Monthly Archives: December 2009

Behind MIT’s DARPA Weather Balloon challenge win

MIT Red Balloon Challenge Team

As many of you know, the MIT Red Balloon Challenge Team (part of the MIT Media Lab) won the race to locate the DARPA Red Weather Balloon Challenge.    Their system was a reverse Ponzi scheme where those finding the balloon got $2000, and those progressively farther back the invite chain in finding those people got progressively lower payouts;  the surplus got donated to charity.  (Because the payoffs were cut in 1/2 with every additional degree of separation from the balloon finder, there is no way that MIT could owe more than $4000 per balloon, even if path links to MIT were very long, and MIT assumed that many of the path lengths would be short.)

MIT team members reported that they sent out 2 million SMS messages as one of their strategies but that was a complete bust as far as finding the balloons.  Twitter and Facebook on the other hand were far more effective.  They are going to be subsequently distilling their findings on effective viral communication and sharing it at an appropriate venue.

Their victory was a victory of human connections (“social capital“) over number crunching.  A Google Team was racing them using number crunching and image recognition techniques (e.g., crawling the web in real time for images of red balloons) and had spotted 9 of the 10 balloons when the MIT Team found 10.  The MIT Team noted that the balloon finders were using Google Map to determine the coordinates of their balloon sighting (to report to the MIT team) and Google could have captured that information and used it for their own proprietary team but didn’t.

See a blog post about the basic architecture of the MIT reward structure.  DARPA’s network challenge obviously has implications for how to effectively and rapidly spread information in the event of an attack, although clearly the task here (spotting a red balloon) is infinitely easier than other possible challenges which are less observable to the the naked eye (infectious diseases or biological attacks) or actions who cause is less clear (a plane crashing for instance).

DARPA noted how the challenge explored “how broad-scope problems can be tackled using social networking tools. The Challenge explores basic research issues such as mobilization, collaboration, and trust in diverse social networking constructs and could serve to fuel innovation across a wide spectrum of applications.”

Read more here.

Health benefits of volunteering

The NY Times had an interesting piece yesterday on the health benefits of volunteering.

They cite Stephen Post’s work, which I have discussed earlier, but also notes a 2002 Boston College study (Paul Arnstein et al.) and a California Buck Institute for Age Research study (Doug Oman et al.).

The article discusses several studies that suggest that the causal pathway may run through lower stress and a “helper’s high”.

See the underlying article:  “In Month of Giving, a Healthy Reward” (NYT, Science Times, Tara Parker-Pope, 12/1/09)

Wired to cooperate?

An interesting article in the Science Times section of the NY Times discusses research by developmental psychologist Michael Tomasello (appearing in Why We Cooperate).

At 18 months of age, Tomasello finds that toddlers almost universally immediately help out an adult who needs assistance because his/her arms are full.  What Tomasello, co-director of the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany, doesn’t know if altruism is innate; 18 months is typically an age before toddlers are taught how to behave.

Not so fast.  As the Times points out: “It’s probably safe to assume that they haven’t been explicitly and directly taught to do this,” said Elizabeth Spelke, a developmental psychologist at Harvard. “On the other hand, they’ve had lots of opportunities to experience acts of helping by others. I think the jury is out on the innateness question.”

But Tomasello observes these same helping behaviors across cultures, despite the fact that they teach social behavior on different schedules and may have different beliefs about appropriate social rules.  Moreover, the helping that infants observe is not enhanced by rewards, causing Tomasello to doubt that it has been strongly reinforced.

In some cases Tomasello observes helping behavior with regard to information in infants 12 months old, and even chimpanzees under certain conditions exhibit helping behavior at a young age.

More specifically related to “social capital”, Tomasello finds that age 3, children become less indiscriminately helpful and are more likely to reciprocate earlier helpful behavior from another person.  In addition, they start to develop social norms, like the reciprocity at the heart of social capital.  Parents can help foster these norms through “inductive parenting”, helping children to learn the consequences of their actions.

As the Times points out:  “The basic sociability of human nature does not mean, of course, that people are nice to each other all the time. Social structure requires that things be done to maintain it, some of which involve negative attitudes toward others. The instinct for enforcing norms is powerful, as is the instinct for fairness. Experiments have shown that people will reject unfair distributions of money even it means they receive nothing.” [See discussion of dictator games and ultimatum games in this blog post.]

Read the very interesting article “We May Be Born With An Urge To Help” (NYT, Nicholas Wade, 12/1/09) about how humans try to sort out whether to behave selfishly or altruistically.

Loneliness contagious? An oxymoron? (UPDATED 5/2013)

It seems contradictory.  How can the lonely (who are largely outside of social networks) get their loneliness through social networks?

It makes sense when you think of people’s movements in social networks over some period of time.  The lonely may not always have been lonely, but gradually, they tend to cluster together on the periphery of social networks, suggesting that the social connections with other lonely people exacerbates any de facto loneliness they experience.  That the lonely would be somewhere on the periphery of social networks is somewhat tautological, but that the lonely would cluster together at the periphery is not, and is surprising.

The study is by well-respected researchers (psychologist John Cacioppo from U. Chicago, Nick Christakis from Harvard’s School of Public Health, and James Fowler, a political scientist at UCSD) and will appear  in the December issue of the Journal of Personality and Social Psychology.

The Washington Post notes:

Although the study did not examine how loneliness spreads, Cacioppo said other research has provided clues. People who feel lonely tend to act in negative ways toward those they do have contact with, perpetuating the behavior and the emotion, he said.

“Let’s say for whatever reason — the loss of a spouse, a divorce — you get lonely. You then interact with other people in a more negative fashion. That puts them in a negative mood and makes them more likely to interact with other people in a negative fashion and they minimize their social ties and become lonely,” Cacioppo said.

[The research comprising almost 5,000 people  interviewed every two years between 1991 and 2001] showed that having a social connection to a lonely person increased the chances of developing feelings of loneliness. A friend of a lonely person was 52 percent more likely to develop feelings of loneliness by the time of the next interview, the analysis showed. A friend of that person was 25 percent more likely, and a friend of a friend of a friend was 15 percent more likely.The effect was most powerful for a friend, followed by a neighbor, and was much weaker on spouses and siblings, the researchers found. Loneliness spread more easily among women than men, perhaps because women were more likely to articulate emotions, Cacioppo said.

The researchers said the effect could not be the result of lonely people being more likely to associate with other lonely people because they showed the effect over time. “It’s not a birds-of-a-feather-flock-together effect,” Christakis said.

See “Feeling Lonely?  Chances Are You’re Not Alone” (Washington Post, Rob Stein, 12/1/09)

See also blog posts on the contagion effect of happiness, smoking, and obesity.

See article “The Science of Loneliness” (New Republic, Judith Shulevitz, May 2013)