Digital Engagement #4

On content creation sites such as YouTube, Instagram, or Tik Tok, a large appeal of using the site is being able to “follow” or “subscribe” to individuals you like for their content. For many people, these creators are “idols” they develop a fondness for due to their unique or engaging personality. However, how much of their personality, editing, or marketing choices has been influenced by the site they are creating on? The question I decided to address in my fourth digital engagement assignment is how AI shapes current human behavior, specifically, how AI tools shape decision-making in digital environments.

First, I decided to consult AI text generators to see how they would respond to this question. I was curious to see what the algorithm knew about the extent to which the internet and internet traffic is influenced by AI. When I asked, “How much is YouTube’s algorithm influenced by artificial intelligence?” Microsoft Copilot answered by saying that personalized algorithms, engagement metrics, and content moderation are all influenced by AI tools. Interestingly, it also mentions the bias and echo chambers that AI can create within one’s feed. ChatGPT also mentioned how AI impacts YouTube’s engagement metrics and personalization. Additionally, the tool mentions how the algorithm can run content analysis to understand its topics, context, and relevance. When I asked both tools about other social media sites such as Instagram or Tik Tok, they both made extremely similar points, such as how artificial intelligence influences user behaviors, content analysis, and the personalization of their feeds. However, Microsoft Copilot also mentions that AI affects the user’s search engine on Instagram and Tik Tok, giving the user a list of recommended topics to search for.

After consulting the AI tools, I decided to start searching for articles on the topic from business and technology companies. Similar to what the AI tools had previously stated, Digital Innovation and Transformation, Search Engine Land, and Forbes all explain how YouTube uses a “two-face” algorithm to moderate content that could potentially break the site’s policies and generate the personalized algorithm. One “face” of the algorithm is “candidate generation”, which compares a user’s history with other users’ key information such as their demographics and how many and the types of videos watched. The second face is the “ranking network”, where YouTube measures the time that users spend watching videos as a proxy for the “quality” of the videos and uses that data to make predictions on how likely they are to watch a video. The Digital Innovation and Transformation article mentions how YouTube also used AI when creating a tool that adds ‘chapters’ to videos, which allows users to see clips of videos before watching them. Sked Social explained how Instagram uses a set of AI “system cards” to evaluate a user’s information in order to make decisions on what type of content they should receive. Some factors they consider when creating the “Explore” page include the user’s history with the author (if they follow, viewed, or have liked their content previously), other users’ interactions with the author’s recent posts, the author’s profile, and what device the user is using. In comparison, on the “Reels” page (where videos are automatically put), the algorithm takes a more impersonal approach, instead basing the page on the author and the reel’s overall popularity on the site. Lastly, an article from LinkedIn explained how Tik Tok’s algorithm is uniquely efficient compared to other social media sites because of its “real-time adaptation”, which gives it the ability to adapt to changes in user preferences and behavior in real time.

I could not find any research to suggest that content creators are influenced to change their personality within their content in order to appease the AI algorithm. However, it should be noted that Instagram heavily promotes the “Reels” page in comparison to the “Explore” page, most likely due to the popularity of Tik Tok’s video format, and many anecdotal stories by Instagram creators have stated that they make much more of a profit when they create Reels than posts. However, it should be noted that many other content creators, specifically YouTubers, benefit from “teaching” other creators how to use the algorithm to their advantage. Many creators will use specific keywords in the caption or description of their post that they believe the algorithm will favor. Although this may seem like a smart decision choice, all of these companies have made the exact methods of their algorithm private, so there’s no foolproof effective method that will guarantee your content will become successful on their site.