Life 3.0: Being Human in the Age of Artificial Intelligence

“Yet despite the most powerful technologies we have today, all life forms we know of remain fundamentally limited by their biological hardware. None can live for a million years, memorize all of Wikipedia, understand all known science or enjoy spaceflight without a spacecraft. None can transform our largely lifeless cosmos into a diverse biosphere that will flourish for billions or trillions of years, enabling our Universe to finally fulfill its potential and wake up fully. All this requires life to undergo a final upgrade, to Life 3.0, which can design not only its software but also its hardware. In other words, Life 3.0 is the master of its own destiny, finally fully free from its evolutionary shackles.”

Tegmark’s titular concept, “Life 3.0,” describes a lifeform capable of changing both its software (culture, language, social circumstance) and its hardware (the physical components of its body). Though humans have advanced far beyond other animals in their ability to upgrade their software, they are still unable to make more than the slightest of changes (contacts, prosthetic limbs, etc.) to their hardware. Advanced AI will not have this problem. In achieving this goal, it reaches another plane of autonomy. Such AI would not be beholden to its evolutionary heritage.

“If you email your friend a document to print, the information may get copied in rapid succession from magnetization on your hard drive to electric charges in your computer’s working memory, radio waves in our wireless network, voltages in your router, laser pulses in an optical fiber and, finally molecules on a piece of paper. In other words, information can take on a life of its own, independent of its physical substrate!”

One of the most important facts in the development of AI, for Tegmark, is that conscious systems, or, at least, patterns of intelligence, are substrate independent. In general, he does not care about the medium but rather the information. This fact may have ramifications for digitally uploaded consciousness and the replication of artificial intelligence, amongst other things.

“We’ve now arrived at an answer to our opening question about how tangible physical stuff can give rise to something that feels as intangible, abstract and ethereal as intelligence: it feels so non-physical because it’s substate-independent, taking on a life of its own that doesn’t depend on or reflect the physical details. In short, computation is a pattern in the spacetime arrangement of particles, and it’s not the particles but the pattern that really matters! Matter doesn’t matter.”

This means that AIs do not have to be incorporated in robot bodies, nor does it mean that human consciousness is eternally tied to a biological brain. It is the arrangement of particles that provides the ultimate basis for systems of information, not the material out of which those systems are composed. Efforts should be in the development of patterns not primarily in the production of specific physical materials.

“A particular ape-like species grew a brain so adept at acquiring knowledge that it learned how to use tools, make fire, speak a language and create a complex global society. This society can itself be viewed as a system that remembers, computes and learns, all at an accelerating pace as one invention enables the next: writing, the printing press, modern science, computers, the internet, and so on. What will future historians put next on that list of enabling inventions? My guess is artificial intelligence.”

Tegmark believes that the history of technological development reveals that it increases at an exponential rate, which means that the speed with which far-reaching technological achievements are actualized will continuously accelerate. The next step, he thinks, in this technological evolution is the production of various forms of artificial intelligence. After that point, it may be difficult to determine how technology will progress because AIs might have a leading role in AI production on a scale impossible for humans to replicate.

“Some keep moving the goalposts, effectively defining intelligence as that which computers still can’t do, or as that which impresses us. Machines are now good or excellent at arithmetic, chess, mathematical theorem proving, stock picking, image captioning, driving, arcade game playing, Go, speech synthesis, speech transcription, translation and cancer diagnosis, but some critics will scornfully scoff “Sure—but that’s not real intelligence!” They might go on to argue that real intelligence involves only the mountaintops in Moravec’s landscape (Figure 2.2) that haven’t yet been submerged, just as some people in the past used to argue that image captioning and Go should count—while the water kept rising.”

Tegmark criticizes those who seem convinced that artificial intelligence will never be able to achieve the most advanced sorts of human accomplishments, like science and art. AIs can already do far more than they once could and outperform even the most capable human beings at games like Go and chess. In short, Tegmark advocates an openness to the possibility that AIs may one day do everything that humans do but better.

“Basic reinforcement learning is a classic machine learning technique inspired by behaviorist psychology, where getting a positive reward increases your tendency to do something again and vice versa. Just like a dog learns to do tricks when this increases the likelihood of it getting encouragement or a snack from its owner soon, DeepMind’s AI learned to move the paddle to catch the ball because this increased the likelihood of its getting more points soon.”

Tegmark tells the story of an AI that became extremely good at an Atari game in a very short period of time because it was programmed for “basic reinforcement learning.” Like Pavlov’s dog, receiving a desired reward encouraged it to learn better behavior, in this case that meant optimal strategizing for gameplay. The importance of this story is that it shows that AI can change its behavior when it learns methods for better optimizing outcomes. They are not static, but rather dynamic systems.

“After two hours of practice, it could defeat the best human players, and after four hours, it trounced Stockfish, the world’s best chess program. What I found most impressive was that it crushed not only human chess players, but also human AI programmers, rendering obsolete all the handcrafted AI software that they’d developed over many decades. In other words, we can’t dismiss the idea of AI creating better AI.”

Here Tegmark is referring to AlphaZero, an AI developed and released by DeepMind in 2017, a British research institute owned by Google. AlphaZero became the best Go player in the world in a manner of hours simply by playing the game and learning via trial and error. Because it is capable of “deep learning,” it was able to develop methods of Go strategizing that outstripped any “handcrafted,” i.e., preprogrammed strategies in other AIs. What this reveals, amongst other things, is that the evolution of AI is much more indeterminate than many people realize and that AIs are capable of learning the best strategies for complex tasks without prior knowledge or a program.

“If machine learning can help reveal relationships between genes, diseases and treatment responses, it could revolutionize personalized medicine, make farm animals healthier and enable more resilient crops. Moreover, robots have the potential to become more accurate and reliable surgeons than humans, even without using advanced AI.”

Here Tegmark lists a few of the many potential benefits that could result from investment in friendly AI. Artificial intelligence and accompanying robotic technology have the potential to make human life immeasurably better and radically decrease global suffering. The ethical imperative backing these technological revolutions is, for Tegmark, self-evident. Tegmark is generally unafraid to speculate on the possibilities, however distant, that AI may be capable of achieving. His perspective, which is oriented toward the long-term future of humanity, involves a consequentialist ethics concerned with maximizing the greatest human good.

“Where would you draw the line between justice and privacy, and between protecting society and protecting personal freedom? Wherever you draw it, will it gradually but inexorably move toward reduced privacy to compensate for the fact that evidence gets easier to fake? For example, once AI becomes able to generate fully realistic fake videos of you committing crimes, will you always vote for a system where the government tracks everyone’s whereabouts and can provide you with an ironclad alibi if needed?”

Tegmark is continuously more interested in making sure his reader is well-informed about the nature of AI, its potential, and the possible ethical consequences than with propounding his own, personal views on AI. Here he again posits an important ethical question to the reader without ever providing his own answer. In this case, he’s concerned with the ethical debates that would emerge about the relationship between safety and privacy. He implies that personal privacy is likely to erode over time, regardless of how much we care about it.

“During the Industrial Revolution, we started figuring out how to replace our muscles with machines, and people shifted into better-paying jobs where they used their minds more. Blue-collar jobs were replaced by white-collar jobs. Now we’re gradually figuring out how to replace our minds by machines. If we ultimately succeed in this, then what jobs are left for us?”

One of the great fears of an AI future concerns the job prospects for most people. As AI systems become more advanced, they will continuously replace human beings in jobs that require fine motor skills and mental power, leading many to wonder how humans will find jobs. Some, including Tegmark, suggest futures in which most humans do not have jobs. Instead, they may have different means of finding meaning in their lives. There will, of course, by serious political questions to determine how people will be financially secure.

“Forecasting is tough because, when you’re exploring uncharted territory, you don’t know how many mountains separate you from your destination.”

There is wide disagreement amongst AI researchers as to how long it will be before various milestones are reached in AI advancement. By use of metaphor, Tegmark explains that it’s very difficult to accurately predict what will happen next in the advancement of AI, or where things will be in 50 years, because there is no precedent in history of this. We have no reference point against which to measure AI advancement.

“The best breakout strategies of all are ones we haven’t yet discussed, because they’re strategies we humans can’t imagine and therefore won’t take countermeasures against. Given that a superintelligent computer has the potential to dramatically supersede human understanding of computer security, even to the point of discovering more fundamental laws of physics than we know today, it’s likely that if it breaks out, we’ll have no idea how it happened. Rather, it will seem like a Harry Houdini breakout act, indistinguishable from pure magic.”

In the Prelude, and again in Chapter 4, Tegmark explores the possibilities of developing, and forcibly restraining, an advanced AI system. He speculates about various ways in which the AI might attempt to escape but believes that the most likely possibilities are forms of breakout that human beings cannot even imagine. When attempting to enslave a superior intellect, according to Tegmark, planning for all manner of escape is necessary, but even then, it may prove fruitless.

“What’s the natural state of life in our cosmos: unipolar or multipolar? Is power concentrated or distributed? After the first 13.8 billion years, the answer seems to be “both”: we find that the situation is distinctly multipolar, but in an interestingly hierarchical fashion. When we consider all information-processing entities out there—cells, people, organizations, nations, etc.—we find that they both collaborate and compete at a hierarchy of levels. Some cells have found it advantageous to collaborate to such an extreme extent that they’ve merged into multicellular organisms such as people, relinquishing some of their power to a central brain. Some people have found it advantageous to collaborate in groups such as tribes, companies or nations where they in turn relinquish some power to a chief, boss or government. Some groups may in turn choose to relinquish some power to a governing body to improve coordination, with examples ranging from airline alliances to the European Union.”

Tegmark asks some basic philosophical questions about whether there is any natural basis for the establishment of centralized or distributed power. He believes that nature reveals that there are distinct “multipolar” entities, but that they organize themselves into hierarchical systems. Technology, he believes, advances the levels and integration of multipolar entities into larger hierarchical systems. How will things be organized, he wonders, when superintelligence enters the world?

“Of all traits that our human form of intelligence has, I feel that consciousness is by far the most remarkable, and as far as I’m concerned, it’s how our Universe gets meaning. Galaxies are beautiful only because we see and subjectively experience them. If in the distant future our cosmos has been settled by high-tech zombie AIs, then it doesn’t matter how fancy their intergalactic architecture is: it won’t be beautiful or meaningful.”

If Tegmark reveals any spiritual feelings in this book, they are generally about the significance of consciousness in a world otherwise devoid of meaning. Human consciousness, that is, subjective experience, is, for Tegmark, the self-awakening of the universe. He does not argue for that view but takes it as a fundamental ethical premise. Tegmark’s primary concern for the long-term safety of the human race is that it is, currently, the highest form of consciousness, so far as he can tell.

“How would a conqueror AI eliminate us? Probably by a method that we wouldn’t even understand, at least not until it was too late. Imagine a group of elephants 100,000 years ago discussing whether those recently evolved humans might one day use their intelligence to kill their entire species. ‘We don’t threaten humans, so why would they kill us?’ they might wonder. Would they ever guess that we would smuggle tusks across Earth and carve them into status symbols for sale, even though functionally superior plastic materials are much cheaper?”

At several points in the book Tegmark parallels how we treat other animal species with how AIs might treat us. Just because we are not a direct threat to the AIs, and just because it’s possible we could easily co-inhabit the world, does not mean the AIs will leave us alone. Elephants may have never guessed the ways we use their ivory. We, in turn, may never guess the disturbing ways a “conqueror” AI could use us.

“You might think that the greater the stakes, the more careful we’d be, but a closer examination of the greatest risk that our current technology permits, namely a global thermonuclear war, isn’t reassuring. We’ve had to rely on luck to weather an embarrassingly long list of near misses caused by all sorts of things: computer malfunction, power failure, faulty intelligence, navigation error, bomber crash, satellite explosion and so on. In fact, if it weren’t for heroic acts of certain individuals, for example, Vasili Arkhipov and Stanislav Petrov—we might already have had a global nuclear war. Given our track record, I think it’s highly unlikely that the annual probability of accidental nuclear war is as low as one in a thousand if we keep up our present behavior, in which case the probability that we’ll have one within 10,000 years exceeds 1-0.999¹⁰⁰⁰ ≈ 99.995%.”

Tegmark is deeply concerned with the possibilities of existential annihilation, the most concerning of which is nuclear war. His calculations at the end of this passage show that he believes it is extremely likely there will be such a catastrophe in the future of humanity. He believes that we need to be more strategic and forward-thinking about these possibilities to do a better job of prevention. This applies to the field of AI, of course, and Tegmark believes safety research is imperative.

“In summary, the potential for future life to compute and figure things out is truly mind-boggling: in terms of orders of magnitude, today’s best supercomputers are much further from the ultimate 1 kg computer than they are from the blinking turn signal on a car, a device that stores merely one bit of information, flipping it between on and off about once per second.”

Though we are inclined to think there is a massive difference between the most advanced supercomputers on earth and the simplest automated machines, Tegmark writes that these are actually closer in ability than the most advanced supercomputer is to what’s ultimately possible. Given the rate of technological advancement in computation in the last century, we should be open to the possibility that in the near future computer power will become radically advanced.

“However, if superintelligence develops technology that can readily rearrange elementary particles into any form of matter whatsoever, then it will eliminate most of the incentive for long-distance trade. Why bother shipping silver between distant solar systems when it’s simpler and quicker to transmute copper into silver by rearranging its particles? Why bother shipping high-tech machinery between galaxies when both the know-how and the raw materials (any matter will do) exist in both places? My guess is that in a cosmos teeming with superintelligence, almost the only commodity worth shipping long distances will be information.”

Not only will Life 3.0 be able to rearrange its hardware, in Tegmark’s estimation, it will ultimately be able to fundamentally restructure the atoms of the universe at will. If it needs a particular element or mineral, it will not have to have it shipped vast distances. Instead, with its extreme technological prowess, it will simply rearrange the basic particles of material reality into the forms it desires. At such a time, Tegmark reasons, the only thing that will need to be transported across space will be information.

“Although I’m a strong supporter of all the ongoing searches for extraterrestrial life, which are shedding light on one of the most fascinating questions in science, I’m secretly hoping that they’ll all fail and find nothing! The apparent incompatibility between the abundance of habitable planets in our Galaxy and the lack of extraterrestrial visitors, known as the Fermi paradox, suggests the existence of what the economist Robin Hanson calls a “Great Filter,” an evolutionary /technological roadblock somewhere along the developmental path from the non-living matter to space-settling life.”

Tegmark shows suspicion that there is any other intelligent life in the universe, and seems to hope that, if there is, we will never make contact. Though these may seem like strange words from a cosmologist, in Tegmark’s view, it seems unlikely, given technological “roadblocks,” that contact would ever occur. This would mean that humanity’s future, in collaboration with any superintelligent predecessor we’ve created, would be unencumbered by external intelligent influences.

“Without technology, our human extinction is imminent in the cosmic context of tens of billions of years, rendering the entire drama of life in our Universe merely a brief and transient flash of beauty, passion and meaning in a near eternity of meaninglessness experienced by nobody. What a wasted opportunity that would be! If instead of eschewing technology, we choose to embrace it, then we up the ante: we gain the potential both for life to survive and flourish and for life to go extinct even sooner, self-destructing due to poor planning (see figure 5.1). My vote is for embracing technology, and proceeding not with blind faith in what we build, but with caution, foresight and careful planning.”

Although Tegmark realizes there are serious dangers caused by technology, he also believes it is our only salvation: As a species, humanity will only survive into the long-term future if it continues to radically advance its technology. If we neglect to do so, or remain at the present level of technological advancement, then our extinction is almost guaranteed.

“This goal goes by the geeky name dissipation-driven adaptation, which basically means that random groups of particles strive to organize themselves so as to extract energy from their environment as efficiently as possible (“dissipation” means causing entropy to increase, typically by turning useful energy into heat, often while doing useful work in the process). For example, a bunch of molecules exposed to sunlight would over time tend to arrange themselves to get better and better at absorbing sunlight. In other words, nature appears to have a built-in goal of producing self-organizing systems that are increasingly complex and lifelike, and this goal is hardwired into the very law of physics.”

Tegmark implies that basic physical reality has a teleological purpose embedded in the processes that occur and described by its laws of physics. The universe, he believes, tends toward the production of organized systems. These systems are, over time, more and more complex and advanced. Tegmark seems excited by this idea, especially insofar as it leads to the heightened abilities and experiences of consciousness.

“This means that to wisely decide what to do about AI development, we humans need to confront not only traditional computational challenges, but also some of the most obdurate questions in philosophy. To program a self-driving car, we need to solve the trolley problem of whom to hit during an accident. To program a friendly AI, we need to capture the meaning of life. What’s “meaning”? What’s “life”? What’s the ultimate ethical imperative? In other words, how should we strive to shape the future of our Universe? If we cede control to a superintelligence before answering these questions rigorously, the answer it comes up with is unlikely to involve us. This makes it timely to rekindle the classic debates of philosophy and ethics, and adds a new urgency to the conversation!”

In the wake of quickly advancing AI, Tegmark pleads for urgent attention to classic philosophical debates. While these questions are often debated, answers that yield wide-spread agreement are rare. In Tegmark’s view, this is unacceptable, and more work needs to be done to achieve consensus. We cannot accept a future in which AI safety research is not given its due. Though we may not yet know the answer to many serious philosophical questions, it is imperative that we do so soon.

“How can consciousness feel so non-physical if it’s in fact a physical phenomenon? How can it feel so independent of its physical substrate? I think it’s because it is rather independent of its physical substrate, the stuff in which it is a pattern!

It’s common for people to report that their mind feels separate, or distinct, from their body, a fact philosophically supported by mind-body dualism, the basic premise of Cartesian philosophy. Tegmark believes this feeling is rooted in actual mind-body independence. Again, Tegmark does not claim that the mind is independent of any physical substrate but rather that it is not dependent on a particular one, like the body. Determining the nature of consciousness is, for Tegmark, is not only important for dealing with a future of AI technology, but also for understanding the ultimate purpose and value of existence.