Springboarding off my earlier Climate Science post, and now Borepatch's post...
- Does world average temperature mean anything useful at all? Or is it largely a useless aggregation of data that don't really go together?
- How do you square evolving as distance runners vs. doing high-intensity anaerobic exercise?
- What was the ESE like? How long was it stable? Was there a big genetic jump between the ESE and now? Ridley posits the introduction of trade pretty recently...does that invalidate a lot of prior ESE thinking, or just make it psychologically difficult for humans?
- Which factors actually matter most to life-outcomes out of: Big-5 Personality (esp. Conscientiousness), Self-Efficacy, IQ, Positive-ness, Patience, Social/Emotional Intelligence? (Aretae guess: Patience, Self-efficacy, Social Awareness, IQ, Big-5, Positivity). Is there something measurable and sensible that matters even more? (Aretae guess: yes -- motivation). How do you measure motivation independent of action?
- Do we actually have >99% certainty on anything at all? Drinking 9000 calories a day of Coke is bad for you is a bit narrower than I mean.
- How best to implement scarce, saleable titles of nobility which could work as a status-competition for rich folks with real benefits, but while minimizing the distortionary effects on real goods.
- What would count as disconfirmatory evidence for the formalist hypothesis, even in a partial, Bayesian sense?0
Posts
7 comments:
"What would count as disconfirmatory evidence for the formalist hypothesis, even in a partial, Bayesian sense?"
Hmm, I would have to understand exactly what the formalist hypothesis is first.
Competitively what would be the libertarian hypothesis?
Grim,
I think the formalist hypothesis is something along the lines of:
Efficiency and effectiveness are the features of government most amenable to a good life for the citizenry.
The libertarian hypothesis:
Markets are better than government for solving effectively ALL problems.
Efficiency and effectiveness are the features of government most amenable to a good life for the citizenry.
I'm reasonably certain that history has shown that efficient government is pretty much the opposite of GOOD government.
What would count as disconfirmatory evidence for the formalist hypothesis, even in a partial, Bayesian sense?
The nice thing about formalism is that a) it's more a set of practical wisdom than a grand ideology and b) it's a nascent and changing philosophy.
So every time I encounter evidence against my formalist principles, I update the principles to reflect the new information.
For example, studying Leopold in Belgium or the feudal princes in medieval Germany did make me rethink some things. Witnessing an actual thriving anarcho-capitalist system in Africa or Latin America, or witnessing a thriving well run democracy in one of those areas, would definitely make me reconsider elements of formalism. Or if a formalist experiment was tried, and an African country converted to rule by a legitimize, for-profit corporation, and then that corporation liquidated the residents for biofuel, that would make me rethink formalism too.
Efficiency and effectiveness are the features of government most amenable to a good life for the citizenry.
I'm not crazy about that definition.
I just tried to take a stab at defining formalism succinctly, chez moi.
Average world temperature does make sense as a concept, but only if you remember three things:
1. The noise (annual variations) swamps the signal you're trying to measure. It's likely 100x the signal, in absolute terms.
2. The signal is only meaningful over multi-decadal periods. The silliness we read in the paper is, well, silly.
3. We only have really good temperature data going back to 1979 (satellite measurement). We have data of questionable quality for some of the world going back to before WWII. We have data of questionable quality for a small part of the world (the USA and Europe) going back to the mid-nineteenth century. We have essentially no temperature data at all for the parts of the globe covered by ocean, except for the satellite data. We have "proxy" data that seem to sort of match some of the temperature data maybe 60% of the time. But proxies are the only way people can even guess what the temperature was thousands or millions of years ago.
Remember #3 especially when you hear mouth breathing journalists huffing stories about "hottest year of the millennium". The error bands are simply enormous, and the chance that this is actually true is very small indeed.
Borepatch,
I appreciate those critiques...but does it even work as far as you're saying? Don't we see situations where the Artic is 3 degrees warmer, but the tropics are 3 degrees cooler being listed as a 1 degree decrease in global temperature...which tells us almost nothing about life anywhere? Of am I missing some of the data about how it's useful?
Aretae, the only thing that we know with confidence is that the data we have are very unreliable. The refusal of Mike Mann, the CRU, etc to release their data is all you need to see to understand how solid the scientists think the data is.
The Virginia Attorney General is suing Mann to get the data and computer code that Mann developed while being paid by the state of Virginia. That's got to be some rock-solid code and data, right there ...
[/sarcasm]
My point is that the only thing we know about longer than 80 year climate variations is "we don't know very much". Your comment is a great example of how the data we have is anything but obvious.
I elaborate at some length here.
Post a Comment