海角大神

At first, they are just blurry blobs. But when Douglas Blackiston turns a knob on the microscope, they suddenly come to life in sharp relief.

They look like miniature, lumpy matzo balls suspended in water, but they move as if swimming with intention.

These are no ordinary lifeforms. Rather, Dr. Blackiston built them using frog stem cells and the guidance of artificial intelligence. They are the first 鈥渓iving robots.鈥�

At least that鈥檚 how Dr. Blackiston of Tufts University and his colleagues at Tufts and the University of Vermont (UVM) describe them. But these 鈥渪enobots鈥� 鈥� a nod to the Xenopus laevis, or African clawed frog 鈥� don鈥檛 fit neatly among nature鈥檚 creatures or traditional robots.

The team built these millimeter-wide biobots to move toward a target, form a group to push pellets to a central location, and self-repair, according to a paper published earlier this month. Future iterations could, conceivably, be used to clean microplastics from waterways or to deliver medications to precise locations in the body. But for now, these biological blobs are little more than a proof of concept.

Still, as prototypes for what might later be possible, the existence of xenobots raises ethical questions. The answers could have widespread implications. How we choose to define and categorize these creations could inform how we relate to them: Are they biological organisms? Technological machines? Or both?

鈥淭he terms that we use to describe them can be really important, because they make us think along certain lines,鈥� says Simon Coghlan, an ethicist at the University of Melbourne in Australia, and a veterinarian, who was not involved in the paper. 鈥淲e have all these impressions and ideas and associations that are linked with robots, which are different from the kind of associations and responses that we have to living organisms.鈥�

How to build a biobot

Perhaps the best way to start to figure out what to call a xenobot is to understand how it came to be.聽

Before being assembled in a petri dish, xenobots were formed on a supercomputer. UVM researchers tasked a machine learning program with testing out various body configurations. Dr. Blackiston at Tufts then recreated the virtual bots that exhibited desired behaviors in their lab.聽

He began by extracting stem cells from frog embryos. When clumped together, those cells naturally ball up. Then, using microsurgical tools, he would shape them into the AI-selected shapes.

Biobots are not made of metal or plastic like common machines. But if you define a machine as something that is designed to perform a specific task, says Michael Levin, director of the Allen Discovery Center at Tufts and a senior author on the paper, the xenobots measure up.

But, more importantly for ethical questions, are they alive?聽

鈥淲e think we know what a robot is. We think we know what a living thing is. But for all these things in biology, there are all these in-between cases,鈥� Dr. Levin says. 鈥淲hat this project really pushes the field to do is to think about better definitions for all these terms.鈥�

It鈥檚 easy to say whether most things we see are alive or not. Everyone would look at a fox or a tree, for example, and agree that it鈥檚 living. But determining a precise definition of what it means to be alive is a bit fuzzier. And if something is considered to be living, is it necessarily an organism?

鈥淚n this case, with the xenobots, the curious thing is that they are a bunch of frog cells stitched together,鈥� says Dr. Coghlan. 鈥淵ou could say those individual cells are living 鈥撀爐hey鈥檙e frog cells, we鈥檇 categorize them as life, in a sense 鈥撀燽ut they鈥檙e stitched together to form these new kinds of small, one-millimeter-sized bodies.鈥�

They behave in some ways biologists would use to describe an organism 鈥撀爐hey can move around and perform certain actions 鈥撀燽ut are lacking some other characteristics often cited. They also have not shown the ability to reproduce, or to find food in their environment or make energy to sustain themselves in any other way.聽

Something can be considered an organism and not receive much moral consideration from humans, however. The welfare of plants or bacteria, for example, doesn鈥檛 typically get much attention. But animals usually do.聽

The xenobot experiment was reviewed by an ethics committee, Dr. Levin says, as any research using frog embryos would be. 鈥淭hey ask questions like, what kind of experience is this creature likely to be capable of, and what are you doing to it, and is it worth the benefits to society for whatever that is?鈥�

What a life-form is likely to experience may be key to the moral question, says Dr. Coghlan. Although these xenobots are fairly simple and lack nervous systems, if future programs 鈥渃ould perhaps develop new kinds of organisms that are capable of feeling, or pain, or suffering, then I think that would raise more serious ethical questions.鈥�

Hubris and unintended consequences?

The human creation of a life-form also presents a broader concern for some that scientists are perhaps being too hubristic in trying to control biology.

Creating something new from something that already exists in biology isn鈥檛 a radical concept, Dr. Coghlan points out. Humans have been domesticating animals, creating hybrids (like a mule), and selectively breeding crops for centuries. But, he says, that concern about hubris might 鈥減oint to some deeper ethical or moral response to the ways in which we have this kind of power over nature.鈥�

Those concerns extend beyond the welfare of biobots themselves to worries about the unintended, cascading consequences of altering something in nature.

鈥淟iving organisms evolve and change, and oftentimes evolve and change in ways that we don鈥檛 expect, particularly once you actually release them into the environment and they interact with other living things,鈥� says Nita Farahany, who studies the ethical, legal, and social implications of emerging technologies at Duke University. 鈥淭he implication could be quite problematic.鈥�

Science fiction films have played out such scenarios to terrifying ends. Think about the often-quoted line from 鈥淛urassic Park鈥�: 鈥渓ife finds a way.鈥� What if the xenobots found a way to reproduce in the wild, or bred with other organisms? Or what if their behavior could alter in a nonlaboratory environment, to detrimental effects?

The researchers are in no rush to let the xenobots out of the lab. 鈥淟et鈥檚 be super clear, okay? We鈥檙e not about to dump these in the waterways,鈥� Dr. Levin says. 鈥淟et鈥檚 just not have any misunderstanding there.鈥�

Rather, he and Dr. Blackiston see their biobots as a 鈥渢estbed鈥� for ethical quandaries. They鈥檙e eager to use the xenobots to study cellular function in the laboratory as a way of better understanding what consequences occur when humans manipulate nature, and how best to harness such technologies.