by Robyn Repko Waller
Xenobots are millimeter-sized life forms comprised of complexes of frog stem cells (from the species Xenopus laevis), whose shape is tailor engineered by AI according to their human-prescribed task. Capable of locomotion and transport, thanks to their rigid structure and contracting, xenobots have the potential to aid in localized medicine delivery and vital maintenance and clearing functions within a living system. Xenobots may even help gather and clear microplastics. All this from a biodegradable lifeforms, capable of sustaining itself without supplementary nutrients for weeks.
And now researchers Joshua Bongard and Michael Levin and a team of scientists at University of Vermont, Tufts, and Harvard have demonstrated that xenobots can self-replicate, or reproduce, with “parent” xenobots constructing “offspring” xenobots who then in turn produce third-generation offspring xenobots, etc. Evolutionary success? The key being their characteristic “Pac-man” form — a form determined by an evolutionary algorithm from billions of configurations to be optimally conducive to self-replication.
That xenobots are a medical and environmental breakthrough is obvious and scientifically exhilarating. Here I will set aside their groundbreaking applications and technological potential to ask a more philosophically motivated question.
Are xenobots robots? Are they living robots? Are they the first living robots? The first living robots to reproduce? That all depends, fundamentally, on what’s a robot. (And I guess by automated Turing standard tests of “Are you a robot?”, designed to weed out spambots, xenobots would fail. But so would my sweet hound dog Rudi.)
More seriously, we can get a hint of what the researchers mean to operationalize by ‘robot’ from the researchers themselves. One informative statement in the original press release “Team Builds the First Living Robot” from the University of Vermont picks out the xenobots as “programable organisms”:
“These are novel living machines,” says Joshua Bongard, a computer scientist and robotics expert at the University of Vermont who co-led the new research. “They’re neither a traditional robot nor a known species of animal. It’s a new class of artifact: a living, programmable organism.”
Biological machine covers a lot of ground, though, as I will argue, and hybrid biological-artificial systems are not new. Why classify xenobots as robots if they fall outside of our traditional usage of the concept? Here a statement by the co-lead Joshua Bongard, reported in CNN, is helpful:
“Most people think of robots as made of metals and ceramics but it’s not so much what a robot is made from but what it does, which is act on its own on behalf of people,” said Josh Bongard, a computer science professor and robotics expert at the University of Vermont and lead author of the study.
Robots, then, are just a class of entities or systems defined by their function. In this case, the stipulated characteristic function is to act — perhaps autonomously — on behalf of people for some defined task. The physical stuff doesn’t matter. Robot is a functional role category.
Putting this together, one can charitably propose that xenobots are the first living robots — with the scope of the ‘first’ ranging over ‘living robots’ — first in virtue of their biological composition (complexes of developed frog stem cells) and their autonomous task function prescribed by and for humans. Whereas previous robotics deal in artificial form, xenobots are biological life forms.
But, now, one might worry, how sparsely populated is the class of living robots? If xenobots count, surely other more mundane life forms get in. Consider humans themselves. Humans and indeed other animals and plants have long been proposed to be biological machines of sorts. Living systems that are describable in law-like terms, whose functions could be explicated in terms of causal mechanisms. The most contested but enticing application of which is the potential explanation of the human mind and brain, in its complexity and consciousness, as a biological machine itself. And, assuming the human brain — or any species’ biological brain — to be an evolved organ, it is programable to the tasks of homo sapiens (or that respective species).
But perhaps I’m getting off track. Perhaps what’s central is that robots must be created by humans for some purpose. So living robots must be created by humans. Entities with a derived intentionality or function. Plus a biological make-up. That is the heart of a ‘programable organism.’
Even still, we don’t have to go that complex to find members of this class. Bacteria and viruses are engineered by humans for medical research. Life-saving vaccines and infectious disease studies are a testament to this. These life forms, too, self-replicate, particularly in the case of bacteria. Beyond the microscopic, humans have selectively bred countless animal species for our purposes, my dog Rudi among those species, which in turn reproduce on their own. The tent of self-replicating living robot is a big one, it seems.
One wonders, then, at the end of this list if the robot aspect of xenobots doesn’t primarily trace back to its artificial intelligence origins, the algorithms that engineer the optimal form of the stem cell complexes for each task. The robots in this sense — the AI, then, aren’t living or self-replicating. The robots are artificial and engineering. This application of AI, the xenobots, is remarkable and laudable, nonetheless, promising great return for society, regardless of xenobots designation as living robots.