A set of instructions for making a decision or performing a task. Sorting names alphabetically is a kind of algorithm; so is a recipe for making chocolate chip cookies. But these simple formulas bear only a distant relationship to the computer code used by social media giants like Twitter, Facebook’s parent company Meta Platforms Inc., and Alphabet Inc.’s Google.
Early iterations of social media platforms often recorded user posts and comments in reverse chronological order, with the most recently posted appearing first in people’s feeds and timelines. That all changed in 2009, when Facebook started filling news feeds with content that the algorithm determined that individual users would be most interested in seeing. Other platforms followed suit and began testing their own versions, citing user experience as motivation. Engagement levels are highest when the “best” posts, as defined by the algorithm, are more easily found. This truth underpins the success of search engine giants like Google, which rely entirely on algorithms to deliver quick and relevant results to queries. The change is credited with driving the growth of the platforms and making them much more profitable, as ads are easier to sell to engaged readers than bored scrollers.
Current social media algorithms rank posts in a user’s feed based on a combination of signals that indicate relevance, only one of which is an item’s post date. Companies can use the vast caches of data they collect about individual and group behaviors to personalize recommendations. For example, platforms may prioritize content from your closest friends and family if those are the accounts you interact with most often. Users who frequently interact with content of a specific political bias are likely to receive more content judged to meet that preference. Each social media company prioritizes different criteria when determining what content is most relevant, although the details remain under wraps. Facebook says it prioritizes meaningful interactions to generate friendly conversation, while Twitter says it prioritizes quality and relevance to display the best content.
4. What does Musk want to change?
Musk has released a number of notions in the name of defending what he describes as ideals of free speech, including a pledge to take a minimalist approach to content restrictions. During a panel discussion at the Ted2022 conference on April 14, Musk said he would be “very reluctant to delete stuff” and “very cautious with permanent bans – I think timeouts are better.” He talked about trying to “authenticate all real humans” as a way of differentiating between bots and legitimate accounts. And he advocated opening up Twitter’s content algorithms to public scrutiny, which identify spam or posts that violate the site’s terms of service, as well as sort items to sort user feeds, among other things.
5. What does this mean?
Most likely, Twitter’s proprietary “open source” software would make some of its algorithms available for public inspection. Once others can read the code, they can use it for their own applications or they can make suggestions to Twitter’s own developers for changes. In other industries, such as cryptography, almost all relevant algorithms are open source so that entire communities of coders can collaborate on improving design and security features. What is unclear in the context of social media is how much transparency would make an algorithm’s code available, as seeing the code doesn’t tell you exactly how it works in practice when operating in collaboration with the vast amounts of data that feed it. And with these immensely complicated systems, it’s unclear how much seeing isolated pieces of algorithmic code would reveal about how all the pieces work together.
6. What can this mean for the industry?
Potential competitors may have a better chance of establishing themselves in a market now dominated by giants, because they would be able to leverage the intellectual property of one of the biggest companies in the industry. This could eventually increase competition. New social media platforms particularly struggle with content moderation – it’s hard to grow to a size where you can pay moderators if potential users are turned off by unmoderated content. On the other hand, algorithms can be of limited use without the constant flow of new data that existing social media companies use to continue training and tweaking their formulas.
Yes, for both users and social media platforms. Some experts warn that Twitter’s open source can bring several problems. For example, it can give bad actors a greater understanding of how to manipulate the system and can fuel the proliferation of bots on the platform. “When you put something on the open market and make it available for everyone to use, it can be misused,” said Arun Kumar, director of data technology and marketing at advertising firm Interpublic Group of Cos. Recommendation algorithms are proprietary and form the core intellectual property of platforms like Twitter, Facebook and TikTok.
One option is a compromise that allows user communities to play a role in the development of the platform while still protecting their commercially sensitive data. Jason Mars, a professor of computer science at the University of Michigan, said Twitter could establish an open-source version of its recommender system and use it to inform the core algorithm that determines what users actually see.
• A timeline of Musk’s journey from Twitter shareholder to potential owner
• QuickTakes on the controversy over Facebook’s algorithms and Twitter’s spam bots.
• An article in MIT Technology Review about Musk’s plans for Twitter and open algorithms.
• Scientific American advocates more transparency in social media algorithms
• A 2021 study by Twitter on whether its algorithms amplify political content.
More stories like this are available at bloomberg.com