Discover more from The Stoic CTO
#5: The Progression into Life 3.0
From bacteria to humans to self-modifying species. Insights from the book by Max Tegmark.
In a break from tech and work-related topics, let’s discuss something that induces curiosity.
To receive this newsletter in your inbox, consider subscribing 👇
I heard about Max Tegmark a few years ago after listening to his appearance on Lex Fridman podcast. Recently he has been more well-known as the president of the Future of Life Institute, which advocates for a 6 months pause on advanced AI developments.
I was impressed by his eloquent thoughts on life, AI, and the future of humanity. He wrote two books, one about AI and one about Math. I finished reading his AI book very fast, but I struggled to finish the Math one. Probably just because I didn’t have enough curiosity in Math yet.
In this post, I will share a summary and most important takeaways of the Life 3.0 book, however, I will focus more on the Life aspects, rather than the AI and Superintelligence aspects. Currently, there are a lot of discussions already regarding AI, AGI, and Superintelligence.
On the topic, I read other AI books as well, some that I have finished such as AI Superpower and Genius Makers. And some that are still in progress, such as Superintelligence. I usually read books in parallel, depending on where my interests take me.
In discussing life, we will start by defining life, structuring our thinking around it, and categorizing life by how advanced they are. The thinking of life in this book relates to the life that has been happening on Earth and what may happen here in the future. We are excluding the hypotheses of alien or extraterrestrial life here. However, I believe the core concepts would still be applicable to any form of life.
Whenever I read something mind-expanding, I usually take a lot of highlights and try to take notes so that I can refer back to them later. Hopefully, it can become one that I can share with my friends. I'm super interested in this topic and would love to discuss if you are also into it.
The book was written more than 6 years ago in 2017. There are a lot of advancements in biotech, medicine, and AI, so I’m feeling some of the content here is getting more real pretty fast.
What is Life?
Defining life is hard, and so is defining intelligence, memory, the mind, and numerous aspects of life. Competing definitions exist, and some include highly specific biological requirements such as being composed of cells, which may disqualify future intelligent machines and extraterrestrial aliens.
To have a generalized enough definition. life is simply a process that can retain its complexity and replicate. What is being replicated is not matter (atoms) but information (bits) specifying how the atoms are configured.
When bacteria replicates (makes a copy of its DNA), no new atoms are created, but a new set of atoms are arranged in the same pattern as the original, essentially copying the information.
We can also think of life as a self-replicating information-processing system whose information (software) determines both its behavior and the blueprints for its hardware.
In the book, Max breaks down Life into three versions with advancing complexity. It all boils down to their ability to modify their own software and hardware.
Life 1.0 cannot redesign its hardware or software during its lifetime: both are determined by its DNA and change only through evolution over many generations. Examples are common lifeforms such as bacteria, plants, fungi, and almost every non-human living being.
Many bacteria have a sensor for measuring sugar concentration in the liquids around them and can swim using a whip-like organ called flagella. The sensors and the flagella can work together to control the movements of the bacteria, optimizing for their survival and replication. They may implement a simple but useful algorithm like: "When the sugar concentration around me lowers than what was a few seconds ago, adjust my flagella and swim back".
Humans can learn to speak and have hundreds of other skills. Bacteria, however, aren't good learners. Their DNA specifies not only the design of their hardware such as their sugar sensors and flagella but also the design of their software. They never learn to swim toward sugar, instead, that procedure was hard-coded into their DNA from the start.
This is a form of learning, but not within the lifetime of an individual bacteria. Rather, it occurs during the preceding evolution of that species of bacteria. It went through a lot of trial and error across many generations where then natural selection happens, which results in mutations. Some mutations might improve the structure of the flagella or other hardware, while other mutations might improve the sugar-finding algorithms or other software.
So, bacteria is a form of Life 1.0 where both their hardware and software are evolved, rather than designed.
You and I, humans, are examples of Life 2.0: whose hardware is evolved, but the software is mainly designed.
If bacteria evolves only biologically, then humans not only evolve biologically but also culturally. Both through learning during a human's lifetime and passing down the learnings across generations. During our lifetime we can redesign much of our software: we can learn complex new skills—for example, languages, sports, and professions—and can fundamentally update our worldview and goals.
We weren’t able to perform any of those tasks when we were born, so all this software was programmed into our brains later through the process we call learning. Whereas our childhood curriculum is largely designed by our family and teachers, who decide what we should learn, we gradually gain more power to design our own software.
In schools and across various learning institutions we encounter after childhood, we were offered: Do we want to install a software module into our brain that enables us to work with Linear Algebra? Or do we want one that allows us to speak French? Or do we want to become a software engineer? Do we want to learn more about AI and the future of life by reading a book about it?
The fact that most of the human software is added after birth (through learning) is very useful since our intelligence is not limited by our DNA's information storage capacity. The synaptic connections in our brain have 100,000x more information than the DNA we are born with. Our synapses store all of our knowledge, skills, and experience in around 100 TB while our DNA only stores around 1GB, barely enough to store a movie.
This ability of Life 2.0 to design its software has numerous advantages compared to 1.0. It is smarter and more flexible. For example, bacteria could learn antibiotic resistance by evolving through many generations. However, when humans have an allergy to peanuts, we can immediately change our behavior by avoiding peanuts in the future. This gives Life 2.0 a dominating edge across the population level, even though our DNA barely changes in the past 50 thousand years.
By installing a software module enabling us to read and write, we are able to store and share vastly more information than we could ever memorize. By developing a software module capable of producing technology (i.e. by studying science and engineering), humanity could communicate with each other instantly and enable much of the world's information to be accessible in just a few clicks.
But despite the most powerful technologies we have today, all life forms we know remain fundamentally limited by their biological hardware. No one can live a million years, memorize the whole internet, and fly into space without equipment. To be fully free from the evolutionary handcuffs, life must undergo the final upgrade to Life 3.0 by having the ability to design not only its software but also its own hardware.
Humans are probably Life 2.1. We can redesign our software and some of our hardware (with medicine and various prosthetics). We can hack some of the defects and inefficiencies in our hardware, but very minimally.
A fully technological Life 3.0 doesn't exist yet on Earth. Theoretically, it should be able to dramatically redesign not only its software but its hardware as well, rather than having to wait for it to gradually evolve across generations.
After more than 13 billion years of cosmic evolution, significant developments are already happening here on Earth: Life 1.0 developed about 4 billion years ago, Life 2.0 (humans) arrived around 100 thousand years ago, and many AI researchers think that Life 3.0 may arrive in this century, perhaps even during our lifetime, accelerated by the progress of AI.
Looking at the scientific developments in the coming decade, I’m very curious about what the realistic next version of life on Earth is going to be. Will it be an upgraded us, a Life 3.0-style human with a synthetic and repairable body? Or will it be a technological agent that we invented?
This is another book that I'm currently reading. I've always been fascinated by emergent properties. How could something seemingly come out of nothing? How could some physical matter produce chemical reactions that produce a life that produces consciousness?
This piece introduced me to The Kardashev Scale, another way of classifying life (more specifically, intelligent civilizations) and their scale of advancement into three broad categories based on their energy consumption. For the most part, this piece elaborates in deep detail about The Fermi Paradox (why we haven't seen aliens).
Also by Tim Urban. This is probably the shortest introduction if you want to come down the rabbit hole of superintelligence. It was written over eight years ago, but some of the predictions aged eerily well today.
If you are interested to learn more about the future of life and AI, consider buying the book: Life 3.0, by Max Tegmark.
See you next time!
Thanks for reading The Stoic CTO! Subscribe for free to receive new posts and support my work.