single-celled organisms like bacteria; for her, hardware is indistinguishable from software. The bacteria's behavior is encoded in its DNA; she cannot learn anything new.
Life 2.0 is the life of people on the spectrum. We are kind of stuck with our equipment, but we can change our own program, making choices in the learning process. For example, we can learn Spanish instead of Italian. Like space management on a smartphone, the brain's hardware allows you to download a specific set of “pockets,” but in theory you can learn new behaviors without changing the underlying genetic code.
Life 3.0 moves away from this: creatures can change both the hardware and software shells using feedback. Tegmark sees this as a true artificial intelligence - as soon as he learns to change his base code, there will be an explosion of intelligence. Perhaps thanks to CRISPR and other gene editing techniques, we can use our own "software" to modify our own "hardware."
Programmable Matter carries this analogy to the objects of our world: what if your sofa could “learn” how to become a table? What if, instead of an army of Swiss knives with dozens of tools, you had a single tool that “knew” how to become any other tool for your needs, at your command? In the crowded cities of the future, houses could be replaced by apartments with one room. This would save space and resources.
Anyway, these are the dreams.
Since it is so difficult to design and manufacture individual devices, it is not hard to imagine that the things described above, which can turn into many different objects, will be extremely complex. Professor Skylar Tibbits of MIT calls it 4D printing. His research team identified the key ingredients for self-assembly as a simple set of responsive building blocks, energies and interactions from which almost any material and process can be recreated. Self-assembly promises breakthroughs in many industries, from biology to materials science, computer science, robotics, manufacturing, transportation, infrastructure, construction, arts and more. Even in cooking and space exploration.
Promotional video:
These projects are still in their infancy, but Tibbits' Self-Assembly Lab and others are already laying the groundwork for their development.
For example, there is a project for self-assembly of cell phones. Creepy factories come to mind, where they independently assemble mobile phones from 3D printed parts around the clock, without requiring human or robotic intervention. These phones are unlikely to fly off the shelves like hot cakes, but the cost of production for such a project will be negligible. This is a proof of concept.
One of the main obstacles that must be overcome when creating programmable matter is choosing the right fundamental blocks. Balance matters. To create small details, you need not very large "bricks", otherwise the final design will look lumpy. Because of this, building blocks can be useless for some applications - for example, when you need to create tools for subtle manipulation. With large chunks, it can be difficult to model a number of textures. On the other hand, if the parts are too small, other problems may arise.
Imagine a setup in which every detail is represented by a small robot. The robot must have a power supply and a brain, or at least some kind of signal generator and signal processor, all in one compact unit. You can imagine that a number of textures and tensions can be modeled by changing the strength of the "bond" between the individual units - the table should be slightly harder than your bed.
The first steps in this direction were taken by those who develop modular robots. There are many groups of scientists working on this, including MIT, Lausanne and the University of Brussels.
In the latest configuration, a single robot acts as a central decision-making department (you can call it the brain), and additional robots can join this central department as needed if the shape and structure of the overall system needs to be changed. There are currently only ten separate units in the system, but again, this is a proof of concept that a modular robot system can be controlled; perhaps in the future, small versions of the same system will form the basis of components for Material 3.0.
It’s easy to imagine how these swarms of robots learn to overcome obstacles and respond to changing environments more easily and faster than a single robot using machine learning algorithms. For example, a robot system could quickly rebuild so that a bullet passes without damage, thus forming an invulnerable system.
Speaking of robotics, the shape of the ideal robot has been the subject of much debate. One of the recent major robotics competitions hosted by DARPA, the Robotics Challenge, was won by a robot that can adapt. He defeated the famous humanoid Boston Dynamics ATLAS by simply adding a wheel that allowed him to ride.
Instead of building robots in the form of humans (although this is sometimes useful), you can allow them to evolve, evolve, find the perfect shape for the task. This will be especially useful in the event of a disaster, when expensive robots can replace humans, but must be prepared to adapt to unpredictable circumstances.
Many futurists envision the possibility of creating tiny nanobots that can create anything from raw materials. But this is optional. Programmable matter that can respond and respond to the environment will be useful in any industrial application. Imagine a pipe that can be strengthened or weakened as needed, or changed direction of flow on command. Or fabric, which may become more or less dense depending on conditions.
We are still far from the days when our beds can be transformed into bicycles. Perhaps the traditional non-technological solution, as is often the case, will be much more practical and economical. But as a person tries to shove a chip into every inedible object, inanimate objects will become a little more animate every year.
Ilya Khel