Life 3.0: Being Human in the Age of Artificial Intelligence

In Chapter 1, Tegmark introduces the fundamental concepts of the book, the “beneficial AI” movement, common misconception and myths about AI, and an outline of the remainder of the book.

He first presents his spiritual vision for the future of highly intelligent life in the universe based on his love for and wonder at “cosmic awakening” (22). “Cosmic awakening” refers to the development of consciousness in an unconscious universe: “Thirteen point eight billion years after its birth, our Universe has awoken and become aware of itself” (22). This seems to be an unquestioned, faith-based belief for Tegmark. Consciousness is not an eternal part of reality, nor was reality forged by a conscious entity. Rather, consciousness is an aspect of reality that emerged at a particular place and time: Earth 13.8 billion years into existence. This fact is of deep spiritual and ethical resonance for Tegmark: “Had our universe never awoken, then, as far as I’m concerned, it would have been completely pointless—merely a gigantic waste of space” (22). Tegmark’s interests in the future, the advancement of AI, and the extension of human life, are all part of his missionary zeal to colonize a dead universe with intelligent, conscious lifeforms.

Tegmark lays out three stages of life: Life 1.0, Life 2.0, and Life 3.0. Humans are members of “Life 2.0,” which distinguishes us from other animals (Life 1.0) because we are capable of upgrading our “software,” i.e., culture, language, art, religion. However, like other lifeforms, we remain limited by our “biological hardware” (29). Life 3.0 is not. Such lifeforms would be completely freed from all biological limitations because they would be capable of changing their material/physical bodies. Figure 1.1 describes this as the technological stage of life’s development.

Tegmark discusses controversies and ethical issues surrounding the future of superintelligent AI. “Techno-skeptics” believe that superintelligent AI will not exist until the far future. Others, who think they might arrive within the next 100 years, are divided into three categories: Luddites (who believe AI will be terrible), Digital Utopians (who believe AI will be great), and the “Beneficial AI Movement,” a box in which Tegmark includes himself: “we felt that technology was giving life the power either to flourish like never before or to self-destruct, and we preferred the former” (34). Those in the Beneficial AI Movement believe in the necessity of AI safety research and the conscious direction of the future of AI engineering toward the good of human beings. Tegmark discusses the Future of Life Institute, his nonprofit dedicated, in part, to AI safety. He also references many intellectuals and businesspersons who are also interested in AI safety.

For Tegmark, the reason the AI question is the “most important conversation of our time” is because of its “urgency and impact” (37). It’s necessary that we proactively direct AI’s future because technology is changing rapidly and will impact both the short-term and long-term futures.

He provides a “cheat sheet” for the most important terms in the book. Some terms are technical while others are mundane but used here in specific ways. He also provides a cheat sheet of common myths and their divergences from the real issues in AI development. One of the most common myths is that AI will become evil. The real problem is that AIs will simply have goals that are different from our own, which could cause serious problems. Another myth is the proliferation of dangerous robots. Tegmark believes that the real worry is “misaligned intelligence” that “it needs no body, only an internet connection” (41). At this point in technological development, if something can take over the internet, it could potentially control the world.

At the end of Chapter 1, Tegmark includes a diagram describing how “speculative” the remaining chapters of the book are. While some are grounded in hard scientific facts politics, economics, etc., others are “extremely speculative” (47). This means that most of Life 3.0 is informed speculation on fascinating concepts and future possibilities by an expert in cosmology, physics, and artificial intelligence rather than a purely scientific text. It is designed to help readers think through possibilities and their ethical import and planning accordingly. For instance, Figure 1.6 shows a tree of questions about AI branching upward and outward to show the relationships between mundane questions about the present development of technology with more philosophical questions of the nature of reality and AI in the far future. The Beneficial AI movement, according to Tegmark, recognizes that “the goal should be to create not undirected intelligence, but beneficial intelligence” (36). It is only by thinking through scenarios, possibilities, and hypotheticals that we can determine our own desires for a shared future with AI.