It’s tough to enter the conversation around AI these days. It’s all to easy to get distracted by people pushing competing predictive techniques; it’s far too difficult to evaluate the claims to “AI” that litter the marketing landscape.
So at a time when critical decisions about our automated future are being made, how are we supposed to understand what’s going on, let alone make proper risk v reward calculations?
Max Tegmark comes to the rescue with Life 3.0.
Grounded in real world case studies, an enthralling hypothetical narrative of the future, and easily understandable (and usable!) definitions of key terms, Life 3.0 serves as your rocket ship to the front of the discussion surrounding AI.
Central to the book is Tegmark’s definition of life itself and its 3 stages:
- Life is anything that can retain its information and replicate, e.g. DNA.
- Life 1.0, which is life that slowly evolves both its hardware (physical “body”) and software (mental processes). This begins with individual atoms and continues through to most complex animals.
- Life 2.0, in which bodies evolve but the ability to create software is unlocked (e.g. writing the mental “program” “how to read English”).
- Life 3.0, in which not only software but also hardware is subject to design over evolution. Imagine what could happen when a machine with the ability to simulate in a matter of hours millions of potential designs for an improved method for bipedal motion gets control of a robotics factory!
So Life 3.0 will be fully upon us when it gains control of its physical manifestation. Cool, let’s move on, because there’s a lot more to cover.
How will we relate to machines that can potentially think millions of times faster or more deeply than we can? (there’s an interesting discussion of the tradeoffs between “fast” and “deep” thought, by the way)
After defining what life is and is not at each of these three stages, Tegmark helps us time travel at three scales: the near future, the next 10,000 years, and the next billion.
Billion years?! you might ask, astonished at the audacity. On this journey, the imminence of the next billion years is almost palpable. We sit at a tremendous precipice in our universe’s history: at the summit of 13.8 billion years of evolution we have unlocked such power that we have the tools to unwind a good portion of life’s progress in a blink of the cosmos’ eye.
And really, that’s why this book is so important. While “AI” serves as a solid framework for thinking about life on the grander scale, there is a clear line between the actions we take today and the array of possible futures that we face. He gives us a picture of what they might each look like, lays out the bounds of discussion, and then invites us to be active participants in our own future.
Coming up, I’ll start an in depth conversation about key choices we have to make now – particularly the affinity and control problems. In the meantime, do yourself a favor and read this book!