How AI is Changing the Way We Talk
The Observation
Sam Altman, a prominent figure in tech, was among the first to notice that online discussions were becoming eerily perfect, almost as if they were written by a machine. He attributes this phenomenon to several factors:
- Mimicry of AI Language: Real people are adopting the way AI communicates.
- Homogenization of Online Communities: Online communities tend to develop a uniform way of speaking.
- Social Media's Influence: Platforms push users to say things that garner more attention.
- Potential Use of Fake Accounts: Some companies might be using bots to enhance their products' image.
Research Findings
Researchers at the Max Planck Institute in Berlin analyzed millions of texts and videos. They discovered that after the release of ChatGPT, people started using more words like "delve" and "examine"—terms frequently used by AI. This suggests that AI is influencing human language.
The Problem
A study from the University of California-Berkeley revealed that ChatGPT favors Standard American English. This can be frustrating for non-American users and can make online conversations feel artificial.
Fighting the Trend
Vaikunthan Rajaratnam, a nerve surgeon, attempted to make ChatGPT sound like him by providing specific instructions and refining it over time. He found that while ChatGPT can make communication clearer and more structured, it can also make people sound less authentic.
The Human Advantage
Mark Cuban, a well-known investor, pointed out that AI has a significant weakness: it can't admit when it doesn't know something. This is a trait humans excel at, suggesting that AI might not be able to replace human communication entirely.