Crimson Hexagon founder spearheads academic research partnership with Facebook

A Q&A between Jo Ann Corkran, managing partner of Golden Seeds and member of the board of Crimson Hexagon, and Gary King, Crimson Hexagon founder and board chair

June 7, 2018


In the following Q&A, Golden Seeds Managing Partner Jo Ann Corkran talks with Dr. Gary King (left), who is the Albert J. Weatherhead III University Professor and director of the Institute for Quantitative Social Science at Harvard University, and the founder and board chair of Crimson Hexagon, a Golden Seeds company. Crimson Hexagon helps global brands better understand consumers through instant access to the world’s largest volume of unstructured text and images across social, online public, and enterprise-held data sources.

Jo Ann and Gary discuss the partnership he recently helped broker between Facebook and the academic research community to analyze the impact of social media and digital technologies on societal issues, including political elections.

GK: The Cambridge Analytica scandal was a wake-up call about consumer privacy and election interference. The model we recently announced with Facebook is designed to curb concerns by engaging members of the academic community to study the impact of social media on democracy. The goal is to unearth potential abuses of data, without compromising the privacy of Facebook’s 2 billion users. Historically, academic researchers haven’t been able to access private Facebook data for social science advancement. Now, trusted academics can do so to help address some of society’s most pressing challenges.

GK: Nate Persily of Stanford Law School and I set up this model to be completely independent of Facebook, as well as free from the influence of financial or political incentives. Instead, a commission of respected scholars will oversee the peer review process for data access. Facebook won’t have the opportunity to approve access or review research findings prior to publication.

GK: Social scientists use data for the greater good. There’s more data now than ever before, but academics can access a small fraction of it because companies aren’t willing or able to make it available. There’s a way to do that that is acceptable to all parties. We’re not going to take everyone’s data and make it available. In certain cases, social scientists will be able to see pieces of data related to specific questions or categories. Granting requests will be a collective responsibility of researchers who know how to protect highly sensitive data — we have data at Harvard, for example, that could threaten lives if revealed. The research we’ll enable with aggregated Facebook data is for finding social good. Academics all over the world will be able to apply, and some smaller number will be approved after peer review. Facebook retains the power to disallow certain questions or make certain data available, but if it does, the commission will be obligated let the public know this occurred.

GK: If you look inside a company, you’ll probably find an HR system, a finance system, an air-conditioning system and others. If you turn on a spigot attached to any of those, out comes an enormous amount of information. That data exhaust can be repurposed. We could use it to learn about a lot of things related to the business, but also things not related.

Facebook has 2 billion people on its platform, so we can learn a lot of things outside of what’s relevant to the company for its own business purposes. A simple case might be about determining the average income in certain regions or among certain demographics. That kind of information can benefit future generations, similar to how patients in randomized clinical trials for, say, cancer, don’t benefit from those trials, but future generations will. This is a less serious scale, but a similar idea. We expect that a year from now, there will be a lot of requests for proposals and papers that rely on this data. I suspect many will be important for society but irrelevant to Facebook.

GK: Facebook is legally required to maximize shareholder value, and that’s it. Perhaps going after social good increases value, but perhaps not. So why do it? First, here are people in universities that Facebook couldn’t hire, and vice versa. The flow of information back and forth is valuable for everyone. They’d also like to hire our students, and our students would like to go work there. However, the way I argue Facebook should think about it is if we learn they should really change a particular feature that shouldn’t be there, and they learn from our findings and fix the problem, they should get the credit for that. They might even have a competitive advantage by claiming an outside group of academics has checked over the system. So, next time Mark Zuckerberg has to talk to Congress, he can say we had the brightest minds in academia scrutinizing this; you don’t have to take our word for it. We have impartial, expert analysis.

GK: I created the original algorithm behind Crimson Hexagon’s AI-powered Consumer Insights Platform, based on my experience in statistical social science research. We continue to develop the technology to understand unstructured social media, and our new algorithm enables us to answer more precise questions about what people are saying. Like brands, social scientists are not interested in what any one person is saying, but in what everyone is saying. The relevant data changes all the time. You start with unstructured text, and then you can also analyze images, video, audio, hashtags and connections between people.

GK: Science is about a community of scholars working in competition and collaboration to achieve the same set of goals. If Facebook or its data scientists want to participate in that, that’s great. There are plenty of people in the academic community who will respond. And if Facebook ever wants to issue a release with a disagreement about an academic paper based on its data, it’s allowed to do that. We would learn more if one researcher had a finding and goes back and forth with Facebook or another researcher about the details, the underlying data, attempts to replicate the findings, etc. That is the scientific process. Disagreements can yield progress.

Learn about the work of Golden Seeds.