Oct 20, 2014 · 3 minutes

Earlier this month, the Washington Post published an interview with Graham Dodge, founder of the social media health app Sickweather. The headline read, "Could you use social media to track Ebola or Enterovirus? Sickweather thinks so."

By looking at the title, a reader might reasonably assume that Sickweather, which aggregates social media posts in a user's area to chart the spread of diseases like the flu, thinks it can track Ebola. It fits nicely into that 21st century narrative where iPhones, Twitter, and other consumer Internet products made by billion dollar tech companies can solve everything.

But despite leading with Ebola in the headline (I guess the flu, which Americans are far more likely to contract, just wasn't sexy enough), Dodge specifically told the interviewer that Sickweather is not tracking the disease. He suggest say that, in the event of a widespread outbreak in a developed nation or megacity, an app like Sickweather could be useful in tracking Ebola so the headline isn't exactly dishonest. But it is a bit misleading, considering there have only been three confirmed cases in the United States. And the US's high smartphone and Internet penetration rates make it a far more effective place for Sickweather to analyze outbreak information than the West African communities most affected by the disease.

I wanted to follow up with Dodge, who I spoke with last month when his company raised $350,000 on AngelList, to talk a little more about why social media isn't the best way to track the spread of Ebola.

"The Ebola cases as they are identified are obviously making headline news," Dodge said. "We’re talking about literally three cases in the US this month. That’s compared to millions of tweets about Ebola."

It's an example of how the reaction to Ebola from the US media, both social and otherwise, is not a very accurate reflection of reality.

"99.999% of the people talking about Ebola don’t have it," Dodge adds. "That’s not the case with the flu."

Certainly, false positives can arise when it comes to flu analysis as well -- In fact, an audit of Google's much-heralded Flu Trends tracker found that it wildly over-reported the disease. But Sickweather, Dodge says, doesn't just track searches for the flu -- it looks for specific, declarative statements like "I have the flu." As proof of the app's effectiveness, Sickweather predicted the start of the 2012 flu season six weeks before the CDC did.

As more and more of our lives become intricately entwined with digital processes, it's comforting to hope that all this software will eventually help save us, and not just enslave us. But it's also arrogant to think that all the secrets to solving the world's problems lie in the head of a Stanford grad living in Palo Alto. Furthermore, it ignores the many infrastructural and political problems that often lead to or exacerbate these problems. As Nathaniel Mott wrote after Quartz reported on "apps to stop the spread of Ebola,"

Ebola has become such a threat because the systems put in place to prevent its spread have failed or been let down since it first became a problem. The World Health Organization may have responded to the threat sooner if its government funding hadn’t been cut. Workers might be more willing to clean the streets or bury Ebola’s victims if they were offered hazard pay. The organizations tasked with helping Ebola sufferers might be able to save more lives, or at least bring in more patients from the streets, if they had sufficient resources to handle them.

To quote Mott's headline, software can help save the world. But it can't do it alone. Nor can a clever app that's good at tracking flu in the United States be expected to do the same for Ebola in Africa.

[illustration by Brad Jonas]