Back in January, a team of scientists from the University of Vermont repurposed living Xenopus frog cells and assembled them into entirely new life-forms. These living machines – known as Xenobots – are neither a robot, nor a known species of animal, nor are they genetically modified. They are a new class of artefact: a living, programmable organism. But how did the Xenobots come about? The team used machine learning and an evolutionary algorithm to create thousands of candidate designs for the new life-form in a virtual petri dish. They told the supercomputer what they wish the eventual life-form to do, and the supercomputer tried billions of different combinations of virtual frog skin and heart cells to achieve the desired outcome. The produced virtual designs were then built using actual cells and, in many cases, the Xenobots moved in the way the computer predicted it would.
On the left, the anatomical blueprint for a computer-designed organism, discovered on a UVM supercomputer. On the right, the living organism, built entirely from frog skin (green) and heart muscle (red) cells.
Why create this ungodly creature in the first place? Two potential applications have been identified; environmental remediation tasks and drug delivery. Xenobots could be used to clean our polluted oceans by collecting microplastics. Similarly, bots designed with carefully shaped “pouches” might be able to carry drugs into human bodies. There is concern that these or similar bots could be used for malicious purposes. Another concern is that very large swarms of the bots could get out of control and do more damage than good.
In other news, researchers at the MIT have identified a powerful new antibiotic compound which in laboratory tests, has killed many of the world’s most problematic disease-causing bacteria, including some strains that are resistant to all known antibiotics! Our current methods for screening new antibiotics are often costly, require a significant time investment, and are usually limited to a narrow spectrum of chemical diversity. So, the team at MIT used machine learning to screen more than a hundred million chemical compounds in a matter of days. The algorithm was designed to explore, in silico, large chemical spaces and then, to pick out potential antibiotics that kill bacteria. The newly created compound – named Halicin after the fictional artificial intelligence system from “2001: A Space Odyssey,” – was then test by another machine learning model to see if it would likely have low toxicity to human cells. The researchers believe they can also train the model to add features that would make a particular antibiotic target only certain bacteria, preventing it from killing beneficial bacteria in a patient’s digestive tract.