Topic: General

A.I. Yai Yai!

Economist Robin Hanson suspects that the world economy may soon be doubling every week or two. He arrived at that suspicion based on historical extrapolations, but he also has a theory as to how it might happen: the development, in the near future, of intelligent machines.

According to Hanson, efforts to computationally model the human brain, neuron by neuron, could reach fruition within the next 25 to 50 years. He plays pretty fast and loose with the details, though, so let’s take a closer look at the bleeding edge of the field.

The mother of all brain simulation projects is Blue Brain, a joint project of IBM and Ecole Polytecnique Fédérale de Lausanne (a Swiss town also known for its lower tech, but tastier, fondue). Announced with much fanfare in 2005, Blue Brain has as its anything-but-modest mission to create a complete and exhaustively accurate simulation of the human brain within a decade or so. They figure they can knock out the neo-cortex in the next few years.

Somebody buy these guys a calculator.

The current Blue Brain hardware has 8,000 processors and they have apparently set it up so that one chip models one or two neurons. That has allowed them, as of this month, to model a 10,000 neuron grouping called a “column.” That’s hugely impressive. But, umm, 10,000 down, 99,999,990,000 to go.

The human brain is estimated to have about 100 billion neurons, so they’re going to need another 10 million or so Blue Brain computers to finish the job given the current specs. They’re having a tough time convincing the Swiss government to spring for another one or two of them.

Mind you, computer processing power per dollar has been increasing exponentially over time since the earliest electromechanical computers. So maybe, in a generation, we’ll be revisiting this question. But that’s an awfully big maybe.

It seems doubtful that we’ll be able to create Marvin the paranoid android any time this century. And it’ll be some time after that before the technology is commercialized and we all have “plastic pals” who are “fun to be with!”

So keep contributing to that 401K. It’s gonna be a while before it starts doubling every week or two.

Since When Does the Right Have Any Health Care Prescriptions?

Ezra Klein responds to my post on the Citizens’ Health Care Working Group, where I lament that its tax-and-spend approach to health care shows that it was hijacked by the left—despite having a business-community chairman and Bush’s HHS Secretary on the panel. Klein provides an alternative interpretation: “the right’s prescriptions on health care didn’t even convince the panel’s Republican ringers.”

I’m new to the blogosphere, so I’m not clear what the rules are regarding subtlety. My point was that the right doesn’t have any prescriptions or even premises when it comes to health care—at least, not of their own. To compensate, they adopt the left’s premises (“we need to expand health coverage”) and prescriptions (“we need a Medicare Rx benefit”). For examples, click here, here, here, here, here, here, here, here, and here. Even when the right comes across a good idea (e.g., health savings accounts), they often gravitate toward it not because they understand why it’s a good idea, but because they think it serves leftist premises (“HSAs will … reduce the number of uninsured!”). The exception that proves the rule would have to be the GOP’s Rep. John Shadegg and Sen. Jim DeMint, who have a smashing health care proposal and who know why it’s a good proposal.

In general, Republicans and conservatives do not understand health policy, and do not have a health policy agenda that distinguishes them from the Left. Then, lo and behold, we get the GOP’s top health care guy sitting on a panel that (tentatively) recommends universal coverage financed by tax increases.

Go figure.

Private Answer to Stem Cell Debate

Harvard has announced that it is launching a privately funded, multi-million dollar program to clone human embryos for use in stem cell research. In this 2004 column, I argued for exactly this kind of private sector initiative to solve the politically divisive debate over stem cells.

I wrote in part,

By its very nature, government politicizes everything it touches. Science is no exception. Stem cell research needs neither government money nor politics. It is better to get the government out and let the private sector continue its good work. Those people calling for increased funding could take out their checkbooks and support it. Those who oppose embryonic stem cell research would not be forced to pay for it.

Harvard is proving one again that civil society can do what government can’t.

Taylor vs. Corn

FYI, I’ll be on Kudlow & Company (probably around 5:30 EST) to debate ethanol subsidies with Frank Gaffney.  

Something to chew over while you wait for the beating to come: wholesale ethanol is selling today for July delivery at the Chicago Board of Trade for $3.40 a gallon.  Given that ethanol has only two-thirds the energy content of conventional unleaded gasoline, we have to multiply that price by 1.5 if we want to compare apples with apples.  So to get the same amount of energy from ethanol that we would get with a gallon of conventional unleaded, we would have to pay $5.10 a gallon.  What is the wholesale price today for conventional unleaded for July delivery?  All of $2.10 a gallon at the NYMEX.

But that’s not all.  Gasoline moves from refineries to retail distribution centers via pipelines, and transportation costs are low.  It costs a lot more money, however, to move ethanol from processing plants in the Midwest to retail distribution centers because it must be moved by truck and barge (one can’t use pipelines to move ethanol for various technical reasons).  So add another couple of dimes to the differential between the price of ethanol and conventional unleaded to account for that, and perhaps another dime or more if you’re shipping that ethanol to the Atlantic or Pacific coasts.

If the answer is ethanol, what exactly is the question?

A Case for a Different Libertarian Party

All of this blogtalk about which major party is likely to be more receptive to libertarian policy positions, I suggest, is a waste of time unless the winning candidate of either party is dependent on the votes of libertarians.

Increased outrage about the state of American politics and the prospect for a larger number of close elections increases the potential effectiveness of a different libertarian party – one that sometimes endorses one or the other major party candidate but does not run a party candidate for that position.

The Libertarian Party’s efforts to promote their policy positions by running Libertarian candidates is counter-productive when they reduce the vote for their favored major party candidates. A disciplined group that is prepared to endorse one or the other major party candidate in a close election, however, can have a substantial effect on the issue positions of both major party candidates. The following conditions must be met to achieve this effectiveness:

  1. The party cannot run a separate candidate.
  2. The size of the party must be larger than the expected vote difference between the major party candidates.
  3. After the major party candidates are selected, the party leadership must have the opportunity to bargain with both major party candidates on the issue positions of highest priority for the party.
  4. The party, as much as possible, must act in concert to support the major party candidate who is chosen by the members of the party in that district.

There is no reason for this libertarian party to be active in any district for which the party does not meet all four of the above conditions. (For most libertarians, the most difficult of these conditions to meet, I suspect, is condition 4.) In addition, the party should not emphasize the same issues in every district, because the choice of these issues should depend on those for which the major party candidates are willing to bargain.

This is a strategy to increase the approval of libertarian policy positions rather than the usually counter-productive effort to increase the number of votes for Libertarian candidates. Maybe it is better to term the organization that I have described as a libertarian political action group, not a libertarian party.

Marriage Amendment Failure

Supporters of the Marriage Protection Amendment say that even though it failed in the Senate on Wednesday, they are pleased that it did better than two years earlier. But let’s do the math. In 2004 supporters lost a cloture vote 48-50, with two opponents not voting. So their strength on moving the amendment to a floor vote was 48-52. This year the vote was 49-48, far short of the 60 needed to invoke cloture or the 67 for a constitutional amendment. If all senators had voted, the vote would likely have been 50-50. So maybe that’s a pickup of two votes for amendment supporters.

But the Republicans picked up four Senate seats in the 2004 election. So relative to the number of Republicans in the Senate, support for the amendment actually slipped by two votes. Supporters picked up no Democrats, and they lost two Republicans. Judiciary Committee chairman Arlen Specter voted for cloture in 2004, though he would have voted against the amendment itself; this year he voted against cloture and quoted two Cato publications in his Senate speech. Judd Gregg joined his New Hampshire colleague John Sununu in voting for federalism over centralism after realizing that the 2003 Massachusetts court ruling for marriage equality in that state is not being replicated nationwide. Given that younger voters are much more supportive of same-sex marriage than older voters, it seems unlikely that support for an amendment will grow in future years.