= 27d567bc2d48d9f40e9ccc20ea370b15)
Alsadius comments on Open thread, Dec. 1 - Dec. 7, 2014 - Less Wrong Discussion
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (346)
Which is better - weak evidence, or none?
An interesting question. Let me offer a different angle.
You don't have weak evidence. You have data. The difference is that "evidence" implies a particular hypothesis that the data is evidence for or against.
One problem with being in love with Bayes is that the very important step of generating hypotheses is underappreciated. Notably, if you don't have the right hypothesis in the set of hypotheses that you are considering, all the data and/or evidence in the world is not going to help you.
To give a medical example, if you are trying to figure out what causes ulcers and you are looking at whether evidence points at diet, stress, or genetic predisposition, well, you are likely to find lots of weak evidence (and people actually did). Unfortunately, ulcers turned out to be an bacterial disease and all that evidence, actually, meant nothing.
Another problem with weak evidence is that "weak" can be defined as evidence that doesn't move you away from your prior. And if you don't move away from your prior, well, nothing much changed, has it?
"Weak" means that it doesn't change your beliefs very much - if the prior probability is 50%, and the posterior probability is 51%, calling it weak evidence seems pretty natural. But it still helps improve your estimates.
Only if it's actually good evidence and you interpret it correctly. Another plausible interpretation of "weak" is "uncertain".
Consider a situation where you unknowingly decided to treat some noise as evidence. It's weak and it only changed your 50% prior to a 51% posterior, but it did not improve your estimate.
Often none.
For example, if a piece of evidence E is such that:
- I ought to, in response to it, update my confidence in some belief B by some amount A, but
- I in fact update my confidence in B by A2,
and updating by A2 gets me further from justified confidence than I started out, then to the extent that I value justified confidence in propositions I was better off without E.
Incidentally, this is also what I understood RowanE to be referring to as well.
But it's only bad because you made the mistake of updating by A2. I often notice a different problem of people to always argue A=0 and then present alternative belief C with no evidence. On some issues, we can't get a great A, but if the best evidence available points to B we should still assume it's B.
Agreed.
Agreed.
Yes, I notice that too, and I agree both that it's a problem, and that it's a different problem.
Overconfidence is a huge problem. Knowing that you don't understand how the world works is important. To the extend that people believe that they can learn significant things from history, "weak evidence" can often produce problems.
If you look at the Western Ukraine policy they didn't make a treaty to accept Russian annexion of the Krim in return for stability in the rest of Ukraine. That might have prevented the mess we have at the moment.
In general political decisions in cases like this should be made by doing scenario planning.
It on thing to say that Britian and France should have declared war on Germany earlier. It quite another thing to argue that the West should take military action against Russia.
Might have, but my money isn't on it. You think Putin cares about treaties? He's a raw-power sort of guy.
And yes, the scenarios are not identical - if nothing else, Russia has many more ICBMs than Hitler did. Still, there's ways to take action that are likely to de-escalate the situation - security guarantees, repositioning military assets, joint exercises, and other ways of drawing a clear line in the sand. We can't kick him out, but we can tell him where the limits are.
(Agreed on your broader point, though - we should ensure we don't draw too many conclusions).
Putin does care about the fact that Ukraine might join NATO or the EU free trade zone. He probably did feel threatened by what he perceived as a color revolution with a resulting pro-Western Ukrainian government.
At the end of the day Putin doesn't want the crisis to drag on indefinitely so sooner or later it's in Russia's interest to have a settlement. Russia relies on selling it's gas to Europe.
Having the Krim under embargo is quite bad for Russia. It means that it's costly to keep up the economy of the Krim in a way that it's population doesn't think the Krim decayed under Russian rule and there unrest.
On the other hand it's not quite clear the US foreign policy has a problem with dragging out the crisis. It keeps NATO together even through Europeans are annoyed of getting spied at by the US. It makes it defensibly to have foreign miltary bases inside Germany that spy on Germans.
Do you really think joint exercises contribute to deescalation?
As far as repositioning military assets goes, placing NATO assets inside Ukraine is the opposite of deescalation.
The only real way to descalate is a diplomatic solution and there probably isn't one without affirming Crimea as part of Russia.
There's a certain type of leader, over-represented among strongmen, that will push as far as they think they can and stop when they can't any more. They don't care about diplomacy or treaties, they care about what they can get away with. I think Putin is one of those - weak in most meaningful ways, but strong in will and very willing to exploit our weakness in same. The way to stop someone like that is with strength. Russia simply can't throw down, so if we tell them that they'd have to do so to get anywhere, they'd back off.
Of course, we need to be sure we don't push too far - they can still destroy the world, after all - but Putin is sane, and doesn't have any desire to do anything nearly so dramatic.
Putting gains inner politcs strength from the conflict.
That assumes that you can simply change from being weak to being strong. In poker you can do this as bluffing. In Chess you can't. You actually have to calculate your moves.
Holding joint military exercises isn't strength if you aren't willing to use the military to fight.
Bailing out European countries is expensive enough. There not really the money to additionally prop up Ukraine.
Only as long as he's winning.
NATO is, far and away, the strongest military alliance that has ever existed. They have the ability to be strong. When the missing element is willpower, "Man up, already!" is perfectly viable strategic advice.
Accept an annexation in return for promises of stability? Hmm, reminds me of something...
That's partly the point, we didn't go that route and now have the mess we have at the moment.
And what happened the last time we DID go that route?
Making decisions because on a single data point is not good policy.
Also the alternative to the Munich agreements would have been to start WWII earlier. That might have had advantages but it would still have been very messy.
Sometimes none, if the source of the evidence is biased and you're a mere human.
There are unbiased sources of evidence now?
Some sources of evidence are less biased than others. Some sources of evidence will contain biases which are more problematic than others for the problem at hand.
Of course. But Rowan seemed to be arguing a much stronger claim.
That question doesn't have anything to do with the claim that you can make someone less informed by giving them biased evidence.