Freedom to Tinker

Tim Lee at Tech Liberation Front points out an interesting aspect of the new MovieBeam device – it offers its highest-resolution output only to video displays that use the HDMI format.

(MovieBeam is a $200 box you buy that lets you buy 24-hour access to recent movies. There is a rotating menu of movies. Currently video content is trickled out to MovieBeam boxes via unused broadcast bandwidth rented from PBS stations. Eventually they’ll use the Internet to distribute movies to the devices.)

This is a common tactic these days – transmitting the highest-res content only via HDMI. And it seems like a mistake for Hollywood to insist on this. The biggest problem is that some HDTVs have HDMI inputs and some don’t, and most consumers don’t know the difference. Do you know whether your TV has an HDMI input? If you do, you either (a) don’t have a high-def TV, or (b) are a serious video geek.

Consider a (hypothetical) consumer, Fred, who bought an early high-def set because he wanted to watch movies. Fred buys MovieBeam, or a next-gen DVD player, only to discover that his TV can’t display the movies he wants in full definition, because his TV doesn’t do HDMI.

Fred will be especially angry to learn that his MovieBeam box or high-def DVD player is perfectly capable of sending content at higher definition to the inputs that his TV does have, but because of a bunch of legal mumbo-jumbo that Hollywood insists upon, his set-top box deliberately down-rezzes the video before sending it to his TV. Just imagine what Fred will think when he sees news stories about how pirated content is available in portable, high-def formats that will work with his TV.

The official story is that HDMI is a security measure, designed to stop infringers. It’s been known for years that HDMI has serious security flaws; even Wikipedia discusses them. HDMI’s security woes make a pretty interesting story, which I’ll explore over several posts. First I’ll talk about what HDMI is trying to do. Then I’ll go under the hood and talk about how the critical part of HDMI works and its well-known security flaws. (This part is already in the academic literature; I’ll give a more accessible description.) Finally, I’ll get to what is probably the most interesting part: what the history of HDMI security tells us about the industry’s goals and practices.

Officially, the security portion of HDMI is known as High-bandwidth Digital Content Protection, or HDCP. The core of this security design is the HDCP handshake, which takes place whenever two devices communicate over an HDMI cable. The handshake has two goals. First, it lets each device confirm that the other device is an authorized HDCP device. Second, it lets the two devices agree on a secret encryption key which only those two devices know. Subsequent communication over the cable is encrypted using that key, so that eavesdroppers can’t get their hands on any content that is distributed.

In theory, this is supposed to stop would-be infringers. If an infringer tries to plug an authorized video source (like a MovieBeam box) into a device that can capture and redistribute video content, this won’t work, because the capture device won’t be able to do the handshake – the authorized video source will recognize that it is unauthorized and so will refuse to sent it content. Alternatively, if an infringer tries to capture content off the wire, between an authorized source and an authorized TV set, this will be foiled by encryption. That’s the theory at least. The practice is quite different, as I’ll describe next time.

Recently I’ve been trying to figure out the politics of technology policy. There seem to be regularly drawn battle lines in Congress, but for the most part tech policy doesn’t play out as a Republican vs. Democratic or liberal vs. conservative conflict.

Henry Farrell, in a recent post at Crooked Timber, put his finger on one important factor. This was part of a larger online seminar on Chris Mooney’s book “The Republican War on Science” (which I won’t discuss here). Here’s the core of Henry Farrell’s observation:

There’s a strand of Republican thinking – represented most prominently by Newt Gingrich and by various Republican-affiliated techno-libertarians – that has a much more complicated attitude to science. Chris [Mooney] more or less admits in the book that he doesn’t get Newt, who on the one hand helped gut OTA [the Office of Technology Assessment] (or at the very least stood passively to one side as it was gutted) but on the other hand has been a proponent of more funding for many areas of the sciences. I want to argue that getting Newt is important.

What drives Newt and people like him? Why are they so vigorously in favour of some kinds of science, and so opposed to others? The answer lies, I think, in an almost blindly optimistic set of beliefs about technology and its likely consequences when combined with individual freedom. Technology doesn’t equal science of course; this viewpoint is sometimes pro-science, sometimes anti- and sometimes orthogonal to science as it’s usually practiced. Combining some half-baked sociology with some half arsed intellectual history, I want to argue that there is a pervasive strain of libertarian thought (strongly influenced by a certain kind of science fiction) that sees future technological development as likely to empower individuals, and thus as being highly attractive. When science suggests a future of limitless possibilities for individuals, people with this orientation tend to be vigorously in its favour. When, instead, science suggests that there are limits to how technology can be developed, or problems that aren’t readily solved by technological means, people with this orientation tend either to discount it or to be actively hostile to it.

This mindset is especially dicey when applied to technology policy. It’s one thing to believe, as Farrell implies here, that technology can always subdue nature. That view at least reflects a consistent faith in the power of technology. But in tech policy issues, we’re not thinking so much about technology vs. nature, as about technology vs. other technology. And in a technology vs. technology battle, an unshakable faith in technology isn’t much of a guide to action.

Consider Farrell’s example of the Strategic Defense Initiative, the original Reagan-era plan to develop strong defenses against ballistic missile attacks. At the time, belief that SDI would succeed was a pretty good litmus test for this kind of techno-utopianism. Most reputable scientists said at the time that SDI wasn’t feasible, and they turned out to be right. But the killer argument against SDI was that enemies would adapt to SDI technologies by deploying decoys, or countermeasures, or alternatives to ballistic missiles such as suitcase bombs. SDI was an attempt to defeat technology with technology.

The same is true in the copyright wars. Some techno-utopians see technology – especially DRM – as the solution. The MPAA’s rhetoric about DRM often hits this note – Jack Valenti is a master at professing faith in technology as solving the industry’s problems. But DRM tries to defeat technology with technology, so faith in technology doesn’t get you very far. To make good policy, what you really need is to understand the technologies on both sides of the battle, as well as the surrounding technical landscape that lets you predict the future of the technical battle.

The political challenge here is how to defuse the dangerous instincts of the less-informed techno-utopians. How can we preserve their general faith in technology while helping them see why it won’t solve all human problems?