Unqualified Offerings

By Thoreau

The Finest News Source manages, in 692 words, to say everything that needs to be said about the folly of the wars that the US fights.

By Thoreau

How many people, when they fill out an organ donor card, think to themselves “Who knows, my last act in this world might be to save the life of a Sith Lord”?

Me, I’ve given my wife explicit instructions to give my body to a medical school.  Being very cutting-edge and interdisciplinary, I want my last act in this world to be teaching a graduate-level course outside my field.

By Thoreau

A few weeks ago, when we were discussing the miserable prospects for raises, I said that if this place won’t compensate us we should hit the job market.  A cow-orker started in with some bleeding-heart nonsense about “You know, what we do here is very important.”  The subtext believes that because we are saving the world (or something) we have no right to complain about anything.  (This person also has a rather agnostic view about whether saving the world involves academic standards.)  Anyway, I have a retort:

It’s nice to be liked, but it’s better by far to get paid.  I know that most of the friends that I have don’t really see it that way.  But if you could give them each one wish, how much do you want to bet that they’d wish success for themselves and their friends?  And that would include lots of money.

Translated into academic-speak:

It’s nice to be collegial, but it’s better by far to get paid.  I know that most of the colleagues I have pretend not to see it that way.  But if you could give them each one wish, how much do you want to bet that they’d wish success for themselves and (maybe) their colleagues, and that would include lots of release time, summer salary, consulting fees, and royalties?

By Thoreau

There’s a lot that’s been said about the Treyvon Martin case, and I can’t and won’t rehash it all.  Suffice it to say that it is tragic, and based on the accounts I’ve read I find it difficult to see any scenario in which Zimmerman’s actions can be considered justified.  Even if, for some reason, he honestly believed that he was in danger, my understanding is that the legal standard for self defense is not what the individual believes but rather what a “reasonable person” believes.  I have no reason to believe that Zimmerman was particularly reasonable.  You can say something about Florida’s “stand your ground” law and the flaws thereof, and while I’m not here to defend that law, my understanding is that if anybody would have justification for the use of force under that law it would have been Treyvon Martin, i.e. the person who was stalked and confronted by a crazy man.  I say that not to defend the “stand your ground” law, but simply to point out how weak Zimmerman’s case is even under Florida’s laws.

All that said, I want to go a bit contrarian just for a moment:  I find it a bit awkward to come here and argue that the real problem is that a man is not in jail.  Usually, unless I’m talking about the crimes of the security apparatus, I’m arguing the exact opposite.  And, indeed, I get a bit uncomfortable when there’s a rally calling for somebody to go to jail.  But lest you think I’m here to go after the people at that rally, let me say that the blame for that bad situation lies squarely with the police.  It may very well be that, when all available and legally admissible evidence is examined by a jury, they will find reasonable doubt as to whether all of the necessary elements of a homicide (or whatever charge) were present, and so an unsympathetic defendant will be acquitted.  (Notice how many weasel words I put in there.)  I can live with that.  However, what I cannot live with is the fact that the police did very little to investigate this until the heat was turned up.  Even in the most lenient of all plausible interpretations of the facts, at the very least the police should have investigated more thoroughly than they did.  Does anybody honestly believe that, if the victim had been paler and the shooter darker, there would have been so much foot-dragging by the police?

And so we come to race.  One needn’t have any particular insight into the mind of any individual in this case to note that things always seem to work out certain ways when the skin colors clash.  Moreover, while Zimmerman was most definitely NOT a cop, he was a guy claiming to have confronted a criminal.  Cops always seem to get away with misconduct, so it should be  no great shock that when a guy with dark skin was shot by a paler man claiming to be on the lookout for crime, the cops were in no great hurry to investigate  thoroughly and do anything that might lead to an embarrassing precedent.  And so we get back to the fact that usually I’m the guy saying we need to lock up fewer people, not more people.  The dynamic here is the same:  The same cultural and political factors that generally lead to too many people behind bars also work in favor of the shooter here.

Now, I want to turn to the President’s remarks.  Different questions will get different answers from me.  If you ask me whether I am happy with the mainstream, establishment politician making policy in the Oval Office right now, I will say no.  He has continued the status quo on far too many things.  However, given that we  have a mainstream, establishment politician making policy in the Oval Office, when we go past policy to moments of societal dialogue like this, I am glad that the guy at the bully pulpit is black.  John McCain, John Edwards, Hillary Clinton, and Mitt Romney would all be perfectly capable of bombing Libya and failing to reduce the national debt.  However, they could not point out that Treyvon Martin looks like he could be the Honor Student son of a President.  The awfulness of the mainstream policy juggernaut is going to roll on, and it’s going to crush a lot of all colors in the process.  However, having a man with dark skin in the Oval Office sends certain messages and makes certain conversations possible.  America still has a very, very long way to go in its conversation on how black people are treated by their neighbors and by authorities, but hopefully that conversation will eventually have some impact on the way that the state treats black people.  Perhaps one consequence will be that in some future day the state does not lock up so many black men or excuse the shooting of black men.

And now I want to turn to the crimes of the security apparatus:  We all know that one rarely sees a black person get away with shooting a white honors student but often sees a white person (usually but not always in uniform) get away with shooting a black honors student.  Just as the conversation on that is slowly moving forward, I invite you to consider that attitudes about color, culture, and religion are part of why one sees the deployment of flying killer robots when a brown Muslim fails to detonate a bomb, but one sees jury trials when a white guy shoots an abortion doctor or plots to blow up a government building.  Perhaps, some day, we can have a conversation about that, one in which white non-Muslims (and, frankly, a lot of non-Muslim people of color) manage to say something other than “Look, we have to be safe.”  (For some reason, the “we” in that statement never includes the Muslims who happened to live near the site of the missile strike.)  I would hope that a guy named Barack Hussein Obama, with relatives in Kenya and Indonesia, could join a conversation like that.  Alas, he’s too busy explaining why he doesn’t need permission to use flying killer robots on US citizens.  Well, he himself isn’t Muslim, and he gets a lot of political advantages from this war footing, so I guess it makes sense that he’d be as irresponsible as anybody else in this country.  One can hope that, maybe, some day, he’ll say “You know, I have a cousin who looks like the victim of that drone attack…”  (And, for the record, on the southern Italian side of the family, I have an uncle who looks like he could be Arab.)

By Thoreau

I have just learned of a startup called Tacocopter that hopes to use unmanned aerial vehicles to deliver tacos.  Alas, The Man is keeping them down:

“Current U.S. FAA regulations prevent … using UAVs [Unmanned Aerial Vehicles, like drones] for commercial purposes at the moment,” Simpson said over Gchat. “Honestly I think it’s not totally unreasonable to regulate something as potentially dangerous as having flying robots slinging tacos over people’s heads … [O]n the other hand, it’s a little bit ironic that that’s the case in a country where you can be killed by drone with no judicial review.”

Indeed.

I am not optimistic for them to clear regulatory hurdles in the near future.  Between the safety-Nazis who would worry about these aircraft buzzing around cities, the nutrition-Nazis who will oppose anything related to fast food, and the Nazi-Nazis who will oppose anything that sounds Mexican, they’re up against a formidable coalition.  Their best bet is to persuade the military-industrial complex that drone delivery services represent a chance to grow into new markets.  (At which point it will become illegal to get fast food delivered by people in cars, enforceable by Predator drone strikes on delivery vehicles.  Given the ethnic composition of the pizza delivery workforce in southern California, this might win over the Nazi-Nazis.)

And, since I haven’t quite Godwined this thread enough yet, let me just add that there are also opportunities for soup delivery.

Update:  It turns out that one of the people involved with the project, Star Simpson, was also the MIT student whose shirt with a blinking light display elicited an armed response at the airport in 2007.

By Thoreau

Medical marijuana dispensaries in Los Angeles are unionizing. This is part of a bid to show that medical marijuana dispensaries are transparent organizations with no associations with organized crime.

In all seriousness, this is good news:

The employees joined the ranks of grocery workers, healthcare providers and pharmacists at the United Food and Commercial Workers, Local 770. At a news conference Thursday, the president of the union vowed to leverage the “full force” of its 35,000 members to keep dispensaries open.

I realize that the path from medical pot to broader legalization is not a sure or straight one, but I support anything that moves any aspect of the marijuana trade into an open, regulated mainstream.  Bringing in unions, with rules and paperwork, just makes medical marijuana downright boring.  Associating it with a larger organization that has clout in one of the major political parties is, on balance, a good thing.  It’s a bit risky, because the medical dispensaries will view legalized recreational use as a threat to their business, but it means that the debate we’re having is now a “how much” debate, not an “if” debate.

Next up:  Growers!  And then, eventually, this one dude who’s cool and keeps a low profile.

By Thoreau

We published a paper recently.  We also posted the code used to generate the results.  The day after the paper came out, somebody found an error.  A potentially very serious error.  I went through a lot of fear and panic, and then I thanked the person who found it and my students and I worked on the problem.  The error was a rather embarrassing one, a mistake that we really shouldn’t have made.  The buck stops with me, and what it boils down to is that I had not gone line-by-line through code that somebody else wrote.  I had checked some of the code, but as the project got bigger and the students displayed more competence, I stopped checking line-by-line and started focusing on output.

I was ruthless on checking the output:  We checked every realistic case we could think of in which we knew what the answer had to be.  They came out right.  Or, more precisely, they came out clustered around the right result.  If they had all been exactly at the right result, that would have been just as problematic as a wrong result, because it would have meant that there were no statistical fluctuations in a simulation that really, really ought to have fluctuations.  And, sure enough, there were fluctuations, and of exactly the right magnitude.  Then we perturbed the inputs, sometimes by a little and sometimes by a lot.  The code still converged to the right answer exactly as we would have anticipated.  Then we fed in inputs that ought to give junky outputs, and we got junky results of exactly the sort we’d expect.  We ran somebody else’s code and compared our results with theirs in every realistic case that both codes could handle.  And we got the same results (within those little statistical fluctuations that we expected).  I spent lots and lots of time going over every little blip in the output.  We got the right results in the cases that can be solved by hand, we got results that match others’ code in cases that others have considered, and we got results that matched every qualitative expectation (right down to the junk).  So I was very comfortable submitting this paper.

But I didn’t check every single thing line-by-line.  So a mistake crept in, and that’s my fault.

Anyway, we didn’t contact the journal immediately, because we had already published mistakes once and if we were hasty we might end up publishing mistakes twice.  We did, however, contact a few people who had expressed interest in our work, and told them that we were fixing errors in the code and they should not use it until we got back to them.  To my relief, everyone (including the guy who spotted the error) was amazingly supportive and offered to help.  Publishing a serious mistake is the situation that every scientist dreads.  My students were wonderful and focused on this.  An old mentor who once faced a similar problem gave very good advice.  All of this made me very serene as I prepared myself for the possibility of a retraction.

We eventually fixed the mistake and rechecked all of our cases and got the same results.  It became apparent that the effects of the error were quite minor.  This should not be a surprise, given the checks that I described above.  If the mistake in the code had given us serious errors, we would have spotted them sooner.  Still, we went beyond the cases that we had checked, and went to some very unrealistic inputs before we finally got output that disagreed with other people’s code.  Also, our new code gave “better” results, which is not surprising since our new code was (correctly) implementing an algorithm that has been mathematically proven to give the best possible results (I’m glossing over some details here, but you get the drift).  Based on that, I was confident that we really understood the error.

So, no retraction needed, because the results were the same for all realistic cases, and in our paper we only made claims about the performance of our technique for some physically realistic cases.  However, an erratum (basically, a short notice in the journal saying what the mistake was and if it affects the results) was still needed.  I paused for a few weeks before writing it.  I know that I have an obligation to report errors quickly, but I also have an obligation to make sure that I understand things before I go to the journal.  I needed to clear my head.  So I cleared my head, then I took one more look at the new results, quizzed my students a few more times, nitpicked it a few more times, re-checked the code yet again, then wrote a short article saying that our code mishandled an important step in the calculation, we have posted corrected code (URL given in the paper), and we have redone the calculations in Figures 1 and 2 and have gotten the same results to many decimal places (revised figures included in erratum).

Anyway, I’ve now been through every scientist’s worst nightmare, but I’ve survived, and I have learned that there are people who still respect my work and I have learned to be more careful about what I assume when I am supervising research.  Interestingly, one of the people whom I notified about the mistake has suggested that we collaborate on a project.  I take this to mean that my handling of the mistake has actually earned me more respect, not less.  For that, and for my wise mentor, I am grateful.  It’s times like this when you really understand what’s important.

By Thoreau

Via Greenwald, I learn that Andrew Sullivan is glad to be of service:

I also didn’t realize the impact that the Newsweek Obama cover-story had on Obama donors and staffers until last night. So many people mentioned it.

Greenwald had the perfect response:

It’s important and gratifying as a journalist to know that you’ve given something of great value to a politician’s donors and staffers; isn’t that why the Founders insisted on a free press?

Indeed.

Greenwald also notes that the man accused of a war crime in Afghanistan has gotten better treatment than the man accused of blowing the whistle on war crimes.  When asked for comment, Eric Holder simply said “Never take sides against the family, Fredo.”

By Thoreau

Regarding the cold-blooded murder of 16 human beings in Afghanistan, the LA Times writes:

In American minds, the moral distinction between the accidental and the deliberate, between the carefully judged risk and the deranged act, is incalculable. But for Afghans, the result — the shrouded bodies, the wailing relatives, the bite of shovels into dusty ground — speaks to the numbing sameness of unexpected and violent death.

The print edition has something even worse below the title:

The massacre of 16 villagers reveals a cultural rift, as Americans see a singularly abhorrent act while Afghans see a years-long pattern.

This implies that there’s something exotic in their upbringings and their way of thinking about the world that makes them fundamentally different from Americans.  In that view, we could understand them better if we just went to some sort of cultural diversity seminar led by a prissy liberal who shows us head scarves and Afghan food and explains that Islam is based on the Koran and Pashtuns have very strong family ties.  After getting lectured on “You have to understand, from their perspective…” we could pat ourselves on the back for being more enlightened.

Except that we don’t need a diversity seminar to understand the problem here.  There’s nothing unique to Afghan culture that makes them respond to events in a special way that we Westerners cannot understand.  It’s really quite simple:  The Afghans are human beings who have the misfortune of living in a place that has been at war for decades.  For the past decade they have been under the occupation of a foreign power that has inflicted substantial civilian casualties on an ongoing basis while imprisoning and torturing a lot of people who turned out to be innocent after being handed over by warlords looking for bounties.  They are thoroughly pissed that off their friends and loved ones keep getting killed by US forces.  So it’s no great mystery why they see a years-long pattern.

Now, there’s an obvious US response, something about how those other civilian casualties were not intended, or were calculated to be “worth it” to get a high-value target, while this guy just did it for fun.  Sure, there’s a difference, but is it enough of a difference to matter to them?  All you need to do is pay attention to what’s happening, information that often isn’t belabored in our news but is well-known to those who pay attention, and you’ll see what’s going on there.  To understand their perspective, to understand why they don’t make much of a distinction between civilian casualties that were inflicted knowingly and civilian casualties that were inflicted maliciously, you just need some basic decency and maybe some lessons learned in Sunday School.  No special knowledge of Afghanistan’s unique culture is necessary.

By Thoreau

New evidence provides a very strong refutation of the superluminal neutrino report. Of course, it is possible that the new experiment is wrong, but between the identified snafus in the OPERA report and the extreme implausibility of the result, I know which one I’m betting on.