The advent of AI, particularly tools like ChatGPT, is fundamentally altering the landscape of language and communication.

One intriguing phenomenon is the noticeable increase in the usage of certain words, like “delve,” in academic and scientific writing since the widespread adoption of ChatGPT in 2023.

This shift highlights both the growing influence of AI on language and the complexities it introduces.

The Rise of “Delve” in Scientific Writing

As AI becomes more integrated into daily tasks, it is also subtly reshaping language. For instance, the word “delve” has surged in popularity in scientific literature. According to data from OpenAlex, mentions of “delve” in scientific papers spiked significantly in 2023, correlating with the increased use of ChatGPT.

Researchers, like Dr Jeremy Nguyen from Swinburne University, have noted this trend, observing that “delve” appeared in more than 2,500 PubMed papers in 2023 alone—a fourfold increase from the previous year.

This pattern raises questions about the extent to which AI is influencing language. ChatGPT, now known for its frequent use of the word “delve,” may be indirectly encouraging its adoption by researchers who use AI to assist in drafting papers.

However, while the data suggests a correlation, it does not definitively prove causation. Nguyen himself acknowledges that the word’s rise could also be coincidental, with “delve” simply coming into vogue at the same time as AI tools like ChatGPT became prevalent.

The Feedback Loop: AI, Training Data, and Language Evolution

A key aspect of this phenomenon is how AI models like ChatGPT are trained. These models are developed using vast amounts of public data, including the very texts they help to generate. When a word like “delve” starts appearing more frequently in academic papers and other authoritative sources, it becomes part of the dataset on which AI models are trained.

As a result, the AI learns that “delve” is a commonly used and authoritative word, reinforcing its use in future AI-generated content. This creates a feedback loop where the AI continues to use “delve” more frequently, further embedding it in the language models.

As more researchers and writers use AI tools like ChatGPT, the linguistic patterns of these tools become more widespread. In turn, these patterns influence the training data for future AI models, perpetuating certain words and phrases while potentially sidelining others. This feedback loop could lead to a homogenisation of language, where certain terms dominate simply because they are embedded in AI training data.

The Broader Implications of AI on Language

The increase in the use of “delve” is just one example of how AI is impacting language. As AI tools are employed to automate mundane tasks, they also shape the way we write and communicate. The rhetorical flourishes and stylistic choices made by AI can make texts appear more formal or authoritative, as seen with “delve.”

However, these changes are not without controversy. Some worry that over-reliance on AI could lead to homogenized writing, where certain words and phrases become overused, potentially reducing the richness and diversity of language.

Moreover, the challenge of detecting AI-generated content complicates the situation. While some believe that identifying AI-written text could be as simple as spotting overused words like “delve,” this approach is far from foolproof. Tools designed to detect AI-generated content are still unreliable, and the nuances of language mean that any given word’s presence does not necessarily indicate AI involvement.

Looking Forward: The Future of Language in the Age of AI

The ongoing changes in language brought about by AI highlight a crucial point: the tasks we consider intellectual or creative might be more formulaic than we realize. As AI continues to evolve, it will likely take on more of these tasks, prompting a re-evaluation of what truly requires human intellect and creativity. Researchers like Nguyen are already openly acknowledging their use of AI in academic writing, suggesting a future where collaboration with AI becomes standard practice.

In conclusion, while the rise of words like “delve” in scientific literature underscores the growing influence of AI on language, it also invites us to reflect on the broader implications. As AI continues to shape the way we communicate, it is essential to consider how these tools can be used to enhance, rather than diminish, the richness of human expression. The feedback loop created by AI training on public data presents both opportunities and challenges for the future of language, demanding thoughtful consideration as we move forward in this potential AI linguistic era.