From creators to users: How artificial intelligence is shaping social media feeds

When social media established its foothold as a means of communication, it transformed the way we interact and obtain information online. Now, more than two decades later, artificial intelligence (AI) is transforming social media.

While the use of AI in content generation has often got the spotlight, its wider applications and implications on social media creators and users are often hidden behind the charm of unique images and easy-to-produce text facilitated by the technology.

According to Statista, there are 5.17 billion social media users across the world in 2024, and this number is only expected to grow. Since these platforms have become an integral part of our day-to-day lives, influencing our thoughts, activities and other aspects of life, it is important to understand how technology is shaping their functioning. We speak to experts who provide insights on the use of AI in social media.

AI in social media algorithms
One of the main pillars of social media platforms is the algorithm that dictates their operation. Social media algorithms lend unique features to every platform and are also the link between content production and consumption. They analyse user behaviour and preferences to deliver personalised content. Most of these are powered by AI.

"Social media algorithms use AI to provide a plethora of features, starting with suggesting friends or accounts you would like to connect with to determining the kind of content that appears on a user's feed," explains Rashmi Jain, Associate Professor, Marketing and International Business, K J Somaiya Institute of Management.

"If a user engages or even spends a few extra seconds with a piece of content, the algorithms capture it and AI capabilities use the information to gauge the interest of the user in the content and predict more pieces of content in that area. AI is used to understand the users, their interests, their beliefs and preferences, and then serve the content or marketing messages (ads) that are most suited to the user (profile)," she adds.

AI in social media analytics
Algorithms depend on data analysis. From sentiment to trend analysis, from user behaviour to response analysis, AI is facilitating the comprehensive collection of data on social media platforms.

"AI-powered analytics on social media flourish on user data which includes time spent browsing, likes, shares, comments, new connections, reels watched, etc. All this data is analysed to enhance the experience of the social media platforms but can have grave implications," reveals Jain.

What could be the potential harms, we ask?

"It could lead to user surveillance, data breaches, profiling and discrimination," she responds.

"Take for instance, AI-based personalisation algorithms predict which show, product, or type of news the user will be most interested in, based on the user's past behaviour. The idea is to make the discovery of the desired fast and easy and to avoid the display of (presumably) uninterested things. This creates a filter bubble or a personalisation bubble subjecting users to similar kinds of content and messages. This gives rise to a larger social problem where people are subjected to content and posts similar to their opinions and beliefs, making those beliefs even stronger and robbing people of the opportunity to get exposed to newer perspectives," she elaborates.

"From a marketing perspective, excessive remarketing based on AI-powered analytics or too much bombardment to a niche audience may have a boomerang effect," she adds.

(Boomerang effect refers to a situation where the response to a message is opposite to what was intended.)

AI-powered algorithms and analytics streamline the user experience, but they also pose ethical dilemmas like privacy concerns and the creation of an echo chamber.

AI in social media content creation
The most common use of AI in social media is for content creation. From image generation to video production, generative AI tools like ChatGPT and Midjourney are helping content creators and brands to innovate. While not everyone may use AI-generated output, it serves as a good starting point.

Rishi Jain, a Chennai-based digital marketing and AI coach and co-founder of Digital Scholar, explains, "We utilize AI, specifically ChatGPT, to generate content and brainstorm ideas. Although the initial outputs from ChatGPT sometimes require adjustments, they provide a solid starting point. This approach prevents creative blocks and reduces the need for extensive research. One can then build upon the foundational ideas provided by ChatGPT."

"Additionally, tools like MidJourney can be used for image generation. Traditionally, explaining a visual concept to a designer could take hours. Now, we can input our ideas directly into the tool, create the images ourselves and then pass these on to the designers to refine them. Another application of AI is in video production. For instance, when we need to create videos showing different festivals across India on a map, we input the necessary details into AI tools. These tools generate the required visuals, which can then be incorporated into larger video projects," he adds.

In conclusion, generative AI is a transformative technology that is currently shaping content creation on social media.

Meta AI - A boon or bane?
A recent development in the use of AI in social media is the introduction of Meta AI - an AI assistant that one can interact with on WhatsApp, Instagram, Messenger and Facebook. Its popularity has soared high in India with Mark Zuckerberg revealing earlier this month that the country is the largest market for Meta AI.

Applications
"Meta AI is creating more user convenience. The users of Meta now need not leave the app to search for something, generate an image/ video, draft a message, get content recommendations, etc.," comments Rashmi.

Rishi states the following applications of Meta AI.

• One can get a lot of content creation ideas and recommendations.
• It can look into our conversations and help us out in terms of understanding the emotions of the person. For example, you can ask, "Can you tell me what is the tone of this particular conversation and what should I infer from this?" This helps with better responses and enables making the right decisions at the right time.
• The chatbot, particularly the conversational chatbot on WhatsApp, Instagram and Facebook, is going to be crucial for people with different skills. For example, individuals who do not know how to write or understand English could use this platform to converse. For instance, my 65-year-old mother, who does not understand or write English, could convey her thoughts in Hindi on these platforms, and it could automatically be translated into English.
• This has a lot of business applications as well. If a website has all English content or if someone's emails or messaging sequences are in English, they could be shifted and changed to the language of the particular user.

While there are novel applications of this AI-powered assistant, it doesn't come without a cost.

Downside
"Since Meta has access to the three largest messaging and social media platforms i.e., WhatsApp, Facebook and Instagram, they are very serious about the use of user data for personalisation, preventing certain news/content from reaching a few people (discrimination). Meta AI might not be sensitive to the type of content being served which may lead to the spreading of fake news/misinformation. Further, there are privacy concerns, generating misleading and fake content," states Rashmi.

"The disadvantages of this feature will be primarily over-dependence on AI, leading to a drastic decrease in our thinking capabilities. This could result in our kids and younger generation having difficulty making decisions and relying heavily on these tools. They will probably be poor at grammar and have a tough time remembering spellings because they get corrected within milliseconds by these platforms. There will be hardly any typing or creativity happening, leading to impaired creativity and decision-making skills in the long run," concludes Rishi.

Also Read: Art through the eyes of AI: Here's how artists and creative professionals are adopting AI in their work