LIFT ‘08: Francois Grey and the rise of “volunteer thinking”
February 13th, 2008Note: these are running notes for LIFT ‘08. For more complete blog coverage, check out the official LIFT Conference blog or LunchOverIP.
François Grey is the Head of IT Communications at CERN, the Switzerland-based research center most widely known as the birthplace of the web and home to the world’s largest scientific instrument, the Large Hadron Collider (LHC). In addition to smashing participles and looking for new ones, when the LHC begins operations later this year, it will produce roughly 15 Petabytes (15 million gigabytes) of data annually. To analyze the mountains of the data the project will rely on the LHC Grid, a globally distributed network of computers. The approach is similar to that used by SETI @ Home which taps into the spare computing power of millions of volunteer computers to download and analyze radio telescope data in search of extraterrestrial life. The idea is catching on: Grey says today there are dozens of distributed computing projects in fields such as molecular biology, climatology and particle physics; the Sony PS3 is pre-loaded with software to enable it to connect with Folding@Home, aStanford-based project that’s studying protein folding.
A unique culture has evolved around distributed computing projects, creating tight-knit communities of volunteers. To incentivize participation projects like SETI give volunteers “credits” for donating computing time; individuals and teams self-organize and compete to donate the most hours. Message boards act as a social networking site for volunteers and individual profiles make participation in the project personal: Matthew Tine (aka Stainmaster) is a 32-year old Australian who has been a SETI@home member since 2000 and has to date donated over 18,000 hours of CPU time.
Grey says is science taking the distributed network concept one step further, from volunteer computing to “volunteer thinking”. As he describes it: a new generation of projects is emerging that taps into not only the computers, but the brains of the volunteers, inviting them to analyze scientific data online: cataloguing galaxies, scouring microscope images, or mapping out remote regions. For instance, Galaxy Zoo taps armchair astronomers to help scientists classify galaxies by looking at images and identifying whether it’s a spiral or an elliptical galaxy. Using legions of people is not only efficient, it’s also more effective: collectively human brains are better than computers at pattern recognition, and algorithms would invariably throw out the “unusual, weird or wonderful” patterns that would attract the attention of a curious human.
Herbaria at Home is using a similar crowd-sourcing approach digitize and document the world’s largest collections of herbarium specimens; another project, AfricaMap seeks to create more accurate cartographic maps of Africa by asking volunteers to look at satellite images over the African continent and search for roads, bridges, human settlements, rivers, agriculture fields, etc.
Distributed computing has spawned distributed thinking, a trend Grey calls “citizen cyberscience”, or philanthropic crowdsourcing. It’s a fascinating concept and one that could mobilize an army of amateur scientists into service.
Categories:
Posted by Michele Bowman
