by Arpan Bhattacharya / February 3, 2017
“Archaeologist Sarah Parcak is allowing anyone in the world with an internet connection to participate in discovering new archaeological sites, and protect vulnerable archeological sites from looting. Using the $1 million she got from winning the TED Prize as well as support from the National Geographic Society and DigitalGlobe, she recently launched GlobalXplorer. The organization seeks to engage people from all backgrounds in finding and preserving archaeological sites through the use of satellite images. The methods in question have, according to GlobalXplorer’s website, already produced impressive results. It notes:
So far, Dr. Parcak’s techniques have helped locate 17 potential pyramids, in addition to 3,100 potential forgotten settlements and 1,000 potential lost tombs in Egypt — and she’s also made significant discoveries in the Viking world and Roman Empire. With the help of citizen scientists across the globe, she hopes to uncover much, much more… So far, Sarah’s methods have proved over 90% successful in producing significant discoveries.
Parcak’s vision uses modern satellite technologies to scan the world for archeologically promising data. With the power of crowds, there are sure to be many fascinating finds at staggeringly fast rates.
Currently, the organization is working primarily in Peru, whose rich historical sites have been plundered by many looters. One way GlobalXplorer prevents looting is by allowing users to note which regions seem vulnerable. Archaeologists can vet the cites that illicit high rates of concern. Then, collaborating with the Peruvian government, the organization is able to help protect those areas. This is a welcome solution to a centuries-old problem. The Guardian noted back in 2011, for example, how conquistadores and contemporaries alike had reduced the grand remains of the Moche civilization of northern Peru to a “lunar landscape” and one of the largest pyramids in pre-Columbus Latin America to “an eroded, plundered shell.”
Some archeologists have criticized Parcak’s vision, arguing that her method of crowdsourcing the analysis of satellite images may actually work to encourage looting – by inadvertently pointing them to where the loot is. However, such concerns are moot, according to Kristina Killgrove’s article on the program for Forbes. She describes the robust system through which GlobalXplorer protects against mistakenly informing looters:
The millions of little satellite tiles are displayed to users randomly. The user cannot pan out, navigate around, or see additional nearby tiles. There are no location references or coordinates shown to the user of the platform either. Rather, the unique tile ID is matched with location information after a GlobalXplorer user tags it. The data are then sent to Parcak and her team for further analysis. In this way, users are collecting data without compromising sensitive information about potential archaeological sites.
Although GlobalXplorer uses the wisdom of crowds, it also conceals information to protect against being exploited.
The approach of allowing virtually anyone to contribute information while simultaneously implementing measures to regulate them is an approach shared by arguably the largest information crowdsourcing project: Wikipedia. Wikipedia famously allows anyone to edit any entry while also subjecting every change to the scrutiny of administrators, who vet arguments submitted by readers both for and against the edits. Thus, like GlobalXplorer, Wikipedia has implemented rigorous procedures to allowing users to contribute whatever they like while also minimizing undesirable outcomes of unchecked freedom.
In addition to GlobalXplorer, there are several other examples of researchers using the wisdom of crowds to make discoveries. Phylo, for example, is a game developed at McGill University. It loosely resembles Tetris and exploits humans’ abilities for pattern-recognition to help decode genetic diseases. And Foldit made headlines after just three weeks of its release in 2011 for revealing insights into the structure of an enzyme related to AIDS.”
by Liz Stinson / 02.16.16
“Technological advancements notwithstanding, space archaeology still requires a lot of human involvement. Algorithms might be able to detect variations in Earth’s surface, but they’re not quite good enough to discern what those patterns mean. For that, you need humans. But humans are slow. Parcak can scan 500 square feet of satellite imagery in about five minutes—but a larger dataset can take her team a week or more to scrutinize. What’s more, not every interesting surface feature heralds a discovery; sometimes a dead zone is just a dead zone. “Any success you see is tottering on the edge of a gigantic mound of shit,” she says. Meaning, like most scientific pursuits, much of her work is a systematized version of trial and error. It’s also a major time investment—but Parcak thinks the payoff could be huge. She estimates that humans have uncovered fewer than one one-thousandth of one percent of the archaeological sites along the Nile River delta alone. Imagine the discoveries yet to be made in the rest of the world. “Archaeology is a slow and laborious process,” she says. “That needs to end.”
Parcak wants to expedite discovery, and how she plans to do that is directly connected to her TED Prize wish. She intends to create a crowdsourced citizen science game that allows anyone to search for archaeological sites on satellite images. She’s essentially farming out the most tedious of tasks: looking for discrepancies in images. “It’s basically like using people as parallel processors,” White says.
The platform is still in development, but here’s the gist: People log on, and they’re dealt a card showing a chunk of land, likely 165×165 feet in size. This piece of land will be scrubbed of its location and map data, to keep would-be looters at bay. (“We’re treating archaeological site data like patient data,” Parcak says. “We’re protecting it.”) The game will present players with a decision tree, guiding them with questions about what they’re seeing. It might say something like: “What do you see in this image? Do you see vegetation? Do you see a modern structure? Something else?” Then, even more fine-grained: “Is it a circle? A square? Oh, is it a looting pit?” Parcak plans to provide examples of what a looting pit or tomb or pyramid might look like on the card that they’re looking at. Each card the player is dealt will look different, based on the processing techniques that have been used. If enough people say they saw the same thing, Parcak and her team will file it away as a potential excavation site. They’re still determining how many people have to agree before it becomes useful data. If and when a site is found, Parcak says she’ll only share the data with fellow archaeologists if they promise to take the people who helped them (you and me) into the field through things like Skype, Snapchat, or Periscope.
The game Parcak envisions is in the same vein as Fold It, EyeWire, and Galaxy Zoo, three games that have crowdsourced discoveries surrounding protein folding, neuron structure, and galaxy classification, respectively. They’ve even led to scientific publications in august journals like Nature.
Gamifying her work is Parcak’s latest and most ambitious attempt to boost public interest in archaeology—a field whose place in the public consciousness ranks several orders of magnitude below, say, pop stars. She’s already been tirelessly media trained. Parcak recalls the first time she had to go on camera to talk about her science. She’d just mapped Tanis in Egypt, and the BBC was doing a documentary on it. They put her through a two-day media training bootcamp where she was, in her words, “broken down.” “They were like, ‘You really think you’re good at this? Go head. Explain your science to me,’” she recalls. “And I tried, and I realized that my grandmother wouldn’t understand what I was saying.”
The media hoopla seems trivial next to actual discovery, but Parcak believes communicating her science to the world is no less important than the discoveries. “The scale of the significance of your work? The S-shape curve goes up like this,” she says, swooping her hand toward the ceiling.
Now that she needs the public’s help, it’s even more important that they not just understand but care about what she’s doing. Parcak is well suited for the job. She’s an archaeological evangelist who’s not afraid to play into the tropes of the field in order to popularize it. When I ask her if the term space archaeologist is actually accurate, she says she’s been called out on it before. “So instead I tried: ‘I’m a mid-troposphere, multi-spectral, high-resolution, ancient landscape imagery analyst,’” she says. “And the room just fell asleep. I don’t care anymore. I’m a space archaeologist, yes I’m a landscape archaeologist, yes I’m an Egyptologist, yes I’m an anthropologist. We all have all these hats.” But her mission is bigger than her job title. “I want to get the world excited about archaeology,” she says.
“Satellite images of South Abusir, Egypt. The image on the left was acquired in 2009. Looting holes can be seen in the image on the right, which was acquired in 2011.”
As we finish lunch, Parcak walks me through a rough version of her game. She explains how she recently learned the terms UX and UI, and tells me that she and her partners are trying to make the platform as intuitive as possible to players of all ages. Just watch, she says, “it’s probably going to be, like, a grandmother from Fresno who’s our super-genius finder.” She doesn’t expect everyone who plays the game to become a full-blown archaeologist, but she does hope it will boost the visibility of her field, and draw attention to the increased threat posed by looting. It’s estimated that terrorist groups have made more than $300 million off of looting antiquities in the wake of the Arab Spring. With looting on the rise throughout the Middle East, Parcak says public interest is more important than ever.
Parcak believes that archaeology is the key to understanding not just our past, but our future. Studying ancient cultures is about more than uncovering objects—it’s about gaining a clearer picture of the way humans have lived, and learning lessons from our past so we don’t repeat them in the future. “I would argue that, with all this obsession we have with going to outer space—a colony on Mars!—we should stop for a second, take a deep breath, and think about all the times we’ve tried to colonize places throughout history and how that’s never gone well.”
PREVIOUSLY on #SPECTRE
DECENTRALIZED PREDICTION MARKETS
the ACCURACY of CROWDS
COWS as COMPASS