Our Coming Robot Revolution

Is the Atlantic becoming the National Enquirer? This month, the cover blares, "Why Machines Will Never Beat the Human Mind," which sounded pretty fishy to me. But potentially interesting! So I read the cover story, which turns out to be a fairly modest piece about a guy who took part in this year's Loebner Prize competition, which is designed to find a computer that can carry on a five-minute conversation so realistic that a panel of judges (or 30% of them, anyway) think they're actually talking to a human being.

OK fine. So far no computer program has ever won, and last year the computers lost yet again. But why does this mean machines will never beat the human mind? Brian Christian is the author, and after 8,000 words of saying nothing at all on this subject, he finally says this:

When the world-champion chess player Garry Kasparov defeated Deep Blue, rather convincingly, in their first encounter in 1996, he and IBM readily agreed to return the next year for a rematch. When Deep Blue beat Kasparov (rather less convincingly) in ’97, Kasparov proposed another rematch for ’98, but IBM would have none of it. The company dismantled Deep Blue, which never played chess again.

The apparent implication is that—because technological evolution seems to occur so much faster than biological evolution (measured in years rather than millennia)—once the Homo sapiens species is overtaken, it won’t be able to catch up. Simply put: the Turing Test, once passed, is passed forever. I don’t buy it.

Rather, IBM’s odd anxiousness to get out of Dodge after the ’97 match suggests a kind of insecurity on its part that I think proves my point. The fact is, the human race got to where it is by being the most adaptive, flexible, innovative, and quick-learning species on the planet. We’re not going to take defeat lying down.

No, I think that, while the first year that computers pass the Turing Test will certainly be a historic one, it will not mark the end of the story. Indeed, the next year’s Turing Test will truly be the one to watch—the one where we humans, knocked to the canvas, must pull ourselves up; the one where we learn how to be better friends, artists, teachers, parents, lovers; the one where we come back. More human than ever.

Seriously? That's it? That's what the cover headline is based on? Surely a better answer is that IBM built Deep Blue solely for its PR value, and once IBM won they had gotten all the PR out of it that they ever would. Win or lose, there was no point in continuing.

Plus there's the fact that computers have gotten better since 1997. Hell, there are mobile phones that play grandmaster-level chess these days.

What a letdown. I was hoping the piece would actually have something interesting to say about AI, but it didn't. This is especially disappointing because I'm so thoroughly convinced that human-level AI is not just possible, but inevitable. I figure all we need is hardware about a million times more powerful than we have now (current hardware doesn't allow us to come anywhere close to human-level AI because current hardware has about the same processing power as an earthworm) plus another few decades of software development. But I might be wrong! And I'd like to hear a good argument from anyone who thinks I am.

I often think that the reason I'm so bullish on artificial intelligence isn't because I have such a high regard for AI, but because I have such a low regard for biologic intelligence. Christian argues that computers are good at analytic, left-brain stuff, but humans will fight back with our awesomely passionate right-brain abilities (i.e., we'll "learn how to be better friends, artists, teachers, parents, lovers.") But you know what? I've watched The Bachelor, and it mostly proves that love can be cynically engineered just as well as any suspension bridge.1 Human emotions, as near as I can tell, are just as much the result of prosaic neurochemical reactions as anything else, and there's really no reason to think that computers won't eventually be able to emulate them just as well as they emulate anything else. Maybe better. After all, it's not emotion that separates us from the rest of the animal kingdom. Human behavior just isn't as deep as we like to think.

But like I said, maybe I'm wrong. Christian's piece sure didn't make much of a stab at persuading me, though.

1I'm serious about this, which makes me sort of gobsmacked at the show's popularity. Don't viewers realize that the real moral of the show is that you can pretty much guarantee to produce — real! honest! heartfelt! —  love if you simply follow a fairly simple set of cookbook steps? Why would anyone want to have that lesson pounded home season after season?

Keep Spouses Out of It

Here's the latest jockeying for position on the constitutionality of the healthcare reform law:

Seventy-four House Democrats have signed a letter to Clarence Thomas asking the Supreme Court justice to recuse himself from any deliberations on the constitutionality of the national health care overhaul, arguing that his wife's work as a lobbyist creates "the appearance of a conflict of interest."

....The House Democrats' letter follows a suggestion made by Sen. Orrin Hatch (R-Utah) last week that Supreme Court Justice Elana Kagan should recuse herself from any consideration of the health care law's constitutionality because of her previous position as U.S. Solicitor General.

I get that this is mostly just rhetorical jousting, and I get that Democrats are mostly just responding to Hatch. Still, I think it's a bad idea to be making this argument. We really shouldn't be promoting either the idea that judges' spouses need to be apolitical creatures or that judges are responsible for what their spouses do. For starters, even in our current enlightened era it's a lot more likely for a male spouse to be politically active than a female spouse, which means this kind of argument hurts women a lot more than men. And if it's true of judges, why not members of Congress too? And legislators? And mayors?

If there's actual money involved, that's one thing. But it's pernicious to suggest that politicians or judges are acting badly if they're involved in legislation that's merely supported by their spouse. We should be beyond that by now.

Webb Bows Out

Sen. James Webb (D–Va.) has decided not to run for reelection in 2012. So what does the Democratic grass roots in Virginia think of this? Here's Virginia email pal #1:

Crap. I know he wasn't the greatest advocate for progressive causes, but Webb was still a D. Now, he will relinquish his seat to the guy he beat 6 years ago, George Allen. No Democrat can beat Allen in 2012 unless he self-immolates again. I cannot begin to tell you how much of a Republican throwback Allen truly is.

That's not very encouraging. Maybe Virginia email pal #2 will be a little more optimistic:

This really pisses me off. He could have kicked Allen's ass. He has let a lot of people down and has probably handed another Senate seat to the Republicans. What a selfish prick.

OK then! Looks like Webb has a wee bit of fence mending to do.

The Power of Single-Mindedness

Via Tyler Cowen, Inc. magazine has a piece in its current issue about how entrepreneurs think. It's based on research from Saras Sarasvathy, a professor at the University of Virginia's Darden School of Business, and here's one bit:

That is not to say entrepreneurs don’t have goals, only that those goals are broad and—like luggage—may shift during flight. Rather than meticulously segment customers according to potential return, they itch to get to market as quickly and cheaply as possible, a principle Sarasvathy calls affordable loss. Repeatedly, the entrepreneurs in her study expressed impatience with anything that smacked of extensive planning, particularly traditional market research. (Inc.’s own research backs this up. One survey of Inc. 500 CEOs found that 60 percent had not written business plans before launching their companies. Just 12 percent had done market research.)

....Sarasvathy explains that entrepreneurs’ aversion to market research is symptomatic of a larger lesson they have learned: They do not believe in prediction of any kind. “If you give them data that has to do with the future, they just dismiss it,” she says. “They don’t believe the future is predictable…or they don’t want to be in a space that is very predictable.”

This reminds me of a study that I read about years ago. (No link, unfortunately, since I don't remember where I saw it.) The gist of it was that a team of researchers tried to figure out what made entrepreneurs different from ordinary corporate executives. Were they less risk averse? It turned out they weren't. Did they have different kinds of social and/or analytic skills? Not really. Were they more energetic? More visionary? Better able to understand new markets? No, no, and no.

So what was it? They discovered there was one metric on which entrepreneurs scored far higher than other executives: self confidence. Entrepreneurs, it turned out, weren't right about things any more often than other executives. But they were convinced of their rightness far more, and this fits pretty well with Sarasvathy's conclusions. Why bother with market research if you know you're right? Why bother with a business plan? Why wait to get to market? Why listen to other people's predictions? There's just no point in any of this stuff if you know you're right.

I don't think this is a hugely surprising result. And it certainly explains a lot about entrepreneurial success and failure. If you're right, and you're absolutely convinced you're right, it's a big advantage. You'll go full bore all the time, ignore distractions, and just generally take the most aggressive possible actions whenever you can. And if you're wrong? You'll do the same thing and flame out spectacularly. The power of single-mindedness is not to be underestimated.

Let's Tax Abortions!

Nick Baumann, who broke the story last week about the Republican anti-abortion bill that would have limited the definition of "rape," explains today what else is in the bill. Aside from making the Hyde Amendment permanent (currently it has to be renewed every year), there's this:

Another part of the law is a sweeping attack on tax benefits and deductions that affect abortion. It would, for example, forbid self-employed people from deducting abortion costs as medical expenses and would outlaw the use of funds from tax-exempt Health Savings Accounts to pay for abortions. In effect, this would raise the taxes of nearly anyone who had an abortion or purchased insurance that covered abortion. "Going after the tax subsidies that affect abortion" would represent a "substantial victory for the pro-life movement in America," Timothy Jost, an opponent of abortion rights and an expert on health law at Washington and Lee University, told Mother Jones last year.

Finally, a tax increase conservatives can love! But do Republicans have any chance of actually passing this bill? In the House, sure. But in the Senate, it would be immediately filibustered and there's nothing close to 60 votes in favor of passing it. However, Nick suggests that a more moderate version of the bill might have a chance if Republicans manage to attach it to some kind of "must pass" legislation.

But this leaves me confused: amendments can be filibustered too, can't they? The Senate considers must-pass legislation all the time, and if the minority party could attach amendments like this willy nilly, we'd see a helluva lot of minority amendments attached to must-pass legislation. But we don't. So something's wrong here. Can anyone explain to me whether this danger is real or not?

Obamacare vs. PPACA

Jon Cohn defends the term "Obamacare":

I get what the polls are saying. This is still not a particularly popular piece of legislation. People may not want to repeal it, but they're not about to celebrate it, either. I imagine the White House and the Democrats have strategists who have run surveys on this and concluded the term is not particularly helpful.

Even so, I like the term. I think this bill will be popular someday and, in the meantime, I think it's a reminder that this administration did something that will help millions of Americans while starting to put our health care system in order. Maybe I'm wrong — I've certainly been wrong about this before — but I think that within a few years, and maybe even by 2012, association with the health care plan will be a net plus.

I'd say that 2012 is pretty optimistic, but I certainly agree that eventually Obamacare will be a popular program. My question is whether this is ever likely to catch on. Social Security isn't RooseveltSecurity and Medicare isn't JohnsonCare. Presidents really don't get their names associated like this very often except with broad world views like Bush Doctrine, Reaganomics, etc. And even that's not very common.

The real problem here is that Democrats, once again, failed Legislation 101. This was their bill. They could name it anything they wanted. So what did they choose? PPACA. That's very memorable, isn't it? What's wrong with these guys?

WikiLeaks: Saudi Oil May Have Peaked Already

The Guardian reports today on another WikiLeaks cable, this time about oil production in Saudi Arabia. Based on conversations with Sadad al-Husseini, a geologist and former head of exploration at Aramco, the Saudi state oil company, the U.S. consul general thinks the Saudis have been significantly overstating both the size of their reserves and their production capacity:

The cables, released by WikiLeaks, urge Washington to take seriously a warning from a senior Saudi government oil executive that the kingdom's crude oil reserves may have been overstated by as much as 300bn barrels — nearly 40%....According to the cables, which date between 2007-09, Husseini said Saudi Arabia might reach an output of 12m barrels a day in 10 years but before then — possibly as early as 2012 — global oil production would have hit its highest point. This crunch point is known as "peak oil".

....The US consul then told Washington: "While al-Husseini fundamentally contradicts the Aramco company line, he is no doomsday theorist. His pedigree, experience and outlook demand that his predictions be thoughtfully considered."....While fears of premature "peak oil" and Saudi production problems had been expressed before, no US official has come close to saying this in public.

This won't come as a surprise to anyone who's been following the oil industry over the past few years. Matthew Simmons' Twilight in the Desert, which I reviewed six years ago, made a detailed case that Saudi Arabia's production capacity had pretty much maxed out already, and Business Week published an article three years ago based on internal Saudi documents that said much the same: the Saudis could pump 12 million barrels a day in short spurts but only 10 million barrels on a steady basis — and that's all there is. Production capacity just isn't going up.

There's always Iraq, of course, which certainly has more production capacity if it can develop it, but Saudi Arabia increasingly looks like it's peaked already. And if that's true, it probably means that the global peak in production, which was delayed a few years by the 2008 recession, is most likely not too far away. Our future is going to be increasingly oil free whether we like it or not.

Occupational Licensing: Should We Care?

Should states require that interior designers be licensed? How about manicurists? Hair stylists? Florists? Why not let people do these things without bothering to put hurdles in their way that, in the end, mostly just protect existing manicurists/hair stylists/florists from competition? It's not as if a bad manicure is really that big a deal, after all. Today, Felix Salmon takes a crack at making the progressive case for licensing:

Broadly speaking, the more constraints you have on a profession, the less likely you are to see massive inequality within that profession. If you got rid of licensing for profession X, you’d see many more low-paid Xs than you do right now, and you’d also see a significant uptick in earnings at the very top of the X profession. It’s a second-order effect, to be sure, but I’m pretty sure that at the margin, licensing helps to reduce inequality.

I’m in favor of reducing the number of gratuitous licensing laws; I’m also a fan of motherhood and apple pie. But at the same time I think they are, in a sense, a form of worker protection which is acceptable to Republicans — think of them as unions for people who hate unions. And that’s not entirely a bad thing.

Hmmm. Putting aside the question of whether this is a good justification for licensing regimes, is it actually true? Does licensing really reduce income inequality within professions? Felix links to an article about growing inequality at the tops of big law firms, but I'm not sure how that supports his case. Law, after all, is one of the most strongly licensed fields we have. If three years of law school plus a bar exam plus large amounts of professional regulation aren't enough to keep income inequality in check in the legal field, I'm not sure that nine months of barbering instruction would have much effect at all in the barbering field.

Beyond that, though, I'd like to have some idea of the raw numbers here. Take a look at this page, for example, which lists the licensed occupations in the state of New Jersey. Barbers are there, of course, and maybe they shouldn't be. I'm not sure. But the majority of the occupations are in the healthcare industry, where I think we can widely agree that some kind of certification is a good idea. And even some of the others look pretty justifiable if you think about them a bit. Plumbers and electrical contractors? The state probably has a legitimate interest in making sure that nitwits don't damage public property (sewer systems) or cause neighborhood fires because they don't know what they're doing. Locksmiths and burglar alarm installers? That might be a crime control thing. Landscape architects? I'm not sure about that one. But maybe drainage issues are important for more than just the house getting the landscaping?

Still, it's mostly healthcare. The last time I had some blood drawn I asked how one gets to be a blood drawer. The answer, first of all, is that the proper term is phlebotomist. And second, you take a six week class in order to get certified. (In California, anyway.) During the first few weeks of this class the aspiring phlebotomists learn the basics by drawing each other's blood many times each day. Sounds grim, doesn't it? My phlebotomist assured me it was. Still, she did a pretty good job of drawing my blood, for which I'm grateful. So has every other phlebotomist who's ever drawn my blood. That sounds like a win for phlebotomy certification.

In any case, my question is this: sure, you can find silly cases like Louisiana's florist licensing requirement. And the hairdressing/manicuring industry is a big one where licensing requirements are at least questionable. But overall, what percentage of licensed professionals are in sectors where the requirement is genuinely dubious? Once you take out the healthcare folks, where licensing seems pretty reasonable, as well as the other areas where licensing seems justifiable, how much is left outside the barbering/manicuring industry? Are the funny examples of stupid licensing requirements really widespread, or are they just the kind of rare horror stories that all human endeavors inevitably produce? In other words, is runaway occupational licensing really a big deal? Or just a minor issue that deserves a minor bit of attention? It would be nice to see some actual facts and figures on this.

UPDATE: Adam Ozimek has a useful response here. It doesn't have all the figures I'd like to see, but it does have a few comparative numbers of interest.

How To Tell Time In Germany

Like those apocryphal Eskimos with their endless names for snow, Frank Jacobs reports that the famously punctual Germans have four different ways to tell time:

In a large part of north-western Germany, from the Danish to the French border, the preferred option is viertel nach zehn ('quarter after ten')....In what used to be East Germany, the same clock time is referenced as viertel elf ('quarter eleven')....An option limited to German-speaking Switzerland is to call this particular time viertel ab zehn ('quarter from ten')....Another national Option is viertel über zehn ('quarter over ten'), used only in central parts of Austria.

Wait a second. Shouldn't there be five ways to say this? What about the German version of "ten fifteen"? Doesn't anyone use that? After all, what do they do when it's 10:12 and there's no handy "quarter" or "half" shortcut to use? 

Centrist Dems Plot Against the Mandate

Politico reports that a few centrist Dems are planning to do what centrist Dems always do:

A handful of moderate Senate Democrats are looking for ways to roll back the highly contentious individual mandate — the pillar of President Barack Obama’s health care law — a sign that red-state senators are prepared to assert their independence ahead of the 2012 elections.

....And it’s not just health care. The senators are prepared to break with the White House on a wide range of issues: embracing deeper spending cuts, scaling back business regulations and overhauling environmental rules. The moderates most likely to buck their party include [the usual suspects....]

Second things first: on non-healthcare issues, I understand this entirely. Red-state Democrats want to demonstrate their moderate bona fides in order to shore up their right flank before the 2012 election. This is standard behavior, completely predictable, and might even work.

But on healthcare, I wonder what they could possibly be thinking? The individual mandate can't simply be rolled back, it would have to be replaced. And there are ways to do that. But there's no way that any Republicans will ever vote for any of them. Their base wouldn't stand for it, it would wreck their chance of having the Supreme Court throw out the whole thing, it would tacitly accept that PPACA is here to stay, and it would provide political cover for vulnerable Democrats. Hell, Republicans wouldn't even vote for repeal of the 1099 provision last year — a move that virtually everyone in both parties supported — because it might have taken a slight bit of tea party pressure off of Democrats. And that was after the election. There's just no way that a repeal of the invididual mandate will gain even the slightest traction among conservatives. The whole thing reeks of desperation.

Politics being what it is, it's vanishingly unlikely that there will be any substantial changes to PPACA until after the 2012 election and after the Supreme Court rules on the mandate. Until then, Republicans will hold out hope that they can use it to fan the flames of tea party fury and either (a) win the 2012 election or (b) win the Supreme Court case. Or both! Only once that's over, and PPACA is fundamentally here to stay, will it be possible to enact any improvements to the law.