Hacker News Reader

Comment by EGreg | Hacker News Reader

There is no THE theory of abiogenesis.

The main question for me is abiogenesis.

How did the first self-replicating things evolve?

It seems that the kolmogorov complexity of the simplest quines in most languages is relatively high. One would imagine that, in terms of chemistry, the kolmogorov complexity of something that can self-replicate and also evolve with mutations which do not prevent the replication from taking place over generations - ie resilience - is too high to have arisen by mere chance.

View on HN

Be wary of anthropic bias. If this universe didn't happen to support life, then we wouldn't exist here to observe it. Perhaps there are other universes with different laws of physics, and we would exist in one of them instead. And even in our own universe, there are many trillions of planets. Yet we only know of one that has evolved life. The fact that we observe life here tells us nothing about how probable it is.

In fact as far as we can tell, life is extremely rare. We've only found one instance of it forming on it's own. And it can only exist in a very tiny space between the earths crust and the upper atmosphere (and only in very small quantities in most of that space.) We haven't found self replicating systems on other planets or deep within the earth.

So any theory that posits "life is an inevitable outcome of thermodynamics" is almost certainly wrong. It's a theory that can't be falsified, and has no predictive power. A more interesting theory would be one that can explain why our specific planet and our specific laws of physics support life. Why do exist here instead of somewhere else?

Yep it seems to be rare, thus falsifying the theory that England is developing.

And by the way, can't multiverse thinking be invoked pretty much anytime to explain anything?

"Oh we have fine tuning that is a knife's edge 1 part in 10 to the 127th"

"Yeah well maybe we have even more universes."

That's not real science. That's actually digging your heels in and saying "there is nothing beyond what we already found and understand". How will you go about testing this?

The multiverse isn't falsifiable. It's philosophy, not science.

>And by the way, can't multiverse thinking be invoked pretty much anytime to explain anything?

That's exactly the problem I've always had when it comes up.

Anthropic bias doesn't explain anything. It just says that you can't use your own existence as evidence of something. Because if you didn't exist, you wouldn't be able to make the opposite argument. It's basically just a wider application of survivorship bias. It works whether or not there are actually other universes (although the existence of other universes definitely makes it more intuitive to understand.)

By itself that may be true. As they say, "painting the target around the arrow" proves nothing.

But what if the target was there before? Then you have to explain how the arrow hit it so perfectly. There is a difference.

The difference is, is there anything special about the target's placement?

One can argue that alternative universes where soups of particles didn't give rise to life are all uninteresting.

You can of course say that life is needed to make the judgment for what is interesting.

But are you sure there is no sense in which the target can be said to have been painted before the arrow was fired? It all rests on that.

> In fact as far as we can tell, life is extremely rare.

As far as we can tell, right. Given our extremely narrowed view on the vast universe I'd say we've seen literally nothing yet. I'd not be surprised if life is not rare at all. It may only be rare in the tiny fraction of space that we know about.

When we first sent probes to other worlds, many people seriously expected to find life. The first astronauts on the moon were quarantined for weeks afterwards. Instead it seems like the vast majority of space is dead.

There may be lots of other planets like Earth, sure. But that's a very different claim than "life is an inevitable outcome of thermodynamics". At best it's "life is an inevitable outcome of very specific conditions that happened to exist on ancient Earth and Earth-like planets, under specific combinations of chemicals and temperature..." You need a lot more than just thermodynamics to have life.

It's probably entirely possible to create a self replicating machine using only components found on the surface of the moon. In fact that's been a serious research project of NASA for a long time. But such a machine will never form by chance, because the probability is so incredibly small. And so the moon is lifeless.

> and only in very small quantities in most of that space

It depends on how you define quantity. Total biomass may be low in proportion to crust, sure. But fully one fifth of the atmosphere is being maintained in a clearly unstable state (free oxygen!) by something. There's no way an alien would look at a spectrograph of our atmosphere and think anything but "it's alive".

As far as we can tell intelligent life is extremely rare. The not too distant universe could be well-populated with simple unicellular life (or just individual self-replicating molecules). It may be that self-replicating molecules form relatively easily but the jump to more complex life forms is exponentially more rare; but that would be more a question about the power of Darwinian selection than England's theory.

Also, England's theory never says "life is an inevitable outcome of thermodynamics," Jeremy England said that. There's a big difference. As it stands the theory says nothing about self-replication, it merely predicts chemicals forming seemingly-rare (with respect to the full chemical space) ordered structures. And his theory does have predictive potential and the simulated results in his PNAS paper support his predictions. It's certainly experimentally falsifiable: put a bunch of chemicals together that can form bonds, if you have an approximate understanding of the chemical space, you test apply an appropriate energy source, and see if the resulting chemicals occupy a surprisingly rare part of that space as predicted.

But I do agree that your question "Why do [we] exist here instead of somewhere else?" is one of the more interesting questions.

England's theory tells you that once you get close to a certain free energy basin you will access its optimal minimum. To my understanding it is silent about how you get "close enough"

Chance is not 'mere'. Whatever you think high complexity is, is just as well a low complexity to a different perspective. We are only slightly more remarkable than dirt, and over 4.5 billion years of chemical history, I don't see what you're basing your objection on.

Why are you certain that _self_ replication would be a necessary condition for "life-ish" chemistry to start happening, and to start evolving?

I expect chemical metabolisms pre-dated organisms, and purely metabolism based replicators pre-dated stored instruction replicators.

Can you elaborate on this and describe a complete picture of what you think happened?

Certainly not a complete picture. More of an unfinished amaetur impressionist water color.

But something along the lines of the dynamics seen in reaction diffusion systems.

Once you have a "metabolism" where interacting chemicals produce one another, you have a system that can sustain itself on it's reactants.

In a mess of random reactions, order will emerge through feedback effects, chemicals that are used up are used up, those that are replenished will increase in concentration.

Various regions will have different ratios, and different impurities giving a landscape in which versions of the metabolism may specialize or interact. ... Some regions produce molecules that self assemble into things like walls and membranes. ... RNA, likely around as a component the whole time, at some point it is specialized into an information medium role.

Stir continuously, simmer for 1-billion years or until it photosynthesizes.

Reply on news.ycombinator.com