On Fake News: Part II

You may read part I here. Might make some of terminology and sentiments here more clear but is totally avoidable.

While we may agree that fake news is a nuisance and I may have convinced myself that it is an important enough problem, a question still remains what can be done about it.

Enough people are doing their part in dealing with the problem of fake news and yet I have come across increasing number of news that I believe is directing our collective response in the wrong direction.

I believe our handling of fake news so far hasn’t created large impact because it is directed at a much smaller set of news items, influences the wrong set of people and is too slow to react. Handling of fake news currently being “fact checking” a news item.

Flaws in current solutions

Identification of fake news

Current solutions identify probable fake news based on facts represented and the action we take against the news item is to check if the facts represented are correct or not. What it has helped achieve is that fake news has evolved from outright blatant lies to modification of fine details or misrepresentation of facts to serve the ulterior motive.

Action against fake news

The action taken against fake news is that we check if the facts represented by a news item are correct or not. It is a time consuming process because of which by the time fact checking is finished, damage by the news is already done. Also, it allows some fake news to achieve its ulterior motive and in some cases ends up reinforcing it when the motive is achieved by hiding important information or adding opinion to facts.

Influence of solution

Fact checking solution is generally used by a person who is susceptible of a news. Such a person, in general, will not be easily influenced by the news and hence doesn’t contribute as much to the collective response to the news. In most of the cases, such a person is not the target audience of the fake news to begin with and hence any influence the solution has on them is inconsequential.

What I would do differently

The action against fake news, as per me, shall consist of three aspects: identifying “problematic” news, reducing its impact on individuals and limiting its distribution.

Identifying “problematic” news

I use the word “problematic” because I believe we shall act against not just the news with incorrect facts but any news which can evoke a specific response in its consumer that may be desirable for a section of society.

Any news that creates an emotional response in its consumer shall hence be identified and acted against.

Reducing impact on individuals

While current solutions do provide a tool to verify a news what they fail at is to trigger the desire to verify the news in right people and to be able to verify it at the right time. To actually reduce the impact on individuals it is hence necessary to raise suspicion in the most susceptible individuals and to provide them the tools to verify the news at the time of consuming it.

Limiting distribution

Most important aspect of fake news is that it affects our collective response and hence reducing its reach is necessary. But since not all people are impacted similarly by same news what we need to limit is reach of a polarising news to people who are susceptible to be influenced by it.

What, I think, shall be done

Polarisability of a news

Instead of prioritising certain news items over others, the solution shall universally assess all news items considering different takes on an event as different news items and hence treated independently.

By assessing polarisability, I mean, where does the news item stand on the issues or topics covered by the news. More polarised a news item is, higher is the chance that it is fake and hence making it more susceptible to scrutiny by a user.

The other side

The assumption here is that when a news item polarises against one’s inherent biases they will try to challenge it with facts that contradict the news. Bundling opposite takes on an event hence can help in discovering the truth, though TV debates prove that such a strategy can be used to reinforce biases instead of bringing the truth to light.

The reason TV debates are able to reinforce biases is that the moderator selectively gives more attention to one side of the story than the other. Existence of a mediator even in the form of a fact checker can hence be leveraged to serve one’s ulterior motives.

Allowing a user to explore the larger spectrum of opinions and takes on an event exposes them to a larger perspective and hence assisting them in forming an independent opinion which then leaves the motive of fake news to create a desired response mute.

Promoting neutrality

Reducing the reach of polarising news ensures that susceptible people are less exposed to news that may evoke specific response in them. As the reach of polarising news is reduced, creators will also focus on neutral reporting and using fact based news to influence people. While most of the distribution of news would happen on social media which would be out of anybody’s control, neutrality can be promoted by allowing users to discover the most neutral take on an event and promote that instead of sharing the first news about an event that they come across.

Notes on feasibility

• To ascertain polarisability, ML algorithms for sentiment analysis can be used. One such development was made by OpenAI while dealing with Amazon Reviews..
• It is easy to aggregate multiple takes on an event across platforms, assessing their polarisability and then creating collections of them can help in presenting a spectrum of opinions. Using similar technology as above, specific sections of news can be picked up to reduce repetition for user while still highlighting the spectrum of opinions.
• Search engines and bots that can be invoked to respond with the most neutral take on an event, given any news, can be used to help users identify neutral reporting. Such bots are possible for many platforms like twitter, reddit, telegram etc.