Interesting research, but no we don’t have live and breeding robots


Enlarge / The crescent-shaped balls of cells would move in circles, stacking cells that could turn into mobile clusters.

Sam Kriegman and Douglas Blackiston

Scientists announced on Monday that they had optimized a way to obtain mobile clusters of cells to organize other cells into smaller clusters which, under the right conditions, could themselves be motile. The researchers call this process “kinematic self-replication”, although this is not entirely correct: the copies need the help of humans to start moving on their own, are smaller than originals and the copying process stops after only a few cycles.

So, of course, CNN headlined its coverage “The World’s First Living Robots Can Now Reproduce”.

This is a case where something really interesting is going on, but scientists and some of the development coverage are promoting it a lot more than it actually is. So let’s see what has really been done.

Movable balls of cells

The whole process begins with isolating the embryonic cells from a frog egg. Among the many interesting properties of cells, two are relevant for this work. The first is that these cells stick together. If you leave a group of them in a grow box, they will gather into a ball. You can even dissociate them entirely and leave a collection of them in culture, and they will stick together and form a ball of cells.

The second relevant characteristic of cells is that they self-organize, so that the outer cells have threadlike projections called cilia. These go around in circles, pushing against the surrounding fluid of the growing medium. Over time, neighboring cells coordinate their eyelashes in such a way that eventually the entire ball of cells will eventually spin their eyelashes in synchrony.

And that is enough for the whole ball of cells to start moving. For simplicity, we’ll call these balls of moving cells, or MBCs. Without obstacles, an MBC will make a circular path in the grow box and continue to do so for up to two weeks before running out of energy. These are the “living robots” from the CNN headline. They are not new either; the same research team described them in an article in March.

What is new is that the researchers placed a collection of dissociated cells in the dish with the MBCs. And, as the MBCs circled around, they created a cluster of cells in the center of the circle they drew (think of the path of the moving ball of cells like a donut – the cells clump together in the hole in the cell). donut). And, because these loose cells naturally adhere to themselves, a new group would begin to form out of them.

On their own, the new clusters were too small to do much. But the researchers could remove them from that box and place them in a new one with a dense supply of extra cells. The small clusters would grow by incorporating some of these extra cells and eventually reach the point where they too are motile.

Too mini me

The problem with this process is that it is inefficient. The second generation of MBCs was smaller than the first and drew a smaller circle. At the most, you could pass two generations of offspring MBCs before the whole process comes to a halt. They are therefore not particularly precise or efficient replicators.

The new work examined ways to optimize this by altering the shape of the original MBCs. But surgically modifying an MBC is hard work, and then it takes a few weeks in culture dishes prone to contamination to see how it behaves. Being inherently lazy, the researchers decided to model their behavior using computers, implementing an evolutionary algorithm that created variations in shapes that were then tested for their ability to bring cells together using of a physical simulator.

(Unfortunately, a researcher involved in the work of the Computer Science Department told CNN that this combination of algorithm and modeling is AI. If scientists want the public to better understand what they’re doing, that would help. they gave the public accurate information.)

Since the most efficient replication starts with stacking a large amount of cells, the algorithm essentially found the forms of MBC that did it most efficiently. The solution proposed by the model turned out to be a half-toroid.

To create them, the researchers had to perform surgery on the MBCs, cutting a notch on one side of the bullet after crushing it flat into a disc, then unfolding it. But the results were impressive: some of the crescent-shaped MBCs could produce offspring that could last three more generations before the process fails. But again, this wasn’t exactly self-replication, as subsequent generations wouldn’t take on a crescent shape without surgery.


Comments are closed.