Commentary

Digital data are not a shortcut to fact-based truth

May 15, 2021 9:00 am

State and local officials say they need more federal aid to combat cybercrime. (Creative Commons photo via Pxhere)

Before tackling algorithmic “truth” — data that affirm some views over others — ethicists typically assert that actual, factual truth should be the cornerstone of any value system.

If you rely on machines for perception, you’ll end up seeing or believing things that are not there and overlooking ones that are.

To be sure, perception varies from person to person based on experience, education, culture, religion or another variable, but truth is a constant. Many of humanity’s woes emanate from mistaking one’s perception as reality.

Truth is often relative, although there are a few universal tenets. They include such absolutes as it is wrong to lie, to steal, to humiliate others and right to be generous, responsible and honorable. Nevertheless, the pursuit of fact-based truth sharpens perception, expands awareness and deepens conscience.

Truth is subject to proof, without which we have opinion. And while everyone is entitled to that, machines are thought to transcend any human point of view.

That, of course, is a myth.

The myth of ‘undiscovered truth’ from data

Technologists have embraced that myth for decades. They claim digital data helps discover hitherto unexplored truths.

The problem with that is how data are used, especially in social media, and interpreted by users whose perception is skewed.

Case in point: At 4 a.m. on Nov. 4, as presidential votes were being tallied in multiple states, the polling website fivethirtyeight.com posted this chart showing a huge spike in the Wisconsin vote for Joe Biden. In actuality, all that happened was the city of Milwaukee uploaded absentee ballots, which had taken time to verify and assemble.

This Facebook post, which garnered a mere 38 reactions, was shared 77 times, claiming someone stealthily added votes to tilt the tide to Biden. Others picked up the deduction. An Instagram post intimated an illegal vote dump and conspiracy. Twitter erupted. Within hours, President Trump retweeted the false narrative to his 89 million followers.

By then, news organizations were covering allegations. Soon talking heads took over — some Democrat, some Republican — in disputing or affirming the now mysterious early morning pro-Biden lead.

In the months that followed, this and related voter fraud charges made their way through Wisconsin courts to the U.S. Supreme Court, which finally on March 8 declined to hear the case.

Keep in mind that this is technological — not political — commentary. You may dispute that. After all, social media assumes every example, every utterance, should be viewed through the prism of divisive politics.

If you want to cast blame, level it at how we receive and cipher information.

Big data, bigger lies

In an article titled “Algorithms and data construct ‘truth,’ not discover it,” artificial intelligence expert Kalev Leetaru debunks the idea that algorithms adjust for human bias with “pristine mathematical perfection that captures the world as it is rather than the world biased humans would like.”

Unfortunately, he notes, the world depicted by data is especially prejudiced. Eventually, people realize this. But when governments and organizations do, they manipulate data to replace algorithmic “truth” with “the preordained outcome they desire.”

The result is mass obfuscation of reality.

The Eticas Foundation, which analyzes technological bias in the interest of public debate, acknowledges that “the public largely believes that machines are neutral arbiters.” However, the foundation questions whether algorithms “amplify and extend” discrimination prevalent in society.

In that vein, a recent study finds that algorithms ingest unchecked information that engages people interactively, often via social media, further eroding truth. People then base decisions on the biased data, which algorithms also gather, leading to increasingly unreliable information.

Machine vs. humane values

This above effect is not new. In fact, George Washington — purported never to tell a lie (a historic fabrication) — believed authority figures bamboozled citizens by “concealment of some facts, & the exaggeration of others, (where there is an influence) to bias [a] well-meaning mind — at least for a time.”

His remedy? “Truth will ultimately prevail where pains are taken to bring it to light.”

To this day, that remains the solution, which I have documented since 2004 in my Oxford University Press books, the latest titled “Interpersonal Divide in the Age of the Machine” (2017).

Given the time we spend on social media, now more than two hours per day (apart from other screen time), we have replaced humane values with machine ones. They include:

  • Importance of self over others.
  • Boredom over attentiveness.
  • Oversharing over privacy.
  • Entertainment over knowledge.
  • Distraction over concentration.
  • Incivility over empathy.
  • Affirmation over information.
  • Belief over fact.

If you want to change perception so that it is closer to reality, just reverse the aforementioned tenets. Make others as important as yourself. Be attentive. Protect privacy. Focus. Be empathic, civil. Seek knowledge and fact-based information from a variety of trusted sources.

If you do, as George Washington promised, the truth will come to light.

Our stories may be republished online or in print under Creative Commons license CC BY-NC-ND 4.0. We ask that you edit only for style or to shorten, provide proper attribution and link to our web site. Please see our republishing guidelines for use of photos and graphics.

Michael Bugeja
Michael Bugeja

Michael Bugeja is the author of "Living Media Ethics" (Routledge/Taylor & Francis) and "Interpersonal Divide in the Age of the Machine" (Oxford Univ. Press). He is a regular contributor to Iowa Capital Dispatch and is writing a series of columns on the topic of "Living Ethics."

MORE FROM AUTHOR