Institute For The Future

On Anticipatory Quarantines

 

A leading researcher at Microsoft last week suggested that computers that have been infected with viruses should be, in effect, quarantined from accessing the Internet to avoid spreading potentially dangerous mischief into the broader network. The researcher suggested that safe computers could be issued a "health certificate," once virus-free, indicating that the computer is free, once again, to roam the world.

 

I don't have the technical knowledge to comment on the wisdom of the proposal, but I was struck by the language. The phrase "health certificate" and the concept of a quarantine sounds, well, a bit medical, for something like a computer network. But I think what it's indicative of is a broader effort to rethink questions involving quarantines, anticipating risk and the spread of problems.

 

We're seeing this in health, of course. For example, this recent Fast Company article points toward a new DARPA effort to literally anticipate how a virus might evolve in order to begin to constrain it even before it occurs.

 

 

 

DARPA's worry is that the current state of the art in "antiviral agents and vaccines" involves a retrospective stance, protecting against "viruses that are already endemic, virulent, and medically significant to human or animal health" and that "traditional medicine" looks backwards to well-characterized reference viruses, which may not necessarily represent currently problematic strains that are impacting people. Despite innovations like using cell phones to predict where disease outbreaks may occur, vaccines can take months or even years (if you're talking about small molecule inhibitors) to successfully combat newly emerging viruses. This is demonstrably true, and it's a problem if there's going to be a pandemic (can you remember the fuss about H1N1 and how nobody had enough vaccine?) but in terms of everyday life, it's hardly an inconvenience.

 

The new program is designed to stimulate the development of systems that can totally reverse this issue--and result in accurate predictions about the mutational course a virus may follow as it proceeds through a population. The upshot of this could be that the medical and pharmaceutical industry switch from a reactive mode to a proactive one, where they predict and prepare for potential outbreaks in advance. This could result in far more cost-efficient use of resources as well as speedier response times when an outbreak occurs, as well as reduced durations. It's more than just clever computing of genetic changes or lab-based in-vitro studies, though, the point of Prophecy is to work out if environmental inputs could cause mutations too.

 

 

 

 

Other examples in health include renewed efforts to forecast malaria outbreak, as well as efforts to use social status to anticipate how viruses distribute through social networks.

 

In a more general sense, what I think these questions point to is a very difficult question about how to control risk in a world where problems can spread faster than we can possibly react effectively. In this context, we can't simply think about how to react effectively but have to get better at the much fuzzier challenge of anticipating a problem before it exists.

 

The reaction to the "flash crash" in the stock market earlier this year was just one example of this phenomenon; a poorly written algorithm resulted in the stock market losing 10 percent of its value in 10 minutes. Efforts are underway, now, to figure out what the best strategies that stock exchanges could employ to prevent that level of chaos in the future.

 

But the question also reminds me of a startup company called Recorded Future, which is monitoring Twitter, searches and other sources of real-time information to gain a sense of what might be in store for us in the near future. Among other things, Recorded Future suggests:

 

 

In time, this may lead to the development of apps targeted at consumers, says Ahlberg. "If I'm about to buy an iPhone, I might want to know if I am going to look stupid because they'll launch a new one next week, or how long it usually takes for competitors to launch competing products after a new Apple launch."

 

 

Now, in some sense, this is silly. In the grand scheme of things, it doesn't really matter if someone has a slightly outdated Apple product (though in may matter to the device owner.)

 

But I think it points to the challenge here. Information moves incredibly quickly, and a sudden blip can lead to major panic and a major problem, or it can just be something trivial, like a rumor about Apple. As a result, we don't simply need the ability to anticipate and act on that information, we need the ability to recognize what matters in the first place.

 

The other point is that even while companies like Recorded Future, and projects like the DARPA project I cited above, may be premised on the concept of predicting the near future, ultimately, these are forecasts--best efforts to figure out--what might be happening. If the forecast about Apple's latest move is off, it doesn't really matter. But if you quarantine millions of people who turn out to be fine, or deny millions of people access to the Internet for fear that their computers could spread a virus, you're creating much thornier questions.

 

In the next decade, we'll be facing a lot of questions about how we want to use continuous streams of information to shape our actions in the near future. It's easy to suggest that we should be cautious. The real dilemma will steam from unpacking the previous statement--should we be cautious about protecting the rights of the individually quarantined, or should we be cautious and protect the rights of the whole?