Do you know where you are on the internet?
Per the laws of physics, us humans live in a relativistic world. Is the hill small or large? An answer depends on where the subject of the question is standing in relation to said hill. New information is always colored by some aspect of physical perspective. "Did you hear John is going to quit," one brand of office gossip might go around. "No, where did you hear that," is a natural and common answer to such untested news.
The quality of information depends on its source. The further away one is from a subject, the more distorted and unreliable their perspective. Most people exercise this form of judgement almost automatically. People are naturally skeptical of stories they aren't familiar with.
But what if the subject in question doesn't exist in real space, but in cyberspace. The internet, where a majority of Americans now get their news, doesn't easily reveal the details of physical context that is so important to our judgement. If your loose-lipped neighbor starts spouting paranoid theories about someone on the next block, most people learn over time to practice skepticism in interpreting his or her diatribes.
But what about the stories you read on Facebook. Do you know where they are coming from? Not in terms of the URL of the website these stories are published on, but the actual physical origin, the proximity of the storyteller to the subject or subjects being talked about?
The tools we use to access and navigate the internet mask this type of contextual information in a way that makes it very difficult to determine exactly where a particular strand of news information started or came from.
This phenomenon is studied in an academic research paper published earlier this summer by the Data and Society Research Institute, called 'Media Manipulation and Disinformation Online'. It is also the driving problem statement behind Grapple's forthcoming prototype. More on that down-page.
Ted Cruz and the JFK assassination is a perfect example.
Consider the story of Texas Senator Ted Cruz's involvement in the John F. Kennedy assassination. Did you know that Sen. Cruz's father was involved in the nation's last presidential assassination? If you believed the current president, you would.
Trump heard about Cruz's involvement after reading a story in the print tabloid National Enquirer, then brought it up during an interview with Fox News. His discussion of Cruz's involvement with the JFK shooting made news headlines in publications ranging from Breitbart to The New York Times.
By the time the story reached your screen, it could be shaded in any number of guises. This headline was published hours later on the New York Times' website:
f most people who came into contact with this story knew exactly where this information about the senator's father originated, it may not have had quite so much traction during the Republican primary. Before this story came out of the president's mouth, before it graced the pages of the Enquirer, it came from a website called the Wayne Madsen Report. Madsen is a widely debunked conspiracy theorist and a regular on Alex Jones' InfoWars conspiracy program.
Madsen enjoys a niche of dedicated followers, similar to Jones' and other fringe conspiracy groups. The mass aggregation that occurs on the internet has a tendency to launder sources like this, however, and can lead to widespread confusion on topics such as the Kennedy assassination. The Data and Society Research Institute explains:
This is not nearly the only time that a widely criticized Madsen report made it to the pages of main stream news publications. Madsen convinced a reporter at London's Observer newspaper that he was a former NSA contractor in the vein of Edward Snowden. The newspaper ran a headline proclaiming 'Secret European Deal to Hand Over Private Data to Americans."
That time the power grid (wasn't) hacked by Russia.
A similar thing happened to a less politicized story back in December. Relying on an anonymous source, The Washington Post ran a story that Russian hackers had gained access to the U.S. power grid through an electric utility company in Vermont.
The source, who was never identified, misunderstood a security breach that had taken place at the utility, where a Russian computer virus was detected on a corporate laptop. The laptop was in no way connected to the power grid or any ancillary systems.
The post, possibly short staffed due to the holiday period, was slow to correct the story. By the time they had, several elected officials were talking about the power grid breach in interviews quoted by major network television stations and national newspapers.
The power grid story is yet another example of the internet, and the practice of news aggregation -- or sharing other websites news stories -- masking one of the crucial contextual details about the story: where it came from.
This startup company has an idea to help.
Grapple is building an algorithmic tool that essentially cuts through this news and information labyrinth to answer the proverbial question: "where did you hear that from?"
Our prototype, rendered below, might appear a bit complex. It will be, at first, as we are designing the tool as an aid to help journalists track news sources as the move across the complex web of the internet. As we develop the software, though, we will soon make the process of recognizing where a news thread originated as simple and seamless as the spell checking tools and search engines most of us are now accustomed to looking at.
Even more ambitiously, once our application becomes skilled at mapping out networks of news information, it can be keen on recognizing inconsistencies across the network. Grapple could quickly point out misspelled names, or garbled statistical figures. If a quote was parsed, it could offer a quick clip of the full quote as it appeared in its original context. Making this type of research automatic will change how people interpret the content they are reading online, and in turn will change the demands that the voracious online market places on content producers, like bloggers and newsrooms.
Grapple will never pass judgement onto a news source. We aren't ranking or rating news sources based on some objective or subjective measure of quality or credibility. Our aim is simply to connect the dots for internet news readers, just as Google and its predecessors connected the dots of formerly disorganized hyperlinks years ago.
Search engines allowed people to access more information than ever before, but they relied heavily on algorithmic shortcuts to give users the content that they were most likely to be searching for. These algorithms allow people to find information without putting in the legwork of figuring out exactly where the information they wanted was published.
Our plan is to build something of a patch. Grapple won't at first compete with Google as a standalone search engine, or with Facebook as a social content delivery platform. Grapple will augment the competencies of the tools already out there, seamlessly filling in some of their deficiencies, just as the search engines themselves made up for the deficiencies in the web indexes people used before. Think about AOL, and those hyperlink pages. People used to have to actually click through webpages to find things of interest!
Perhaps someday people will no longer see the internet as an overwhelming stream of content, but an easily sortable network of pieces. After all, just about all of the world's information is now online in some form or another. Finding and understanding it remains the ultimate problem.
In the next month we will be activating a prototype of the software described above. Making the application completely public is cost prohibitive at this point, as we have to pay for the storage space and the data that we use. We do need interested avid news readers and researchers of whatever creed to use the software. Participation is the only way that we can improve this rough prototype into something that everyone can someday have access to.
If you're interested in trying out Grapple's prototype news research application, shoot me an email at firstname.lastname@example.org. We would love to have you!