I have to respond to your yootles critique first!
  Our aim is both to decrease to next to nothing the overhead of applying 
more formal mechanisms to decision-making (and more recently, prediction 
and prediction+decisions) as well as convince you there's plenty to be 
gained.  Bethany and I yootle every day for every little (and big) thing 
imaginable.  (Being both indecisive types it often saves us a ton of 
time.) Granted, we're gigantic dorks and no one else cares yet.

I'm of course not done defending my boyfriend Paul either.  More on that 
later.

But I'm actually delighted that we're making real progress on 
circumscribing the disagreement while identifying common ground (eg, 
Graham is at least right in some other fantasy universe; Trixie won't like 
that concession at all!  and of course I conceded on slavery and may do so 
on health care, where basic human rights are at stake.. although Kevin 
may set you straight on the healthcare issue!).


--- \/	 FROM James W Mickens AT 07.08.31 22:21 (Yesterday)   \/ ---

>> You're characterizing our disagreement as hinging on whether public
>> policy should optimize economics subject to moral constraints or
>> optimize morality subject to economic constraints.  I'm unclear on what
>> either of those really mean for public policy.
>
>
> There's a difference in the intent of your policy and the methods that you 
> will use to evaluate it. For example, suppose that you've devised a new tax 
> code for an underdeveloped country. When you evaluate its success, will you 
> look at how much additional wealth it generated, or some actual measure of 
> utility such as the percentage of citizens who have access to electricity or 
> clean water? I use the term "actual measure of utility" because I think that 
> just examining, say, the increase in GDP is a bad way to measure net social 
> welfare. The net wealth of a society is, at best, an indirect measure of its 
> net welfare because aggregate wealth trends tell us nothing about the 
> *distribution* of wealth or whether that wealth is being used to satisfy some

> particular goal. The classic example is health care. Despite rising levels of

> aggregate wealth in America, many *individual* Americans have poor health and

> inadequate access to proper medical care. Is the solution to this problem the

> creation of even more wealth in the hope that the health care industry will 
> spontaneously reorganize? Or is the solution a targeted policy, whether it be

> nationalized health care, better health education in schools, and/or 
> something else? I argue that the latter approach would be better, 
> particularly since the market has thus far been ineffective in addressing 
> this issue.
>
> The failure of wealth-driven policies is even more obvious in the 
> international pharmaceutical market, where drug companies develop medicines 
> for diseases that affluent people care about (e.g., restless leg syndrome, 
> diabetes) and ignore a huge number of illnesses (e.g., diarrheal diseases) 
> that affect a much larger number of people who have much less money. People 
> who care about net welfare should find this problematic. So, in the 
> international drug market, should we pursue wealth-driven or morality-driven 
> policies? In other words, should we allow drug companies to maximize their 
> profits and hope that they'll turn a charitable eye towards the developing 
> world, or should we force them through regulation, subsidies, tax credits, 
> etc., to address the needs of poorer countries? History suggests that the 
> former strategy will fail if you're trying to optimize for health and not 
> profit.
>
> I understand that it is extremely expensive to develop new drugs and that 
> pharmaceutical companies must be given a way to recoup these costs. However, 
> it's obvious that a market system which focuses on maximizing their profits 
> will not lead to a net increase in global health (an important utility 
> metric).
>
>
>
>> Let me first defend Graham's point.	He concedes whole classes of 
>> exceptions and I think social injustices are included, if not explicitly. 
>> His argument -- that income inequality is not, inherently, unjust -- 
>> remains intact.
>
> Once again, I'm claiming that economic justness does not equal moral 
> justness. When you say that income inequality is "not inherently unjust," you

> should specify whether you refer to the economic definition, the moral 
> definition, or both.
>
> If Graham includes social injustice in his exceptions list, then I suppose 
> that he and I are in agreement. But if Graham believes in the entrenched, 
> pervasive nature of social injustice, why does he spend so much time waxing 
> poetic about the inherent fairness of economic inequality? This fairness only

> exists in an idealized model of the economy which bears little resemblance to

> the real one. The fact that Graham spends most of his time talking about this

> idealized world suggests that either a) he is a hopeless utopian, or b) he 
> does not, in fact, believe that social injustice is entrenched and pervasive 
> ;-).
>
> ~j
>
>
> p.s.
>
>> And just to nip a potential subthread: the non-mathematically inclined are 
>> not allowed to blithely declare human motivation to be irreducible to 
>> mathematics.
>
> Ah, but I claim that the mathematically inclined are not allowed to blithely 
> declare that human motivation *is* reducible to mathematics ;-). The reason 
> that I do not use yootles to determine who will pick me up from the airport 
> is that, in the common case, this decision is not subject to rigorous 
> mathematical or economic constraints, nor should it be. In many scenarios, I 
> only care about approximate notions of fairness. I suppose that if gasoline 
> were $27,000 a gallon, it might be reasonable to employ a strong mathematical

> framework to prevent tragedy (e.g., "Oh no, Todd has taken me to the airport 
> fifteen times but I haven't taken him at all. Todd has now spent $405,000 on 
> gas while I have escaped scot-free."). Absent such extreme conditions, the 
> introduction of mathematics into simple human transactions will often just 
> add overhead and produce little tangible benefit.
>
> It is frequently possible and fruitful to analyze people's behavior using 
> mathematical models. However, that doesn't mean that the underlying 
> psychology of the individual is actually driven by these models, or that 
> giving the math to people will make it easier for them to manage their lives.
>

-- 
http://ai.eecs.umich.edu/people/dreeves  - -  search://"Daniel Reeves"

Dilbert:     "...and we'll buy a dozen of these. We're trying to
	      spend our budget so it doesn't get cut next year."
Salesperson: "This is great! You guys are so dumb that I don't even
	      have to use my fake personality to make the sale!"