Log in

No account? Create an account
Musings On Machine Learning and My Lovecraftian Worldview - Synchronicity swirls and other foolishness

> Recent Entries
> Archive
> Friends
> Profile
> my rpg writing site

April 21st, 2017

Previous Entry Share Next Entry
02:23 am - Musings On Machine Learning and My Lovecraftian Worldview
Here's a fascinating, and to me at least, extremely wonderful and hopeful article: Our Machines Now Have Knowledge We’ll Never Understand It's about machine-learning, and the fact that computers win at go by playing in ways that humans don't, and that this is true of pretty much all modern machine learning and AI problem-solving, they don't do things like humans would, and more importantly, they don't create relatively simple laws and formulas from which they can make predictions. Humans do that, but machines don't. Here's one of the key insights:
Our machines now are letting us see that even if the rules the universe plays by are not all that much more complicated than Go’s, the interplay of everything all at once makes the place more contingent than Aristotle, Newton, Einstein, or even some Chaos theorists thought. It only looked orderly because our instruments were gross, because our conception of knowledge imposes order by simplifying matters until we find it, and because our needs were satisfied with approximations.

That’s fine if you just want to put the 8-ball in the corner pocket. But if you want to know the real path that ball will take, you have to look at the friction created at the molecular level as it passes over each fiber of the felt, at the pull of the moon and the moment’s variation in the Earth’s wobble, at the unequal impact of the photons emitted from the light fixture above the table and the lamp off to the side, and at the change in the air current as your opponent holds her breath. Not to mention the indeterminacy of the quanta. None of that may affect whether you sink the ball, but it is the truth of what’s going on. Even if the universe is governed by rules simple enough for us to understand them, the simplest of events in that universe is not understandable except through gross acts of simplification.
In short, there's every reason to assume that the universe is too complex for us to understand, but we can use machine learning to discover insights, and techniques for working with it that would be impossible for humans to ever find except by chance, and perhaps not even then.

I can see some people finding this idea disturbing or depressing, but I find it exceedingly exciting. I'd honestly be sad if human brains were capable of encompassing the truths of the universe – that would imply a pretty simple, and from my PoV, rather dull, and likely deterministic universe. Instead, we may have a wildly complex. As J.B.S. Haldane wrote The universe is not only stranger than we imagine, it is stranger than we can imagine. In part, this seems obvious, because as I discuss here, it seems likely that all lifeforms have conceptual limits, and there's no reason to suppose we do not.

I find this exciting both because it implies wonders and complexities that will challenge even the post-human beings I hope and rather expect us to become, and also because I describe my worldview as fundamentally Lovecraftian (but definitely not in any sort of nihilistic manner – I find the idea of a carefully ordered universe with an inherent purpose that we did not choose to be vastly more nihilistic than one full of wonder, near infinite possibilities, and just as many choices. I also rather love to idea of gaining insights into the world, which can both provide information and useful technologies, which are based in working with data in ways that humans both don't and can't fully understand. With luck, as we change ourselves to become smarter and capable of deeper insights, the world would continue to be complex beyond our understanding, but also susceptible to ever more sophisticated analysis.

Original post on Dreamwidth - there are comment count unavailable comments there.

(Leave a comment)

> Go to Top