|Mar. 18th, 2015 @ 12:28 pm Harry Potter and the Methods of Rationality|
Random thoughts on |
It's one of the only books supposedly about science that actually centrally presents the scientific method. Harry's natural inclination is very much clever-person-flaw of "try to figure it out by thinking hard" and several times it's emphasised that he should have just tried things more, and that he didn't question his assumptions enough. And that even when the evidence gathered is truly bizarre and contrary to scientific common sense (eg. why DO spells come in faux-latin or faux-Aramaic?) but incontrovertible, Harry and the plot both immediately assume the right thing to do is accept that and try to figure out why. This is the sort of thing it tries to do, and I think it deserves praise for succeeding!
Someone criticised it for not being scientific-method-y enough and I think that's also a valid criticism. A lot of the key insights come from "thinking very hard under pressure" not using the scientific method. That which CAN be a valid source of insight (mostly when you accept the guess you hadn't previously wanted to admit), but isn't usually. The article I can't remember the name of described it as Aristotelian, deriving insights by pure intellect. And I think that's partly that that happens sometimes, but intelligent people often wish that it was more common, and partly that for a plot to work, it often has to involve things being discovered suddenly.
Characterisation of Harry
The characterisation is somewhat inconsistent, trying to meld canon Harry Potter, Yudkowsky, and a theoretical perfect rationalist into one personality, but often feels genuine, and shows aweness of the strengths and weaknesses those people would display.
However, it does try to have it both ways a bit, in showing the mistakes Harry is prone to some of the time (mostly, failing to recognise that someone else might know something he didn't, or that people's feelings are important) but not as much as they maybe need. (Not necessarily saying it's unrealistic, obviously Harry as an 11-year-old wizard with bits of Voldemort's power imposed on him IS going to be weird, but not perfect in the story.)
Especially, it exhibits a flaw in Yudkowsky's writing when he uses something unthinkable as an example, and know's it's doing to be controversial, but apparently underestimates how MUCH blathering about [things you need trigger warnings for] might hurt people. This is one of the many things that, despite liking his writing, makes me also Very Cross about it.
It does much, much better than almost any book at constructing a complicated plot with multiple people scheming at cross purposes, and mostly has everything tied up neatly with all the time-travel, misunderstandings, secret agendas all legitimately justifying why people did various things that seem bizarre at the time.
Yudkowsky tries to do what I always want, of a plot where no-one holds the idiot ball. He writes about this in a series on his tumblr. And does it better than most other things!
But, inevitably, he writes a giant sprawling epic, posted episodically so he can't edit the earlier chapters to fit things that come up later, and doesn't achieve as much consistency as he aimed for.
Ooh, boy. Did I mention sprawling epic? It's surprisingly good considering -- I was certainly gripped all the way through. But it's nigh-impossible to digress in the middle for hundreds of chapters about rationality games and then try to return to the plot of Harry Potter book 1 at the end, even though that's what the medium sort of demanded.
Raising interesting ethical questions
It raised lots of interesting questions in my mind. Unsurprisingly, I didn't agree with all the answers suggested -- I've stopped even being surprised by that, I think that's just how interesting questions work.
Often there's something I'm sure is wrong, but I think I benefit from understanding why I think it's wrong. Like, should you do something for someone's own good? Or do something to society for society's own good? I think we implicitly do this all the time in small ways -- eg. say hello to someone if they say hello to you, respect it if they ask you not to, but don't assume you MUST say "is it ok to say hello to you" if they've greeted you first. But for making big decisions, it's often disastrous, because you think you know what's best, but most of the time, it depends on lots of things you don't know.
Edit: Fix spelling of 'Yudkowsky' twice. Sorry, EY!
You can also comment at http://jack.dreamwidth.org/939441.html using OpenID. comments so far.