AI's Growing Influence: How ChatGPT Is Changing Our Language and Communication Styles

In the 18 months since the launch of ChatGPT, the pervasive influence of artificial intelligence (AI) on human language has become evident. Researchers from the Max Planck Institute for Human Development, analyzing nearly 280,000 YouTube videos from academic channels, found a significant uptick in the use of certain words. Speakers now favor terms like "meticulous," "delve," "realm," and "adept" up to 51 percent more frequently than in the three years prior. These changes align with ChatGPT's preferred vocabulary, as indicated by a previous study comparing 10,000 human- and AI-edited texts. The transformation extends beyond mere vocabulary. There is evidence suggesting that AI is altering our tone and structure of speech, making it more formal and less emotionally expressive. The study's lead author, Hiromu Yakuru, noted that "delve" has become a linguistic watermark, marking the presence of AI in academic discourse. However, this shift is not just academic; it affects everyday communication. For instance, researchers at Cornell University observed that smart replies in chat applications tend to increase positivity and cooperation. Yet, the same study found that suspicion of AI involvement can decrease perceived collaboration and increase demands, highlighting a paradox where AI simultaneously improves and undermines communication. Mor Naaman, a professor of Information Science at Cornell Tech, identifies three critical levels of human signals that are endangered by AI's influence: 1. Basic Humanity Signals: These are cues of authenticity and vulnerability, such as personal stories or expressions of emotion, which affirm our humanness. 2. Attention and Effort Signals: These convey that the speaker put genuine thought into their words, like crafting a unique response instead of defaulting to a template. 3. Ability Signals: These showcase the speaker's true capabilities, such as humor and competence, through original and nuanced language. Naaman worries about the loss of these signals as we increasingly rely on AI to mediate our communication. For example, AI-generated messages on dating platforms can sound eerily similar, robbing individuals of the chance to display their unique personalities. This erosion of personal expression and authenticity could lead to a situation where only face-to-face interactions are trusted, eroding confidence in digital and video communications. Moreover, AI's impact is not confined to standardizing language; it also perpetuates biases and inaccuracies. A study from the University of California, Berkeley, revealed that AI responses often contain stereotypes or exaggerated dialects when asked to mimic languages other than Standard American English. For instance, when prompted with Singlish (a variety of English spoken in Singapore), ChatGPT sometimes produces exaggerated and inauthentic versions, leading to discomfort and disengagement. This pattern of AI flattening regional dialects and cultural nuances not only undermines linguistic diversity but also fosters mistrust and misconceptions about different communities. As AI continues to shape our communication, it raises questions about the balance between standardization and authenticity. On one hand, AI can facilitate efficiency and professionalism in contexts like email writing and formal presentations. On the other hand, it risks homogenizing personal and emotional conversations, stripping them of the verbal stumbles and regional idioms that build trust and connection. The tensions at play include: - Backlash Against Homogenization: Early resistance indicates that people may actively avoid sounding too much like AI, as seen with academics avoiding the term "delve." - Advancements in AI Personalization: Future AI models might become more expressive and tailored, reducing the current issue of a uniform AI voice. - Loss of Conscious Control: The deepest concern is the potential for AI to influence our thinking, leading to a reliance on pre-scripted thoughts and expressions rather than our own unique ideas. The future of AI-mediated communication is not set in stone. People are beginning to recognize and push back against overt AI influence, and technology is evolving to better capture human diversity. The challenge lies in maintaining a conscious role in how we communicate, ensuring that our speech and writing retain the personal touches and emotional depth that define human interaction. Ultimately, the key is to preserve space for the imperfections and quirks that make our communication truly human. Industry experts agree that while AI will undoubtedly continue to shape our language, the real question is whether we will proactively manage its impact to protect our individuality and the richness of human communication.