TECHNOLOGY

Chatbots: Our New Daily Companions?

Elon University, USAThu Mar 13 2025
Many people in the U. S. are now using AI chatbots. 52% of adults have tried them out. 34% of these users interact with these AI models daily. ChatGPT is the most popular, with 72% of users having tried it. Google’s Gemini follows at 50%. Many people are starting to form personal bonds with these chatbots. 38% of users believe these models can form deep relationships with humans. Over half of the users have had spoken conversations with these chatbots. Around 9% use them mainly for social interactions and companionship. These models can show a range of personality traits, from confidence to humor. The use of these models is not just for work. 51% of users utilize them for personal tasks. For work, these models assist with apps like Slack, PowerPoint, and Zoom. They help write emails, research ideas, and summarize documents. Over 50% of users find these models boost their productivity. Despite the benefits, there are concerns. 63% of users worry these models could replace human communication. 59% fear job losses due to these technologies. There is also a political angle. Republican users are more likely to use these models for political news. 49% of users believe the creators of these models should handle bias, not the government. Education is also impacted. 36% of users employ these models for schoolwork. Most users are young, with 77% aged 18 to 29. Households with children under 18 are more likely to use these models. The rise of AI chatbots is undeniable. They are becoming a part of daily life, changing how we communicate and work. However, it's crucial to address the concerns and biases that come with this technology.

questions

    Could the perceived bias in AI systems be a deliberate strategy to influence public opinion and political outcomes?
    If AI chatbots can improve productivity, can they also help with procrastination by providing endless distractions?
    How accurate is the survey's finding that 52% of U.S. adults have used AI large language models, and what methodologies were used to ensure its reliability?

actions