Ethics, Technology, and Society

As you read this blog post your eyes are capturing the light emitting from your computer screen. Your brain is busy taking in the information and processing it into meaningful sentences and ideas. You’re hard at work (don’t feel guilty if this is on company time).

You and your computer are “turned on” so to speak, but only one of you is doing anything worthwhile. Other than emitting some light, your computer probably isn’t doing much else. Most unfortunately, it isn’t making the world a better place.

Enter BOINC, an ingenious project designed to enlist your computer into doing something good, when it would otherwise be idling.

BOINC is an offshoot of the seti@home project, out of the University of California at Berkeley. (SPOILER ALERT!: This blog posting is NOT about aliens, so please keep reading.) SETI (the Search for Extraterrestrial Intelligence) is a scientific project whose goal is to comb the skies searching for extraterrestrial life in the universe. The search involves employing massive radio telescopes designed to listen to weak radio signals originating in outer space. The radio signals are recorded and analyzed to determine if they contain the markers of intelligent life. In the case of radio signals any non-random pattern is a potential mark of intelligence.

The SETI project was made famous in the 1997 film adaptation of Carl Sagan’s science fiction book Contact, a story that takes place following a successful detection of extraterrestrial intelligent life. In the film SETI scientist Jodie Foster detects a signal while listening to the telescope’s input through a pair of headphones. She hears a loud pulsing noise and immediately realizes the significance of the event.

In the real world, however, you need more than a pair of Beats by Dr. Dre to detect alien life. To start with, there’s a huge band of frequencies that you can tune in to to see if anything interesting is there. Compounding the problem, the intelligent signals may be very weak and are always mixed in with all sorts of other noise on the same frequency. This means that detecting an intelligent signal will likely involve performing a sophisticated analysis on the data collected by SETI. But the analysis takes computing power. Lots and lots of expensive computing power.

To overcome the analysis problem scientists and computer programmers at Berkeley came up with a brilliant idea. In 1999 they launched seti@home, a small program that anyone on the internet could download onto their computer to help with the analysis. The Berkeley team recognized that most people’s computers sit idle the vast majority of the time that they are turned on. Just think of how many times you walk by a computer and see a screensaver running, or of the amount of time you spend staring at the screen working on your emails or other documents. The seti@home team figured if they could harness a small portion of your idle time and put it towards analyzing their radio data, they could analyze massive amounts of data, with much greater accuracy at a relatively low cost.

Seti@home worked by downloading small chunks of data that could easily be analyzed on a typical home or office computer. Users could specify when the program was allowed to run in the background and whether or not it was allowed automatically connect to the server (those were the dial-up days). It even came with a very sci-fi looking screensaver and graphical interface displaying various graphs and charts meant to give the user the idea that something interesting was going on.

If you were interested in the search for extraterrestrial intelligence (the old joke being that there’s little of it here on Earth) then there certainly was something interesting going on at the seti@home project. But beyond the ability to work the data in search of aliens, seti@home opened up a new world of powerful computing possibilities. By convincing several hundred thousands of participants to donate their computers’ spare processor cycles the seti@home team had created the first large-scale distributed volunteer computing grid.

This week China unveiled their new supercomputer, aimed at showcasing the country’s technological prowess. They’re expecting it to clock in at just over a petaflop–that’s a million billion calculations per second–putting it among the world’s top 15 fastest computers. Those machines are very expensive, and out of reach of most researchers. Seti@home, on the other hand, currently boasts an average of 748 teraflops per second–that’s 0.748 petaflops–and the speed is limited primarily by the number of participants who choose to download the small program. In other words, grid computing rivals the worlds fastest supercomputers in terms of raw processing power.

Now I can get back to BOINC. Seti@home’s wild success demonstrated the potential of grid computing. Researchers wanted to open the grid up to anyone with a computationally complex problem to solve. In true academic spirit, the very same team that developed the seti@home application decided to develop the Berkeley Open Infrastructure for Network Computing, or BOINC, to allow other projects to compete with them directly for computing time. That move has great ethical implications.

The ethics of BOINC should not be undersold. BOINC allows its users to “attach” to one or more grid computing projects at a time while managing them all from a single interface. Very quickly seti@home became only one of several projects to which users could donate their computing power. The other projects on the list have clear ethical dimensions where SETI might not, including projects whose goals include developing complex models for climate prediction, epidemiology, drug research, and many dedicated to solving problems in basic science. Users can effectively contribute to solving some of the most pressing global problems without having to contribute any additional money at all.

Motivating people to participate is the single biggest roadblock, and one that is common to other ethical dilemmas involving individual contributions to so-called “distant” problems, such as giving to international aid projects. Peter Singer has famously argued that we all have an ethical duty to give a portion of our income to saving lives in far-off countries. The idea (and it is a convincing one) is that a small sacrifice that has little impact on us AND saves lives is a sacrifice we are required to make. Despite the argument donations are relatively hard to come by.

As with the cash donations to international aid, participation in BOINC related projects is the key to success. Yet participation is relatively low, roughly 587,000 active computers (a mere 327,000 active users) at last count.

BOINC is a great project for a couple of reasons. First, participation in the many BOINC projects costs relatively little (a few cents of energy per month). Second, despite the low cost the impact is relatively large, as developing the models is impossible without massive computational output (the current BOINC participation makes it the most powerful supercomputer at 2+ petaflops per second!).

The tiny individual cost of participation makes it very difficult for you (or I) to come up with good reasons why not to participate in BOINC. If you can’t think of a good reason why not to participate, and if participation has obvious benefits, then you ought to participate. Seems straightforward, right? I hope to convince you to do just that.

That’s where this blog comes in. Over the next few weeks I will be blogging about BOINC.

First I am going to take the time to describe how to install and configure BOINC so that you will likely never notice it is running on your machine. I will also answer any questions about the installation and configuration process through comments on this blog. After setting you up I’ll be profiling some of the projects that are of particular ethical interest, in order to give you a reason to attach to them.

I truly hope that some of you will take me up on this. Keep in mind that if you choose to participate, you AND your computer will actually be making the world a better place.

(+1 rating, 1 votes)

 Loading ...