B.S. Detector: After It Hits the Fan
Roughly a week ago, the creators of the B.S. Detector app -- a cheeky Google Chrome plug-in that attempts to identify suspicious information floating in Facebook's content streams -- received some tough criticism after their program wrongly labeled legitimate news sites like ShadowProof, Naked Capitalism, Truthdig and Truthout as unreliable.
Seeing promise in the plug-in, major news organizations like CBS amplified B.S. Detector's harsh and undeserved judgments while the application was still in its nascent form.
As a journalist who views these outlets as important contributions to the media ecosystem, I was one of the critics who panned the B.S. Detector app as little more than a "blacklist."
In a remarkable gesture of good faith, the creator of the app -- Daniel Sieradski -- "liked" my Tweet. He then responded to my critique by inviting me into the project's chat room, listening to my concerns, and agreeing to move toward what I described as a "disclosure-based" model.
Full disclosure: I know and respect the creator of the B.S. Detector app, and the journalists at the media organizations that were unfairly classified.
* ShadowProof's editor-in-chief Kevin Gosztola and I were part of a skeleton crew of four reporters who covered every wrinkle of the Chelsea Manning court-martial, including the little-noticed pre-trial hearings. The coverage of that high-profile case would have been poorer without Gosztola's daily reporting.
* Sieradski and I met through a college friend, but I also know him through his accomplishments as a progressive blogger and activist. Those depicting Sieradski today as a "censor" on a "witch-hunt" either do not know -- or are ignoring -- his long record busting the mainstream Jewish community's taboos in honestly discussing the Israeli-Palestinian conflict.
Sieradski's lifelong commitment to free-expression made me feel that I could be candid.
When I entered the B.S. Detector chat room, I told the team that the problem with their app was that it isolated news outlets they found suspect, instead of providing information that would empower users to make their own informed decisions about their media diets.
To my pleasant surprise, the team listened to my critiques and responded to them thoughtfully, and we discussed and debated how to implement them. Some debates and disagreements got heated. Egos clashed. Methodologies got questioned. How to phrase two or three words became a source of genuine agitation.
In short, it was beginning to look a bit like a newsroom.
Then, the backlash against the B.S. Detector spiraled into every newsroom's nightmare: Publications identified as not credible threatened defamation lawsuits.
One of them, Naked Capitalism, followed through with a letter seeking more than an apology and retraction.
Channeling the righteous indignation -- and arguably the tactics -- of Gawker assassin Peter Thiel, a lawyer for Naked Capitalism's publisher demanded nothing less than to bury the B.S. Detector app and its code.
“You and the individuals working in collaboration with you or under your direction must not release your code to the open source community or use other means of distribution,” James Moody, a lawyer for Naked Capitalism's publisher, wrote in a 4-page letter.
In characteristically blunt fashion, Sieradski replied: "This entire letter is predicated on numerous falsehoods and is a greater attempt at censorship than anything I've been accused of."
On Sunday morning, Sieradski struck a more conciliatory tone in a formal response to Naked Capitalism's litigation threat. He offered his "sincere apologies" to the mislabeled outlets, which he emphasized he removed from the database as soon as he learned of their inclusion. He stressed that his app is a work-in-progress being developed by a team of "volunteer academics, librarians and journalists."
As a working journalist sharing my insights with the group, I should clarify: I joined B.S. Detector's chat room well after Naked Capitalism's inclusion on the list. I never played any role in deciding whether a publication should or should not be deemed credible.
In fact, I repeatedly argued that the B.S. Detector app should not decide whether or not a user should trust a news outlet. It should provide news consumers with information to form their own judgments.
The B.S. Detector team has not adopted all of my suggestions, nor would I expect them to. But, at almost every turn, they have engaged with my ideas seriously and constructively.
A Pulitzer Prize-winning reporter and editor joined the chat room to offer his own critiques and insights, and the group planned on how to reach out to press and free-speech groups in order to build a better and more effective application that does not unfairly stigmatize independent news media.
That's what makes Naked Capitalism's demand -- in my view -- not only excessive and a waste of resources for all involved, but also misguided.
The genie is already out of the bottle: The "fake-news" phenomenon -- and the response to it -- are not going anywhere. Organizations will continue to try to create tools to prevent the spread of misinformation over social media, which is a noble, worthy, and necessary goal. Facebook already announced its intention to address the problem by partnering with third-party verification services that will likely be guided by a profit motive. Some of those tools may be flawed and misfire. Every newsroom makes mistakes.
In my opinion, Naked Capitalism is gunning after a group of people who are volunteering their time on a free and open-source project with a proven record of openness to criticism and change.
Their commitment to transparency was so absolute that they subjected their entire brainstorming process to public scrutiny in an open chat room. Every mistake, argument, criticism, typo or even unformed idea could get picked apart, not only by thoughtful critics, but also by the project's antagonists and defamation attorneys looking for any sign of negligence and malice.
To me, it seemed that the B.S. Detector team had taken the newsroom maxim "Show your work" to an uncomfortable extreme, but it should be noted that Poynter, the leading media ethics institution, endorsed this very model.
So in the spirit of openness, I still believe that the B.S. Detector app remains deeply flawed in its current form. My criticism of its methodology designed by Merrimack College Prof. Melissa Zimdars is a matter of public record. I do not question her clear scholarship, knowledge and goodwill, but I believe she currently places too much emphasis on such subjective criteria as "aesthetic analysis" and "title/domain analysis" -- and too little focus on factual information about a news outlet's structure, masthead, firewalls and financial model. Categories like "state news" from repressive nations are too incoherent, and in theory, could be used to tag a terrific broadcaster like the Qatari-funded Al Jazeera as less reliable than for-profit media giants like Fox News.
Flagging a relatively small number of marginal outlets as fake, hate, satirical, clickbait, and conspiracy theory is not enough. If the app doesn't subject legacy newspapers, wire services, or major broadcast outlets to the same scrutiny, it will stigmatize independent news outlets by design -- despite the best efforts of its accomplished volunteers.
B.S. Detector's crew also has too few journalists on board, and as far as I am aware, the people making decisions about news ratings are librarians, coders and an academic. These are all important perspectives, but a media-education app also needs the insights of working journalists.
My knowledge of the media landscape comes from the first-hand experiences of having worked in news rooms from Manhattan Federal Court to Guantánamo Bay. I have been a member of the credentialed press at a U.S. presidential debate and an Ecuadorean presidential press pool. I am familiar with other news outlets -- from international wire services to independent financial blogs -- from having counted them among my colleagues and peers.
More journalists should share their insights, and Sieradski has welcomed their contributions. But this hard work and debate will not happen if news outlets attempt to kill the idea at its inception.
Instead of identifying B.S., the detector should act as a media-education project that helps users try to sniff it out on their own. This is a daunting task, and it needs thoughtful, knowledgeable and constructive criticism to see it to fruition.