colossus_operators.jpg

The Colossus digital computer, completed in 1943, helped the Allies win WWII.

I’m disappointed in the pace of technological change lately, in particular when you consider that we live in a time of war.

During World War II, some of the most astounding technological developments of all time were developed, both by the Allies and the Axis powers. In some cases existing technologies advanced by leaps and bounds, and in other cases new technologies emerged from practically nothing in less than half a decade. The list of revolutionary innovations that emerged during the war years is staggering:

  • Radar and Sonar
  • Computers and Cryptography
  • Rockets
  • Jet Aircraft
  • Nuclear Reactors
  • Penicillin

Lesser innovations, at least in comparison to these milestones, came as well in fields as diverse as project management, communications, transportation, and medicine. And of course, more soberingly, many military technologies were advanced or fully realized during the war, including aircraft carries, submarines, and atomic bombs.

America is now at war again, against an enemy whose tactics continually confound and frustrate us. And yet, as far as I can tell, our pace of technological innovation over the past six years — longer than all of WWII — has been utterly pathetic.

Backward Priorities

Think of all the areas in which new technologies could be helping us in the so-called “War on Terror”, or in the fight against insurgents in Iraq: Machine translation, bomb and toxin detection, robotics including unmanned aircraft and ground forces, new forms of logistics, management, and troop rotation, new generations of armored vehicles, leaps forward in mobile communications, new medical and psychological treatments for injured soldiers… I’m not a military visionary by a long shot, but even as a layman I can imagine some of the kinds of developments we urgently need and how much better off we could be.

And yet instead of investing in military and technological innovation, our government seems to be spending its energy and money on outdated and ineffective older technologies. Even leaving aside the moral and constitutional questions, our technological strategies are at best stagnant: From anti-missle defense to domestic wiretapping to flat-out torture, our anti-terror strategies seem to have gone backwards in time rather than forwards. The Bush Administration has probably spent more time, energy, and taxpayers’ dollars on lawyers crafting arguments to defend medieval torture techniques than they’ve spent on developing newer and more reliable interrogation technologies.

Maybe it’s all a secret?

Many people will argue that perhaps the government and the Pentagon already are, in fact, developing and using astounding new technologies every day, and that the general public just doesn’t know about it. They will ask: Did the American people know what the physicists were doing with the Manhattan Project? Did the British people know what the mathematicians were up to at Bletchley Park?

Of course, these are the same people who argued that President Bush had mountains of proof about Saddam’s involvement in 9/11, but that because the material was so secret he couldn’t share it even with members of Congress. Or that the evidence against the prisoners at Guantánamo is rock-solid but is so sensitive that nobody even in the American judicial branch can be trusted to see it.

These arguments are preposterous, of course. If Bush had proof for Saddam’s connection to 9/11, he would have released it at all costs. Half of the prisoners at Gitmo, far from being provably guilty, have been freed.

If the Administration was truly committed to increasing America’s technological advantage against our existing and potential enemies, we would see many of the results every day, for example in the security at our airports, toll plazas, banks, and government institutions. We’d see newly-funded research programs at our universities. We’d see new anti-money-laundering technologies in place in the financial sector. We’d see commercial products using technologies developed in the military and finding their way into consumer electronics. And American combat casualties would be reduced.

Let’s face it. There’s no secret weapon or secret plan.

Where we should be:

Many of the technologies that would help us in the War on Terror are exactly the kind of technologies that could easily trickle into the private sector with no harm to our national defense. Alternately, helpful military technologies could emerge, with government support, from our private sector.

Machine translation is a great example of where we have fallen short: In six years, couldn’t we have invested billions of dollars in a program to make a machine that can reliably transcribe spoken Arabic into textual English? Or for that matter, any language into any other? Private sector technologies have slowly been scratching at the surface of this, from commercial speech-to-text software to the many online translation algorithms, such as the legendary Alta Vista Babelfish, that have existed for over a decade. But given that our ability to understand foreign languages has never been more important to national security, shouldn’t there be a Machine Translation Manhattan Project?

Or interrogation: The New York Times had an analysis of the state of America’s interrogation technology a few weeks ago, and it is embarassingly pathetic:

As the Bush administration completes secret new rules governing interrogations, a group of experts advising the intelligence agencies are arguing that the harsh techniques used since the 2001 terrorist attacks are outmoded, amateurish and unreliable.

The psychologists and other specialists, commissioned by the Intelligence Science Board, make the case that more than five years after the Sept. 11 attacks, the Bush administration has yet to create an elite corps of interrogators trained to glean secrets from terrorism suspects.

While billions are spent each year to upgrade satellites and other high-tech spy machinery, the experts say, interrogation methods — possibly the most important source of information on groups like Al Qaeda — are a hodgepodge that date from the 1950s, or are modeled on old Soviet practices.

Meanwhile, this week’s New Yorker includes a survey of current lie-detection technologies, in particular a new technique using MRI scanners to detect deception. The article says that the Pentagon is indeed exploring this technology, but it seems pretty clear that the private sector is leading the charge. Although the tech is still primitive and unreliable, at least for now, it shows some promise. Shouldn’t we have moved from waterboarding to reliable brainscanning by now? Such an advancement in interrogation techniques would not only help us militarily, but also in the court of public opinion: discontinuing physical torture would go a long way towards putting America back in a position of moral advantage.

Any serious presidential contender for 2008 should propose that the USA is technologically a full decade behind where we can and should be in protecting ourselves against today’s new threats. The Administration’s abject neglect, and downright hostility, towards advanced science and technology has gone on long enough. It’s time to bring the eggheads back.


Comments

11 responses to “Technology and War”

  1. Even better: stop creating and feeding these threats, so technology can be use to improve people’s lives, instead of killing them.

  2. @Carlos: I agree in spirit — military technology alone won’t bring world peace — but given the situation of the world we live in, we do need to be able to detect terrorism before it happens, and we need to be able to react better after it happens. Also, most of the potential technologies I’ve listed have nothing to do with killing. Maybe that’s why the Bush Administration doesn’t see any use for them.

  3. Great post. So the historical question is, “Why has this war not produced significant technical innovation, like other major modern conflicts before it?”

    I think the answer to that is that we have very much entered into Orwell’s state of “perpetual war”, which is more of a stasis phenomenon, with an administration that very much has an active hand (whether it’s their intention or not) in perpetuating it.

    Bill Moyers did a great interview with Jon Stewart a while back where they discuss the cognitive disconnect between the administration’s words and their actions, the notion that they are obviously trying to strike a balance between keeping the citizenry just fearful enough to be compliant but not so alarmed that they pay close attention. If you were to ask a Bush official directly I’m sure they’d say they were doing this for everyone’s own good, not for some fiendish Big Brother-ish plot, but that’s precisely their folly: The horrific end result is the same whether it was intended or not. Evil is as evil does.

  4. @Christopher
    Don’t get me wrong, I see what you mean in your post. But I do think that it’s easier to eliminate the causes of terrorism rather than terrorism itself. No matter how many terrorists you spot before blowing up something. They will keep coming, because the root problem that spawns terrorists still exists.

    @David
    I agree with you. War On Terrorism is just an excuse for more state control, just as WMD were an excuse to storm Iraq (and set up permanent military bases in the Middle East?). And it’s the perfect excuse, because Terrorism is not an enemy that can be defeated but a vague concept, thus you can perpetuate this state control ad infinitum.

  5. @Fahey: I don’t want war to be a source of innovation!

    I vote for concentrating developing technology outside of the Pentagon, rather than embracing war as a useful inspiration and motivator. Secrecy and security are counter to our most successful innovation: the internet.

    Your post does not focus on the history of the internet — originally the exclusive, expensive, secretive domain called ARPANet, … but now a platform for incredibly open collaboration and innovation.

    I think that the open development of the network will do a better job at “fighting terror” than any surveillance or interrogation methods — openness is the most elegant anti-terrorism.

    Thank $deity for the pentagon’s investment in it, but it’s time to move on to a new model.

    Communication and community is more inspiring and creative than the top-down, command and conquer world of the pentagon.

  6. 1. Personally I think the use of the word “War” in this phrase is a misnomer: Unlike in WWI and WWII, this war is not mobilizing the entire populace, with a broad-based draft, paper drives, presidents forcing car companies to switch their entire industrial output to tanks, etc. This is a war in which everyday citizens like you and I are being asked to take no notice, and to keep shopping. So lacking any massively disruptive reallocation of economic & technical resources, the fact that we wouldn’t be seeing any major technological breakthroughs shouldn’t be surprising.

    2. I don’t think it works to compare the advances in computers brought about by the first two world wars to brain scans or automated translations. Calculating charts for mortars or making unbreakable codes are the sorts of things that can be automated easily, because the domains are relatively well understood and made mathematical. There’s no sensibly mathematical analogue for brain scanning or automated translation. Much of the low-hanging fruit has already been taken, and now we’re heading into the territory of how the mind works … and despite the best promises of pharmaceutical companies, nobody knows how the mind works, certainly not enough to write an algorithm about it.

    Turing himself argued that automated translation would be impossible, because language and consciousness are so heavily intertwined, and only conscious beings can make good translations.

    Though, to complicate matters even further, there are some highly focused areas that have advanced quickly since September 11th. I’ve been told that facial recognition is moving at a fast clip, for example, since it’s very important for crowd control and validating photo IDs.

    As somebody who makes a living in technology, I still think the solutions are primarily social. My hope for the next President is that he/she works on the more effective social methods of insuring our safety: training more Arabic translators, better human intelligence throughout unstable regions, less broad-based scanning of communications and more calculated surveillance, a lot less arrogance in our matters in the world, and a lot less deaths of brown-skinned people at the hands of our own military. One can hope.

  7. @Francis: To your first point (1), I agree. I think that’s really the undercurrent to my whole peice — that this war is phony and half-baked to begin with (David Sleight’s point, above, too) so naturally we’re falling short of success on every front. From ongoing tax cuts to the “small footprint”, the Administration has underestimated what war really means in every possible way, including the degree of technological investment required.

    I’ve never been totally comfortable with the word “war” to describe what America is doing right now, especially when compared to WWI or II. But America has been in plenty of wars much smaller than this (or these), so the term is accurate even if the scale is smaller than the biggest conflicts. War is war.

    That said, even if we were not at war, we should be investing much more in plain old anti-terror security technology than we are right now.

    Which brings me to your second point (2). Maybe brain scanning is a long ways off, and maybe even machine translation is a pipe dream (more on this in a moment). But how can we say for sure? What if the government were to spend, say, 10 times what they spend now on these fields? What about 100 times what they spend now? Call me ambitious, but I think we could do a lot more than we’re doing now with the right amount of $ and leadership committed to it (see: JFK, Moon Mission).

    With regards to Turing, and with respect, I am sure you are incorrect about his view on machine translation. Turing beleived that consciousness itself was entirely buildable in a machine (he thought that the brain itself was a Turing machine), so why would he think that translation was more difficult than consciousness itself, much less unattainable?

    Facial recognition has come far — but, as with my previous argument, shouldn’t it be much further along? I’ll bet money that some tiny VC-funded web startup with 20 decent programmers is currently doing a better job at facial recognition than the Pentagon is. Once again, think of where we could be with, say, a billion dollars and a thousand of America’s best computer scientists dedicated to this. That’s, what, a week’s worth of what it costs for us to be in Iraq?

    I agree completely with your last paragraph, and with all the other posters who expressed similar sentiments. In no way is my view that we suck at technology intended to suggest either (a) that our wars are justified, or (b) that technology is the best or only solution.

  8. @Chris Blow: I agree in spirit, but ultimately some things cost more money than the private sector is willing to risk. No investor is going to pay $1 billion in R&D money for a translation machine that might not work, or $10 billion for a robot bomb sniffer.

    That said, more innovative solutions can and should emerge from the people. That’s a great idea. How to get people to volunteer their time for a misbegotten war or a deceptive Administration, however, is beyond me at this point.

  9. Yeah, after posting that part on Turing, I realized it didn’t sound right to me either. I think I’m thinking of some other CS luminary, though which one eludes me now.

    As for investing more in brain scans or automated translations: Part of the problem is that major technological breakthoughs tend to be highly non-linear. So if X amount of dollars got you Y results in a extremely difficult domain, it isn’t necessarily a given that you’ll only need 10X dollars to get 10Y results. Maybe you need 100X dollars, or 1000X dollars, who’s to say? (“The Black Swan” touches on this a bit.) These highly concerted efforts tend to be pretty economically exhaustive, so even a big government only gets to do a few of them. Being an engineer for years had made it pretty much impossible to think of solving any meaningful problem without the issue of resource allocation — so the idea of solving automated translation with government money means other problems don’t get solved. Me, I’d rather see a cure for cancer, or even a more universal health care policy, first.

    It’s also not a given that a better understanding of technology, science, or any other technocratic art leads to better tactics or better strategy in any sort of conflict. Nobody proves this better than Robert McNamara in “Fog of War”. He oversaw a great expansion as President of Ford, and excelled at military statistical analysis in World War II, and can still rattle off force multiplier statistics about the Vietcong. What he can’t do, however, is bring himself to admit publically that his policies were a failure. More numbers and more data don’t always make for better decisions. Really great facial recognition technology wouldn’t necessarily lead to less deaths in Iraq. It could plausibly lead to more.

  10. @Francis: You’re thinking of John Searle and his “Chinese Box”.

  11. This also reminds me of an article I read in Wired a few months back called “How NASA Screwed Up (And Four Ways to Fix It)“. Not directly related to war as a driver of innovation, but directly relevant to the notion of sound government outlays driving progress. (Anyone else remember that “What they’re doing up there helps us all down here” ad campaign they did for NASA back in the 80s?)