Dennis Hackethal’s Blog
My blog about philosophy, coding, and anything else that interests me.
Can You Create Life in a Lab?
TL;DR: Yes, but there’s a catch.
Somebody asked about the theory of evolution and the origin of life. The replies got me thinking about the various challenges scientists face in explaining that origin. None of the replies mention one key philosophical problem scientists will have to address.
I commented:
Any successful theory on the origin of life would presumably need to be reproducible. Meaning it would tell you how to create conditions in a lab, say in a petri dish, out of which life arises reliably.
But if you create conditions under which life arises reliably, then YOU are the origin of that life, not the petri dish.
I have no solution to this problem. I challenge others to come up with one.
Someone responded incredulously. A moderator has since removed that comment, so I can’t quote it, but my reply still serves as an elaboration on my points above:
Evolution is about knowledge. If you carefully arrange things in a petri dish such that they start replicating, who/what is the origin of that knowledge? You.
But in a genuine origin-of-life event, it’s the petri dish coming up with that knowledge itself. Somehow. We don’t know how.
Genuine origin of life is the origin of knowledge. So any ‘recipe’ to make life in a petri dish isn’t really an origin of life. So headlines could not read ‘scientists create life in petri dish’ or even ‘scientists figure out origin of life’.
This problem is analogous to the one David Deutsch presents in the context of the simulation of evolution on computers (see his book The Beginning of Infinity, chapter 7). For such simulations to be genuine, the programmers need some way to determine whether these simulations come up with knowledge themselves or are just expressing, and iterating on, knowledge the programmers inadvertently leaked into them.
The same problem applies to origin-of-life experiments: if scientists carefully craft the conditions, and provide the ingredients, for life to emerge in a Petri dish, then those scientists are the origin of that life. Not the Petri dish. And thus whatever life those scientists may think they’ve ‘created’ is actually just a continuation of existing life.
Nevertheless, the genuine creation of life must be possible – or else it wouldn’t have happened on earth. But scientists will need to come up with a way to rule out the leakage of knowledge before even attempting any such experiments.
What people are saying
Why couldn't it just be a probabilistic explanation? Such that, given those conditions and enough time it's not unlikely for a replicator to form.
The theory of evolution itself is of a form: this is how it can happen.
Reply
Scientists like reproducibility. And what you describe is pretty much what we already have with things like RNA world except some details are missing.
Also, consider other origin-of-knowledge events such as creativity: every newborn has it, reliably & predictably (unless there are certain birth defects), yet it’s still genuine knowledge creation.
Reply
Yes, but in an attempt to discover the origin of life the point of making the conditions such that life arises "reliably" is simply to say that such conditions have a probability of arising that's on par with constraints like time, available materials at the time, etc. That completes the how it could've happened explanation. Then you test it independently. Don't see any philosophical problem there. The first bit of knowledge creation, by definition, had to have happened by chance.
Reply
I think scientists would shoot for probability 1 (or very close to 1).
That mode, again, is no better than RNA world. (Though that is good.)
Because you’re not addressing the leakage problem. That’s related but not the same.
Not necessarily. There could be spontaneous processes that reliably kick off knowledge-creating processes. That would also solve the leakage problem.
Reply
"Not necessarily. There could be spontaneous processes that reliably kick off knowledge-creating processes. That would also solve the leakage problem."
If they were spontaneous enough to create the conditions for themselves, than that's exactly the situation where what you call "the leakage problem" is present. If they're not spontaneous enough they'd need some sort of luck in the necessary conditions that they lack the ability to provide.
Reply
I don’t think so. The leakage problem refers to a situation where a knowledge-laden entity (eg, a person) puts knowledge in some other entity and observers then mistake the mere presence and exhibition of said knowledge for genuine knowledge creation on the part of that second entity. The second entity did not create knowledge, it merely inherited it, yet this inheritance goes unnoticed and is mistaken for creation.
A spontaneous process, by definition, is not knowledge-laden, thus there can be no leakage of knowledge there.
Reply
I think you're just bumping the problem up to fine-tuning.
Reply