If you have been active on social media apps recently, you may have seen a trend where people lightheartedly post their old photos and memories from 2016. Jokes aside, I think AI companies could benefit from this trend, using it to train their image generators and maintain their upper hand in marketing.
The popular 2016 trend began in late December 2025 and reached its peak at the turn of the year in January. It uses phrases and hashtags like “2016” and “2026 is the new 2016.” This may seem like a collective desire to push back the clock and rewind to a time before COVID-19 and many other issues. While this may be true, my speculation is that AI and technology industries gain programmable knowledge from this deeper wealth of information.
As most of us know, AI-generated photos and videos already dominate social media. Jonathan Drake Steele, a founder of cybersecurity consulting firm Steele Fortress LLC, called the 2016 trend a “goldmine” according to Cybernews, “because it solves AI’s biggest challenge: temporal data.” With trends like these, AI has immediate access to aging stages and progressions, rather than relying on access to photo libraries or sifting through old posts.
Because these AI models require images of the same person across time, training this way can get expensive to maintain. By adding a “2016” hashtag to your old photos, you could do a lot of the work for the information harvesters.
AI algorithms run most mainstream apps on the market, so it is not irrational to me that these companies may intentionally manipulate trends to their benefit. Even your Spotify DJ is a “personalized AI guide that knows you and your music taste so well that it can choose what to play for you.”
AI is diluting music and social media, and Meta feeds its current users’ data in an attempt to grow their FrankenAI. Arthur from HeyData says, “Meta (formerly Facebook) plans to train its AI models with content shared by users on Facebook and Instagram, including text, images, and reactions.”
So, what do you think? Why would an algorithm benefit from seeing humans age? Is it so that it has an easier time generating images of people, which would make reality even harder to distinguish? Or is it a reason more sinister than we could ever imagine?
Sophie St. James is a junior English major from Central, South Carolina. Sophie can be reached at [email protected].
