It’s always hard to define life. Everyone has their favorite definition – some describe it as a struggle against entropy, while others describe it as an emergent property of chemicals. Countless books have been written on the topic, yet we’re far from a consensus.
Against the backdrop of the second law of thermodynamics, life seems like an improbable accident. When everything tends to go towards disorder, how come life is able to create cities, computers and space ships? How do we reconcile all the beautiful complexity we see around us with the stupidly simple laws we observe in physics?
Recently, I read this beautiful book, named Life as No One Knows It, by Sara Walker and had a minor revelation that maybe we’ve been looking at life from a completely wrong lens.

All this while, we have been fixated on different manifestations of life as we find it on Earth. We see bacteria, zebra fish, or fruit fly and start wondering whether we got lucky here on Earth, or whether our universe contains many more similar instances of life.
But, is this even the right question to ask? If we’re identifying life as we know it on Earth, we’re blinding ourselves to the only instance of life as we know it (with self-replication, cells, DNA, etc.). But, why should we expect life to be the same way in the rest of the universe?
Are cells fundamental? Does life have to be necessarily carbon-based? Sure, we can define life as anything. It’s an overloaded term. But if we keep defining it with traits we observe on Earth, we will be like that guy who only searches for his keys under the street light because that’s the only thing he can see.
Of course, to discover alien life, we need to first know what we’re looking for. But defining all life as being Earth-like is too myopic. What should we do?
Well, only when we define life as a precise physical phenomena that doesn’t refer back to life on Earth, can we expect to discover it at unexpected places in the universe. What could such a definition look like?
My take-home message from the book was this:
Life is a physical process that gives rise to systems that are statistically unlikely to assemble and persist (by themselves)
Sara, in her book, champions Assembly Theory which posits that each object in the universe has two numbers associated with it: assembly index and copy number. Assembly index is effectively the minimal number of steps required to (recursively) assemble an object from available building blocks. Think of it as analogous to number of steps required to assemble a lego shape. Copy number is how many copies of the object exist in the universe.
Once you accept these basic concepts, you can then define life as a process that generates objects with high assembly index and copy number. High assembly index assures that the process to produce it could not go about randomly. You just don’t smash a bunch of lego pieces and make a Taj Mahal. You need knowledge of it, and that knowledge could only come about via a history of selection.
That history is what life is.
So, if you see a complicated object – one that could not have assembled spontaneously from basic physics – you should assume a Darwinian-like selection process to be underway. High copy number of that complicated object simply means you get more statistical assurance that some creative process is underway that’s producing extremely complicated objects in multiple numbers.
Some concepts in the Assembly theory aren’t clear to me. For example, what is an object and what are building blocks? My current understanding is that Assembly theory works wonderfully at the chemical level where you can define objects as molecules, and building blocks as atoms. Then, assembly index can be precisely calculated via the shortest method of synthesising the molecule from atoms. But, outside that application, I’m not sure how will we calculate assembly index of objects like stars, or really complicated ones like kidneys.
Nevertheless, viewed from the lens of statistical impossibility, life suddenly elevates from the realm of Earth-biology to the study of patterns. We can start looking at life not as wet, mushy thing as we know it, but more like a creative force that builds artifacts that could not have come about randomly (because those complicated objects prevent their creative force from dying out.)
In fact, with this lens, life doesn’t even remain a binary thing. Instead of saying something is alive or dead, it’s probably better to talk in terms of degree of aliveness. Aliveness is what an object does to its surroundings. Does it create more complicated outputs v/s the inputs it consumes? Then it’s perpetuating life. The further the distance in complicatedness between an object’s outputs v/s inputs, the more alive something can said to be.
(I love that this definition aligns well with how we view agency in people. High agency people shape their environment, while low agency people get shaped!)
Overall, I strongly recommend reading the book. It’s a unique take on why we should go beyond life as we know it on Earth, and start looking at it as a physical phenomena that subsumes what we see here on Earth. To me, the shift is a bit like finally recognising Earth revolves around the Sun (and not the other way around).
The beautiful thing about rising one level is that it can help us see life at places we never expected it to be. Life is a creative force that refuses to die out by creating objects that can’t assemble spontaneously.
Perhaps AIs are alive in that sense? They’re a complicated object, present in multiple copies and are playing a vital role in creating more statistically unlikely objects in the universe. So, why disqualify them?
Join 200k followers
Follow @paraschopra