Creating life in the lab - Xenobots and Anthrobots | Michael Levin and Lex Fridman

Quick Overview

The discussion between Lex Fridman and Michael Levin centers on the creation and capabilities of living, programmable biological machines called Xenobots (from frog cells) and Anthrobots (from human tracheal cells), emphasizing that their emergent behaviors, like self-repair and collective motion, arise from cellular interactions rather than being explicitly programmed or inherited from their source genome.

Key Points: Xenobots are living robots made from frog embryo cells (Xenopus laevis) that can move, work together, and self-repair. Anthrobots are similar living robots constructed from human tracheal cells, capable of autonomous movement and performing tasks. The emergent properties observed in these systems, such as collective behavior, are not explicitly encoded in the cells' original genome but arise from novel ways the cells interact. The researchers employed evolutionary algorithms (AI) to design these novel forms, suggesting that natural selection on the cellular level can generate previously unseen biological systems. Anthrobots, made from adult human cells, demonstrated novel capacities like moving and clearing debris, with some exhibiting different gene expression profiles than their source cells. Levin suggests that these self-organizing systems, which can learn (like through Pavlovian conditioning), could eventually serve as tools to communicate with or repair complex biological structures like the brain.

Context: This segment features an interview between Lex Fridman and Michael Levin, a leading researcher in synthetic biology and regenerative medicine, focusing on the creation of novel living machines known as Xenobots and Anthrobots. These bio-bots are created by assembling stem cells into configurations that exhibit capabilities not found in the original organism's evolutionary history, representing a convergence of robotics and developmental biology.

Detailed Analysis

The conversation explores the nature of living machines created by assembling cells into novel configurations, specifically Xenobots (from frog embryos) and Anthrobots (from human tracheal cells). Levin emphasizes that these creations possess entirely new behaviors not present in their source organisms or evolutionary history. Xenobots, for instance, can move, work together, and self-repair. Anthrobots, made from adult human cells, show similar capabilities, including autonomous movement and environmental interaction, such as clearing debris. The process involves using AI/evolutionary algorithms to design the initial cell clusters, allowing the cells to self-organize into functional forms without direct genetic alteration. Levin clarifies that the resulting behaviors are emergent properties of cellular interaction within the new form, not inherited from the original frog or human cells. He notes that Anthrobots, derived from adult cells, display novel gene expression patterns compared to the donor cells. Furthermore, these systems exhibit forms of learning, such as Pavlovian conditioning, and can navigate 3D space using their innate cellular machinery (like cilia). Levin positions these bio-bots as potential tools for future medical applications, such as interfacing with or repairing complex tissues like the brain, by understanding the rules governing how cells communicate and organize.