Thomson Reuters and Legaltech News quoted Peter Geovanes, McGuireWoods’ chief innovation and AI officer, in recent news stories on the benefits and costs of integrating generative artificial intelligence in the legal profession.
Geovanes was a panelist at Thomson Reuters’ Generative AI & Emerging Technology Forum, which explored how the professional services sector can leverage the latest advancements in generative AI. Geovanes called generative AI a “mad scientist” that can supercharge the practice and business of law, according to Thomson Reuters’ Oct. 16, 2023, story.
In the practice of law, generative AI could be used to upload an opposing party’s motions to find inconsistencies in statements or quickly summarize transcripts, Geovanes said. On the business side, generative AI could sift through public news reports and data to predict emerging litigation trends, he added. But in both cases, he noted, human input is needed to point AI toward a specific outcome and deliver results for a firm and clients.
“I think that changes the whole dynamic of the law firm and the ability to provide value to those clients and prospects,” Geovanes said.
Legaltech News, an American Lawyer Media publication, quoted Geovanes in an Oct. 12, 2023, story about the cost of integrating generative AI in law firms. The cost will depend on firms’ use cases for the technology, the article noted. Models such as OpenAI’s GPT-4 tie their costs to the use of “tokens,” or pieces of words. The cost covers input token usage — the text uploaded into a language model — and output token usage — the generated response from the model.
Many large language model application programming interface (API) providers “have pricing structures that consider both input and output token usage, so it’s essential to be aware of the token limits and pricing details provided by the specific service you are using,” Geovanes told Legaltech News.
McGuireWoods recently announced that it will utilize CoCounsel, a first-of-its-kind AI legal assistant developed by Casetext and powered by OpenAI’s latest and most advanced large language model, GPT-4.