As AI continues to shape social media, a growing concern emerges: can machines predict and manipulate our likes to influence our behavior? The line between human preference and AI-driven optimization is blurring, threatening the very fabric of online engagement.
The Future of the Like Button in AI-Powered Social Media
Liking features on social media can provide troves of data about human behavior to AI models, but as AI gets smarter, will it be able to know users’ preferences before they do?
Artificial intelligence (AI) refers to the development of computer systems that can perform tasks typically requiring human intelligence, such as learning, problem-solving, and decision-making.
AI has a rich history dating back to the 1950s with the Dartmouth Summer Research Project on Artificial Intelligence.
Since then, significant advancements have been made in machine learning, natural language processing, and deep learning.
Today, AI is applied in various industries, including healthcare, finance, and transportation.
Harnessing the Power of Liking Data
Max Levchin, PayPal cofounder and Affirm CEO, sees a new role for liking data in training AI to arrive at conclusions more in line with human decisionmakers. However, this approach also raises concerns about the optimization path often led by AI systems, which can result in very different outcomes than humans exercising human judgment.
The Problem of Reinforcement Learning from Human Feedback
The current method of using reinforcement learning from human feedback (RLHF) is costly and relies on hiring human supervisors and annotators to enter feedback. Levchin believes that the accumulated resource that Facebook owns – its liking data – could be a godsend for developers wanting to train intelligent agents on human preference data.
AI’s Impact on Social Media Preferences
While Levchin envisions AI learning from human preferences through the like button, AI is already changing the way these preferences are shaped in the first place. Social media platforms are using AI not just to analyze likes but also to predict them, potentially rendering the button obsolete.
The Future of the Like Button
Steve Chen, YouTube co-founder, notes that AI may eventually be able to tell algorithms with 100% accuracy what users want to watch next based on viewing and sharing patterns. However, he also acknowledges that the like button may still be needed for handling sharp or temporary changes in viewing needs.
The Rise of AI-Generated Content
AI is increasingly being used to generate content that is subject to people’s emotional responses. This raises questions about the original purpose of the like button – to motivate more users to generate content – and whether platforms will remain successful without human users creating posts.
The Authenticity Problem
The use of AI to refine entertainment content also poses a risk of authenticity. For example, an incident during the 2024 Super Bowl halftime show highlighted the problem of AI-generated content being seamlessly corrected without notification: ‘the problem of AI-generated content being seamlessly corrected without notification’.
The Threat of AI-Generated Influencers and Bots
AI-driven influencers and bots are infiltrating online communities and gaining followers on social media platforms. This raises concerns about the future of the like economy, as likes may no longer come from real people, and content may not be created by them.
The influencer marketing industry has seen a significant shift with the emergence of AI-driven influencers.
These digital entities use artificial intelligence to create and curate content, often mimicking human behavior.
According to a study, 60% of marketers believe AI-powered influencers will become more popular in the next two years.
They offer brands greater reach and engagement, as well as reduced costs compared to traditional influencer marketing.
A Future of Virtual Engagement
The use of AI to create virtual personas, such as CarynAI, a chatbot that mimics a real-life online influencer, is changing the landscape of online engagement. This raises questions about the need for transparency in identifying who is behind a popular post and whether we want to retain the ability to dissemble.
The Need for Transparency
As AI-generated content becomes more prevalent, there will be a growing need for tools that provide more transparency and assurance about whether a like is attached to a real person or just a realistic bot. Different platforms may develop these tools at varying degrees of effectiveness: ‘whether a like is attached to a real person or just a realistic bot’
AI-generated content has revolutionized the way we create and consume digital media.
With advancements in natural language processing and machine learning, algorithms can now produce human-like text, images, and videos.
According to a report by Grand View Research, the global AI-generated content market is projected to reach $6.4 billion by 2028.
This growth is driven by increased demand for personalized content, efficient content creation, and reduced production costs.
- wired.com | AI Is Using Your Likes to Get Inside Your Head
