Friday, April 8, 2016

Smoothing the Seas of Information

There is a ton of information available today: With over 130 million books and a billion web sites officially registered, there's a lot to read. Add to that enormous library user-generated content, like the over 5 million Wikipedia articles (for just English), or the 300+ hours of video added to YouTube every minute, or the over 4,000,000 blog posts/day, or the incredible amount of information shared and published and consumed every minute...It's obvious that without tools to navigate this sea of information, it is easy to get lost, be misinformed, or to never find what you needed in the first place.

60 seconds of internet: Some numbers

The question, then, is how do we make use of all this information? Is it impossible to make sense of all this "big data"? Can we harness it to harvest useful information?

The problem: how do you know what info to trust?

Since we can't know everything all the time, we must rely on curated sources of information to help us digest what's important to us. This includes authority figures like professors and experts yielding academic journals, reports, and books; information aggregation services like news outlets, regular journals (magazines), newspapers, radio, etc; and other people who may or may not know what they're talking about, consumed through things like videos, conversations, and social media outlets.

My wife is from Latvia, a country formally of the soviet union with a current population of about 50% Latvian, 50% Russian. Having both Latvian and Russian friends, it's amazing to see how differently they understand world events, or how they perceive local ones. I have friends who so distrust their government (which is, in fact, rather dysfunctional), that they easily believe conspiracy theories (like 9/11, for example). Others make rash assumptions about somebody's intentions based solely on race--and I have dozens of those kinds of stories, including several against me personally.

The fact is, the experience and expectations brought by somebody to new information heavily colors the new information. Because we are basing our understanding, opinions, beliefs, and, ultimately, our reasoning for our actions on this information, it's vital that we are able to know how trustworthy that source is. To make matters more difficult, information is often presented with little regard for fact checking (think social media, gossip), while other times, information is specifically crafted to convey only one message, such as propaganda pieces or advertisements.

News outlets, generally seen as a reliable source of information, use this to their advantage, often presenting what news events mean or how you should feel about them rather than focusing on presenting the facts. It's bad enough that there is now a service that occasionally fact-checks major US news outlets, ranking the truthfulness of their information (sadly, several networks are standing at 50% false information). A study by Fairleigh Dickinson University, re-published by the Business Insider and others, shows us that even when much of the information is factual, it doesn't seem to lead to improved understanding of domestic or international events--which is often seen as a primary purpose of news outlets (click the links above for more).
To make any information useful, 4 major problems that must be overcome:
  1. Access. You need to be able to access the information to make any use of it. This seems obvious, but a lot of meaningful, useful information is inaccessible, either because of reader illiteracy (which can mean not being able to read it, or not understanding the topic well enough to make accurate sense of the information), or because the information is controlled. An common example of information control is the paywall, where a reader must pay to access the information (Netflix, many publications). This isn't necessarily bad, since these fees can help keep the content provider afloat, and allows them to curate better information. The problem is that they can also prevent access to those who need access to that information, such as an important medical study being prohibitively expensive to doctors in developing nations.
  2. Authority. It's difficult to know whether a piece of content is trust-worthy or not. Did they consider all sides of the issue? Do they understand the core principles associated with what they are presenting? For example, a reporter named Markham Nolan describes a situation where his news organization received a video accusing a government of throwing bodies off a bridge to hide their murders. He discusses how they went about figuring out if this video is authentic, proving a war crime, or if it is a smear propaganda piece. In this case, it was an authentic video, but to show that authenticity took some rigorous investigation, which many people are unwilling to exert, and so many are able to authenticate themselves and their information too easily.
  3. Bias. It's nearly impossible to present pure information without any bias, or consume information without interpreting it through preexisting biases. A study, published in Business Insider, showed how our own bias changes our perception of a source's trustworthiness (example image below). Additionally, people are prone to confirmation bias (more on that here), or may lock into a single explanation or understanding of an issue like my Latvian and Russian friends. Interestingly, while discussing these ideas with my friends and coworkers, several people told me that they aren't interested in changing or expanding their understanding on some issues, because they are comfortable with their current understanding, even while openly admitting that they are likely wrong. Being aware of the biases and motivations of content providers' and consumers is important for communicating more effectively.
  4. Trustability of news outlets
  5. Level of presentation. This refers to the overall ease of understanding a work, which can be made more difficult by choice of vocabulary or idiomatic phrases, use of technical jargon, or making assumptions such as background understanding or feelings on a topic. For articles, you can have very technical papers that use a lot of jargon or special words, usually designed for telling others in that field about the research, or works that are just hard to read, though they could be written much more simply. You can also have the opposite problem, where something is so simplified that it really doesn't help the media consumer understand the topic. Media should support the audience it's intended for, and people outside that intended audience should do their homework to understand it. That's not to say that more effort to make "doing the homework" easier would be amiss.
  6. Calvin and Hobbs on Clarity

Proposed solutions: some work better than others

I hope it's obvious to you that there's at least one problem. Several solutions against being misled have been offered. A typical example comes from Howard Rheingold's book Net Smart, where he presents three steps to verifying information.
  1. "Triangulate," meaning you find (at least) three different, unrelated sources to verify any piece of information, because if all three say the same thing, it's much more likely to be true
  2. Look at the stakes the author/creator has. If the person is blogging for free, they are likely to have a lower stake than somebody who's job depends on the accuracy of the information. Often, the one with more to lose will be more reliable.
  3. Use your inner "crap detector." If something seems fishy or off, it likely is. This comes from experience, so isn't terribly helpful if you're new to some topic.
The problem with this approach is that companies and special interest groups know this is what you do. There's a lady who works in media, named Sharyl Attkisson, who gave a TED talk discussing how groups use this approach to create multiple, credible-looking sources of information, so that if you rely only on those three basic tools, you will likely believe their misinformation. I was shocked when I first heard this, but the more I thought about it, the more obvious it seemed. I feel it's worth your time to watch her entire talk:
She suggesting looking for tell-tale signs of planting this information, including:
  • Use of inflammatory language, like myth, psudo-, conspiracy, lies, "proves"
  • Claims to debunk myths that aren't actual myths. Often these myths are also fabricated, creating a two-prong lie: convincing those who believe the myth debunked that they have been enlightened, and giving a pat on the back for those who think they're too smart to fall for such a myth. In either case, a non-myth is now considered a myth.
  • Attacks against people or organizations, rather than discussion of ideas or facts
  • Public skepticism directed at those who question their claim
Additional ideas generated in discussions with friends and family as tools against misinformation include:
  • Leveraging subject enthusiasts, as they often can have experience to match experts and tend to be more helpful to newcomers. You can often find them in forums dedicated to the topic in question.
  • Seeking divine revelation about a topic. Most people around the world consider themselves religious, and many of these believe in a God of truth, including truths on social, scientific, and other issues. Thus, they feel they can approach their God to ask for clarity on issues of import to them.
  • Understanding the "basic principles" or core principles of a topic, which could be underlying physics for natural phenomena like climate change, or understanding specific processes, like how voting districts and delegates logistically work for understanding elections in the United States.
  • Discussing the information with peers, friends, and neighbors for consensus. While this isn't always the best way to verify information, often a friend or colleague will have a different understanding of the information, and when combined lead to a greater clarity about the topic.

Another Proposed Solution

These techniques can help us detect problematic information and understand how to integrate new ideas, but unfortunately they don't do much to ultimately forward solutions to these problems, or help to harness the massive amount of data now available. A true solution needs to:
  1. Be able to make use of collaborative effort, which can gain access to multiple levels of understanding, multiple perspectives, and increase likelihood of finding help understanding a topic
  2. Be open to those who need it and have services that cost money so the solution can sustain itself, improve it's offerings, and grow
  3. Provide verification of content providers' authority on the topics
  4. Clearly identify sponsors, supporters, and other potential conflict of interests that color the presentation of that information
  5. Offers multiple levels of presentation and background information to help orient people completely unfamiliar with a topic
I have been working with a couple professors and multiple friends and colleagues on a proposed solution that meets the above criteria. The hope would be to create a source that is seen as authoritative and valuable to the average person and to the experts (see separate blog post about that). Yet even if this tool took off and had 500 million contributing users, it still wouldn't fully solve the problem, because communication remains the glue that holds our society together. And because communication requires humans somewhere in the mix, misinformation, bias, misunderstanding, and malice are prone to entering.

But I believe that by utilizing the tools above, and actively working together on new tools like the one discussed above, we can all sail the seas of information with fewer storms, wrecks, or days of bad weather. And who knows--maybe we'll even make landfall in an incredible new land of promise.

No comments:

Post a Comment