Is It Time For New Guardrails On Social Media Algorithms?

A key reason for the rise of advertising on social media is the ability of algorithms to identify priority customers, message them with more relevant content and ultimately build a strong, emotional bond with them.  Good marketing starts with in-depth research on the customer to ideally find those “hot emotional buttons” that excite them and sustain their loyalty to a brand.  Nothing is more effective for learning more about customer nuances and behavior than what these algorithms can uncover, although privacy has become a big issue, especially in Europe and with Facebook.  .

Unlike traditional media like TV, print or radio, algorithms in social media allow it to be more interactive.  Essentially algorithms are used to sort content in each user’s feed, and every social media platform offers a unique algorithm.  Facebook for example focuses on meaningful customer engagement and its algorithms are designed to rank content based on popularity, content type, relationships and recency.  Conversely LinkedIn is a leader in B2B marketing, with algorithms dedicated to networking rather than building followers.

Twitter was introduced in 2006 and was focused on timely news and posts which were ranked by the day and time, not the content.   However it soon was overflowing with problematic content and many accounts were banned for “engaging in hate speech, promoting Nazi, white nationalist material, child exploitation, suicide, self-harm and harassing users” (this included Trump’s account).  As a result, Twitter created an advisory group, the Trust and Safety Council, to eliminate these hate-filled posts.

While these algorithms brought a fresh path for advertisers to reach their customers, it became apparent that there was a serious downside, especially the spread of divisive misinformation.  Algorithms on social media invisibly select the posts or news, not the user, and this usually means that users see only the news that confirms their existing beliefs, pushing people toward more extreme content.  Engagement driven algorithms often amplify and reward misinformation that reinforces opinions which can be outrageous and even harmful.

The Danger Of One Person Controlling Algorithms

Elon Musk with his recently re-branded Twitter (now “X”) is a good example of how one person or company can wield so much dangerous power to influence people’s opinion and contribute to the hateful environment in America.  Musk proclaims himself a “free speech absolutist”, basically positioning his new X platform as a source for any and all posts, no matter what the truth is and the harmful ramifications for its audience.  Since taking over Twitter last October, 2022, he has disbanded this advisory group, reinstated accounts previously banned for hate speech, and laid off workers responsible for weeding out problematic content.

Musk has a political agenda aligned with the far-right social network and is dedicated to defeating the “woke mind virus”, which he associates with Democratic and progressive policies.  Here are some recent actions by Musk which are particularly alarming and demonstrate the risks of algorithms controlled by unpredictable managers with so much power:

• Antisemitism – After a controversial meeting on August 29 with the Anti-Defamation League (ADL) CEO, Jonathon Greenblatt to “address hate” on the X platform, there has been a surge of anti-Semitic posts under #BanTheADL, spurred by a “like” that Musk added to a post by a far-right activist who called for the banning of ADL.  Musk is even scapegoating Jews for the declining value for X, now estimated at one third of the $44 billion he paid.
• Restricting Ukrainian Military – Musk has helped the Ukrainian war effort by letting it use his SpaceX Starling satellite network for intelligence, but refused last year a Ukrainian request to use it for a drone attack on a Russian naval fleet..
• Siding with Putin – Musk promoted a settlement of the Ukrainian war which reflected Putin’s plan that gives Russia about 20% if its territory. Putin afterwards called Musk “an outstanding person”, similar to the praise Trump expressed for Putin years ago as “brilliant and talented”.

To minimize divisive misinformation, algorithms should be more transparent and provide users with a choice of what they can read or see on social media.  Indeed, freedom of speech should have some limits.