Redefining 'Ordinary Meaning': GenAI and Legal Language

One Judge's Intriguing Experiment

Can generative artificial intelligence help us understand what words mean in legal disputes? Published yesterday, a federal judge's unusual concurring opinion suggested it might be possible—and sparked a fascinating conversation about the future of legal interpretation.

A "Landscaping" Conundrum

The case itself involved an insurance dispute hinging on whether installing an in-ground trampoline qualified as "landscaping." Sounds simple, right? But as Judge Kevin Newsom of the Eleventh Circuit Court of Appeals discovered, pinpointing the "ordinary meaning" of even seemingly straightforward terms can be trickier than it seems.

Judge Newsom, a self-described "plain-language guy," dutifully consulted his dictionaries. Yet, the definitions he found felt incomplete, failing to fully capture the essence of how the word "landscaping" is used in everyday life. He even examined photos of the trampoline installation, his intuition telling him it didn't quite fit the bill. But he struggled to articulate *why*. As he noted, “Nothing in them really struck me as particularly 'landscaping'-y.”

From Dictionaries to Doubt...and an AI Assist

Enter ChatGPT. Out of frustration, Judge Newsom decided—almost as a joke—to ask the AI chatbot for its take on the meaning of "landscaping." To his surprise, ChatGPT delivered a reasoned definition that resonated with his own developing understanding of the term. "Perhaps in a fit of frustration, and most definitely on what can only be described as a lark, I said to one of my clerks, ‘I wonder what ChatGPT thinks about all this.’”

Intrigued, Judge Newsom ventured further. What did ChatGPT think about the trampoline? Could *that* be considered "landscaping"? The AI's answer—a confident "yes," backed by logical explanations—pushed Judge Newsom to consider the potential of this technology in a whole new light. ChatGPT responded: "Yes, installing an in-ground trampoline can be considered a part of landscaping... It’s a deliberate change to the outdoor environment, often aimed at enhancing the overall landscape and usability of the area."

The Promise (and Perils) of AI-Powered Interpretation

Judge Newsom suggests that the appeal lies in AI's ability to tap into something fundamental about legal interpretation: the importance of understanding how words are *actually* used by ordinary people. Large language models (LLMs), trained on vast datasets of online text, offer a potential window into this everyday usage. As he puts it, “The ordinary-meaning rule... has always emphasized 'common language', 'common speech', and 'common parlance'—in short, as I’ve explained it elsewhere, 'how people talk.'”

Moreover, unlike static dictionary definitions, AI can analyze language in context, recognizing nuances and shades of meaning that traditional methods might miss. This "contextual intelligence" is key to unlocking the intended meaning of legal texts. Judge Newsom explains, "The combination of the massive datasets used for training and this cutting-edge ‘mathematization’ of language enables LLMs to absorb and assess the use of terminology in context.”

However, Judge Newsom doesn't shy away from the potential pitfalls. He acknowledges the risk of AI "hallucinations"—generating inaccurate or misleading information—underscoring the need for careful scrutiny of their outputs. “First, the elephant in the room: What about LLMs’ now-infamous 'hallucinations'?" He also raises concerns about potential biases within training data, emphasizing the importance of transparency and inclusivity: “The absence of offline usages from the training pool—and in particular, the implications for underrepresented populations—strikes me as a sufficiently serious concern.”

Kudos to Judge Newsom

Judge Newsom deserves kudos for his courage and foresight. Including the prompts and outputs from ChatGPT as a published appendix in his opinion is not only transparent but also very useful. Moreover, his willingness to share this view, which challenges conventional wisdom, is commendable and reflects an openness to innovation that is crucial for the legal field. As Judge Newsom rightly notes, the experiment with ChatGPT "no longer strikes me as ridiculous" and indeed "might have something useful to say about the common, everyday meaning of the words and phrases used in legal texts."

Proceed with Caution...and Curiosity

So, can ChatGPT replace judges and lawyers? Not quite. But Judge Newsom's experiment offers a compelling glimpse into a future where AI could play a valuable role in legal interpretation. As he aptly puts it, LLMs should be viewed as one tool among many, offering additional insights that can help us better understand the law and ensure its fair and just application. “It seems to me scarcely debatable that the LLMs’ training data are at the very least relevant to the ordinary-meaning analysis.”

The conversation is just beginning, and Judge Newsom is among the first to highlight the true potential of AI in the legal field. While many in legal circles have yet to fully grasp how well this technology fits various legal tasks and processes, Judge Newsom's concurring opinion is just the beginning. This technology is unusually well suited for a wide range of legal interpretation, analysis, and other key functions. As AI continues to develop, the intersection of artificial intelligence and the law promises to be a transformative space to watch.