Bottom-up

Via Techdirt, David Levine spots an interesting breakdown of Metallica’s revenues:

Along with touring revenue — the band pulled in $22.8 million from 55 arena shows reported to Boxscore that drew more than 968,000 fans — Metallica sold 694,000 albums in 2009. The majority of those sales came from its Rick Rubin-produced 2008 release, “Death Magnetic” (297,000). Album sales revenue totaled $1.6 million. And most of Metallica’s track download earnings came from its 1991 hit “Enter Sandman,” which sold 450,000.

This underscores a point I made last month: For many bands, album sales are already a trivial fraction of their revenues. A world in which bands give away music to sell more concert tickets and merchandise would be terrible for record labels, but it wouldn’t be a big change for a lot of bands. And this model is especially attractive for up-and-coming musicians who have the most to gain from using free music as a promotional vehicle.

Eric Raymond predicts that smartphones will disrupt the traditional PC market:

Here’s what I think my computing experience is going to look like, oh, about 2014:

All my software development projects and personal papers live on the same device I make my phone calls from. It looks a lot like the G1 now sitting on the desk inches from my left hand; a handful of buttons, a small flatscreen, and a cable/charger port. My desk has three other things on it: a keyboard about the size of the one I have now, a display larger than the one I have now, and an optical drive. Wires from all three run to a small cradle base in which my phone sits; this also doubles as a USB hub, and has an Ethernet cable running to my house network. And that’s my computer…

When I leave the house, I pull the phone from its cradle and put it in my pocket. At that point, the onboard screen becomes its display. I’m limited to low resolution and a soft keyboard through the phone’s touchscreen…until I get to my local internet cafe, which is full of display-keyboard combinations much like the one I have at home, awaiting my use. If for some reason I need an optical drive, I borrow one and plug it into the device hub that’s servicing my phone.

I think the trend he’s describing is plausible, but the timescale seems wrong. I’m already doing basically what he describes with my laptop. I’ve got one MacBook and two keyboard/display setups–one on my desk at home and one on my desk at school. It’s extremely convenient to have exactly the same computing environment everywhere I go.

However, I don’t think convergence with phones is going to happen quite as quickly as ESR is predicting. My iPhone is an order of magnitude slower and has an order of magnitude less storage than my MacBook. The gap is closing, but it’s happening pretty slowly. You could run a desktop OS on your smartphone, but it would feel like you’d time-traveled back to 2001. I think it’ll be another decade before phones are fast and capacious enough that people will be willing to give up their traditional PCs.

Another possibility is that we’ll see just the opposite: a proliferation of cheap, varied devices, with everything synced to “the cloud.” On this model everyone will have a bunch of $200 devices with a variety of shapes and sizes scattered around—desktops, laptops, tablets, and smartphones—and they’ll be able to access all their stuff from any device because you’ve stored everything on a server run by Google, Microsoft, or Dropbox.

This outcome seems more likely for a couple of reasons. First, there are a lot of different potential form factors for a computing devices. ESR’s model only works well for the two ends of the spectrum. Second, the key economic trend is that these devices are going to keep getting cheaper. People are going to find it convenient to treat them as basically disposable, much as pocket calculators and land-line telephones are treated today. In contrast, ESR’s model suggests that phones will get more and more valuable as it takes on more and more functions. Given how bad most people are about backups, they’re not going to want to put all their files on a device they could lead in the back of a taxicab.

I’m not too enthusiastic about the “cloud” scenario because I like control over my data. But a lot of people basically work this way today, and if I had to bet money, I’d say that’s how most people will be computing a decade in the future.

(Thanks to Chaim for the pointer to the ESR article)

Via Reihan, a trade group called the International Intellectual Property Alliance is apparently lobbying the US Trade Representative to treat preferential adoption of free software as equivalent to copyright infringement. Bobby Johnson at the Guardian explains:

Last year the Indonesian government sent around a circular to all government departments and state-owned businesses, pushing them towards open source. This, says the IIPA, “encourages government agencies to use “FOSS” (Free Open Source Software) with a view toward implementation by the end of 2011, which the Circular states will result in the use of legitimate open source and FOSS software and a reduction in overall costs of software”.

Nothing wrong with that, right? After all, the British government has said it will boost the use of open source software.

But the IIPA suggested that Indonesia deserves Special 301 status because encouraging (not forcing) such takeup “weakens the software industry” and “fails to build respect for intellectual property rights”.

I don’t understand why USTR is in the business of bullying other countries into changing their copyright laws in the first place. But taking that policy as a given, I think Bobbie Johnson’s reaction to this is spot on:

I know open source has a tendency to be linked to socialist ideals, but I also think it’s an example of the free market in action. When companies can’t compete with huge, crushing competitors, they route around it and find another way to reduce costs and compete. Most FOSS isn’t state-owned: it just takes price elasticity to its logical conclusion and uses free as a stick to beat its competitors with (would you ever accuse Google, which gives its main product away for free, of being anti-capitalist?)…

Let’s forget that the statement ignores the fact that there are plenty of businesses built on the OSS model (RedHat, WordPress, Canonical for starters). But beyond that, it seems astonishing to me that anyone should imply that simply recommending open source products – products that can be more easily tailored without infringing licensing rules – “undermines” anything.

Quite right. My only quibble is the suggestion that IIPR are advocates of “free-market capitalism.” There’s nothing free-market about this kind of rent-seeking.

The last time I wrote about the Founders’ Visa concept, I wondered how the law would be drafted to prevent the creation of “sham” startups. Senators Kerry and Lugar have now introduced legislation modeled on Graham’s proposal, and it provides a concrete answer. To qualify for a visa, the startup must raise $250,000 in capital, including $100,000 from an American citizen. And the startup must promise to create 5 jobs, raise another $1 million, or generate $1 million in revenue within the first two years. If the founders fall short of these targets, they get deported.

The first thing to note about this, as Mike Masnick points out, these are really high figures for software startups. Paul Graham, the guy who proposed the founders visa concept in the first place, funds startups with around $20,000 in startup capital. And he has written at length about ramen profitability, the point at which the startup is earning enough money that the founders can pay their basic living expenses. The ramen profitability point is far below $1 million per year. Most startups fail, and a lot of the ones that succeed do so by changing their business strategies several times before they hit paydirt. Getting to ramen profitability within 2 years is fairly difficult. Getting to a million dollars within 2 years is extremely hard.

No genuine angel investor is going to invest $250,000 in a company whose founders could be deported at the end of two years. So the practical effect will be to simply ratchet up the capitalization requirements even higher. Angels will only provide the initial round of funding if they’re prepared to chip in another million bucks at the end of the 2 years. And this, of course, is even more ludicrously out of step with software startups. There are lots and lots of successful software startups who have gotten through their first two years with a lot less capital than $1.25 million.

I don’t know as much about startups in other industries. Maybe there are some where $250,000 is a low threshold. But in any event, this kind of micro-managing is going to significantly limit the usefulness of the visa.

The more fundamental problem with the legislation, I think, is the extent to which it panders to the economic ignorance and xenophobia of American voters. The implicit premise of the proposal seems to be that there are hordes of scary foreigners who want to invade our country and steal our jobs, and it’s vital that we keep them out. However, there are a small number of foreigners who will be so fantastically productive that it’s in our self-interest to hold our collective noses and let them into the country. However, we have to set tight limits so that the bad kind of foreigner (which is almost all of them) don’t use it as a loophole to sneak in and steal our jobs.

I find this kind of bigotry so distasteful that I’m reluctant to indulge it even for a good cause. I’m troubled by the restrictionist elements of my colleague Dan Griswold’s proposal for a guest worker program (which in watered-down form became a key part of the 2007 immigration bill), but I’d vote for it because it would give hundreds of thousands of poor people the opportunity to dramatically improve their standard of living. In the case of the startup visa, we’re talking about a proposal that would help a much smaller group of far more privileged people while pandering much more aggressively to xenophobes.

More to the point, I’m not convinced this strategy will work. The 2007 immigration bill failed because it turns out that xenophobia isn’t a well-organized interest group with which you can negotiate. Proposing ever-narrower visas with ever-more-draconian provisions to “protect American jobs” simply legitimates the Lou Dobbs view of the world while it dissipates whatever political capital advocates of immigration reform have.

This is why I much prefer a bill like the DREAM Act. It, too, is focused on a fairly narrow problem, but it’s narrow in a way that strengthens, rather than undermining, the argument for broader immigration reform. The fundamental argument for the DREAM Act is that immigrants are human beings who have the same rights and deserve the same opportunities as native-born Americans. It powerfully advances this argument by focusing on sympathetic victims of the current immigration regime. The implicit argument for the startup visa is just the opposite: that immigrants are second-class human beings who should only be allowed to pursue the American Dream only if they produce large benefits for native-born Americans in the process. I find that premise both economically confused and morally offensive.

If it was hard for senior American officials in Vietnam to get a straight story about how the war was going, it was much harder for senior officials in Washington to do so. Not only were they several thousands miles away, but the military hierarchy actively blocked them from talking to people who would dispute the military’s official story. The John Paul Vann story continued after he returned stateside. Here’s Halberstam’s take:

Vann went home a very angry man, to find that Saigon had ordered that he not be debriefed in Washington. So he began to give his briefing to friends at the Pentagon. It was a professional presentation indeed, and very different from the usual briefings which were coming in from Saigon. What made it striking was that it was not just impressionistic, it seemed to be based on very hard facts. Vann began to get higher and higher hearings in the Pentagon until finally General Barksdale Hamlett, the Deputy Chief of Staff of the Army, heard the briefing, was impressed and arranged for Vann to meet with the Joint Chiefs. Vann was warned by several high officers that above all he must not appear to be critical of General Harkins, who was the personal choice of Maxwell Taylor (by this time Chairman of the Joint Chiefs), since Taylor seemed to be particularly sensitive and protective of Harkins and his reporting…

The Vann briefing was set for 2 PM on July 8, 1963. At 9:45 he sent a copy to General Krulak’s office. A little later Vann, eager, starched, finally getting his hearing, showed up outside the office of General Earle G. Wheeler, the Chief of Staff, to be on hand in case there were any new developments. He was sitting there when a phone call came in to one of Wheeler’s aides. “Who wants the item removed from the agenda?” the aide asked. The voice at the other end spoke for a few minutes. “Is this the Secretary of Defense’s or the Chairman’s office?” There was more talk. “Is that an order or a request?” Then more talk. “Let me get this right. The Chairman requests that the item be removed.” The aide turned to Vann. “Looks like you don’t brief today, buddy.” He went to Wheeler’s office, returned in a minute, picked up the phone, and dialed a number and said “The Chief agrees to remove the item from the agenda.”

The key thing to note here is that the president and Secretary of Defense were almost certainly oblivious to this little power struggle. Moreover, it’s quite possible that Taylor himself was oblivious to the dispute and this was the work of an aide shielding his boss from awkward criticism. Even if Taylor was responsible for spiking Vann’s presentation, it’s safe to say that there were other, less prominent reports from the field that never made it to Taylor’s level, which would have given Taylor the impression that Vann was a huge outlier.

This is one of the major factors that limit the president’s control over the operations of the military. It’s in the nature of the office for the president to only interact with the top couple of layers of the military bureaucracy. This means that virtually everyone he talks to will have gotten a heavily filtered and sanitized version of events. In this case, the reporting hierarchy made the senior leadership too optimistic, but information can be filtered in lots of other ways as well. Any given unit is going to preferentially report information that serves that unit’s interests and suppress information that does not serve those interests. This means that the information that flows up the reporting hierarchy will be strongly biased in favor of the agendas of the people doing the reporting at each level. If the bulk of the military shares some bias—and in a conformist institution like the military they usually will—the information that flows to the top will strongly reflect that bias.

To be clear, this isn’t because people are consciously trying to mislead their superiors. As I’ve mentioned before, officers in a hierarchy inevitably have to abbreviate and summarize the information they report, and it’s just inevitable that their own biases will inevitably sneak into those reports. So if you think the war is going well (or think your superior wants to hear that it is), you’re going to naturally make your report a little more optimistic than is otherwise warranted. Similarly, if you want to take some action (say, invading a particular Middle Eastern country), you’re going to be somewhat more likely to report information that supports an invasion, and somewhat less likely to report information that would support the case against it.

In the aggregate, this means that the military naturally provides the president with a lot of information supporting decisions the military brass support, and much less information that would support contrary decisions. A president that wants to follow his generals’ advice will always have evidence at his fingertips to defend his decision to Congress and to the general public. In contrast, a president who overrules the military brass will find that his opponents in Congress will have such ammunition at their fingertips, fed to them by top military officials. This doesn’t force presidents to do what their generals want, but it means that all the forces of inertia are on the military’s side. Fighting the military bureaucracy is an exhausting process—even when they technically report to you—because your non-military staff has to do much of the intellectual legwork that would otherwise be done by the military hierarchy itself.

Not everyone in the US military was surprised by the American failure in Vietnam. Two people who clearly saw what was coming were Lieutenant Colonel John Paul Vann and Colonel Dan Porter, officers who were stationed in the Mekong Delta in the early 1960s. Halberstam tells the story:

Vann’s reporting had caused some problems in the past. Now a major storm would center around him in January 1963 when the division he advised was badly defeated and performed with great cowardice at the battle of Ap Bac, which, being close to Saigon, was well covered journalistically. Harkins was furious, not at the Vietnamese or their commander, but at Vann for having called it a defeat and for having talked with American reporters. Harkins planned to fire Vann at the time but was talked out of it by staff members who argued that firing him would bring even more adverse publicity; they also warned that advisory morale was low enough as it was. Instead, Harkins upbraided Vann, and Vann became a nonperson. Anything he wrote or said thereafter was simply disregarded, and important visitors to the country were steered away from his area.

Porter, Vann’s immediate superior, was next. Before he went home after two long years, he had to turn in a final report, and it was brutally frank. Aides suggested that Porter sweeten it by putting in a few positive notes, but he refused. He was angry and bitter over the way his subordinates were being treated, and after consulting with Ladd, Vann, and Wilson, he handed in the most pessimistic report on the war so far, on the nature of the peasant, the enemy, and the ally. Harkins went into a rage over it; normally final reports of senior advisers were circulated for all the top advisers, but Harkins had Porter’s report collected. He told other officers that it would be sanitized and that if it contained anything of interest, he might make it available. It was never seen again, which did not surprise Porter, but enough was enough, he was leaving the Army.

While it’s hard to justify Harkins’ behavior, it’s not actually that hard to explain. Harkins faced the same pressures to submit rosy reports to his superiors that his subordinates faced to report good news to him. Moreover, it’s probable that Harkins sincerely believed that the optimistic reports he was getting from his other subordinates were correct, and that Porter was being a sourpuss—and jeopardizing the war effort in the process.

Incidents like this, repeated across an organization and over time, will powerfully shape the character of a large, hierarchical institution like the US military. People who are willing to toe the party line and sugar-coat bad news get promoted. People who refuse to do so become unpersons or leave the military entirely. In this environment, it would be surprising if the upper echelons of the military contained anything other than “yes men.”

And for this reason it’s utterly unsurprising that Bunker was so oblivious four year later. Bunker’s optimistic views on the war presumably came from conversations with the military brass, who would have been every bit as over-confident as their predecessors. Bunker probably took them at their word, but even if Bunker had taken Halberstam’s argument to heart and wanted to get the real story out of the military hierarchy, it would have been hard to do so. The more junior officers would have faced strong incentives to both give optimistic reports in general and not to speak outside of the chain of command. Even if Bunker had managed to find junior officers willing to give him a frank assessment, it would have been hard for him to put those individual reports into context—to know whether he just happened to find an officer whose region of the country was unusually bad. In short, it would have taken a great deal of effort on Bunker’s part to penetrate the layers of bureaucracy far enough to get a straight story about how the war was actually going.

Recently Steve Jobs issued a Fatwa against “overtly sexual content” on the iPhone. Some Apple employee apparently decided that selling bathing suits to women is an “overtly sexual” activity, and blocked the app.

I had an argument with Adam Thierer earlier this week about how outraged we should get over this policy. Adam made the entirely valid point that it’s Apple’s app store and they’re entitled to carry whatever apps they want. This is true enough. But likewise this is my blog, and I’m entitled to write unflattering things about them in response.

But I think the key thing to focus on isn’t the abstract question of whether porn on iPhones is good or bad. The key thing to recognize is how fundamentally broken the process itself is. “Overtly sexual content” is a concept that seems clear in the abstract but gets leaky once you have to actually classify tens of thousands of applications. Apple is going to make mistakes, and when they do hapless developers are going to find their apps blocked, often with little explanation or recourse. Also, Apple is going to change its mind periodically, and when they do the affected developers are going to find their hard-earned apps rendered worthless overnight. This is no way to run a technology platform. It’s unfair to developers and it doesn’t scale. And this is precisely why it would be better for everyone if Apple could come up with an application distribution scheme that didn’t require so much central planning.

The author of The Best and the Brightest, David Halberstam, was a young reporter who had his first stint in Vietnam in the early 1960s, at a time when the US commitment was relatively small and the prospects for American victory seemed promising, at least superficially. Halberstam returned to Vietnam in 1967, by which time the American troop commitment had ballooned and it was becoming clear that we’d gotten ourselves into a long, bloody, and possibly unwinnable war. He was invited to dinner at the home of Ellsworth Bunker, the American ambassador to Vietnam. By this point Halberstam was convinced the war was hopeless, and he made his case to Ambassador Bunker:

I suggested that we were fighting the birthrate of the nation, that the war was essentially a stalemate—but a stalemate which favored the other side, since eventually we would have to go home… [Bunker] listened politely to what I said. He had, Bunker said, spoken with his generals—he named several of them—all fine men, and they had assured him that, contrary to what I said, everything was on schedule and that there was an inevitability to the victory we sought, given the awesome force we had mounted against the North Vietnamese army and the Vietcong.

Obviously Bunker was wrong. And there’s a temptation to be smug about this. A lot of people reading a story like this would be inclined to say that Bunker and other senior officials connected with the Vietnam war effort were arrogant, blinkered, and surrounded themselves with “yes men.” And of course there’s something to this—on the next page Halberstam cites the example of a general who “had always gotten things right because when he went into the countryside he unpinned his stars.” The best leaders understand the dangers of subordinates telling them what they want to hear and take these kinds of steps to stay connected to the “facts on the ground.”

But this isn’t the whole story. If it were, stories of oblivious managers and bureaucratic catastrophes wouldn’t be so depressingly common. Presidents, generals, and CEOs have powerful incentives to avoid this trap, and yet they fall into it over and over again. There’s something deeper going on here.

One of the most fundamental problems is that the filtering process of hierarchical organizations works not only on ideas but on people as well. I’ve already noted how bureaucratic reporting processes tend to filter out contrarian views. Even more important is the tendency of bureaucratic organizations to simultaneously filter out contrarian people. I’ll look at a classic example from the Vietnam era in my next post.

Incidentally, if you doubt that remix culture can produce works of lasting artistic value, you’ve probably never seen “Ten Things I Hate about Commandments.” It never gets old:

Following up on his video on the evolution of remix culture, Julian contrasts top-down and bottom-up visions of culture:

Current intellectual property law frowns on “copying” as opposed to mere “influence.” If I write and record a song that is manifestly influenced by the sound of the Beatles, that’s just how culture works; if I remix or reperform a medley of their songs, that’s infringing. One way to think about the distinction is to ask how much mutation of the original work has occurred in my head before I send it out into the world. We can imagine my sitting with a guitar playing “Taxman,” beginning by improvising new lyrics, and gradually altering the melody until I’ve produced a song that is sufficiently transformed to count as an original work, though perhaps still a recognizably Beatlesesque one. I’m free and clear under copyright law just so long as I only record and distribute the final product, which consists of enough of my own contribution that it no longer counts as a “copy.”

Implicit in this model is the premise that creativity is fundamentally an individual enterprise—an act of intelligent design. Yet so much of our culture, historically, has not been produced in this way, but by a collective process of mutation and evolution, by the selection of many small tweaks that (whether by chance or owing to some stroke of insight) improve the work, at least in the eyes of the next person to take it up. Perhaps ironically, this is the kind of evolutionary process by which myths evolve—myths of life breathed into mud, or of Athena springing full-grown from the head of Zeus. Our legal system now takes these evolved myths as its paradigm of creation.