Keep pulling the thread on Sean Fitzpatrick.
Sean Fitzpatrick states that hyperscalers like Microsoft, Amazon, and Google are each investing close to $100 billion in foundational AI models.
The cost for LexisNexis to process one million tokens has dropped from approximately $20 two years ago to about $0.10 today.
The LexisNexis AI system is grounded on a proprietary, curated collection of 160 billion documents and records.
LexisNexis's vision is to provide every attorney with a personalized AI assistant that understands their specific practice area, jurisdiction, and prior work product.
LexisNexis employs a model-agnostic, multi-model approach, using different LLMs like an OpenAI model for research and Anthropic's Claude 3 Opus for drafting based on their specific strengths.
LexisNexis's AI system uses an agentic architecture where a "planning agent" receives a user query and allocates sub-tasks to other specialized agents.
LexisNexis's AI products are the fastest-growing in the company's history and are contributing to profit growth.
Judge John Bush has suggested that AI could be used to review historical texts and compile statistics on word meanings to facilitate originalist legal interpretation.
LexisNexis's AI tool, Protege, is designed to assist lawyers with drafting legal documents for court submission, moving beyond traditional legal research.
LexisNexis launched its integrated product suite, Lexus Plus, in 2020 and its AI-enhanced version, Lexus Plus AI, in 2023.
LexisNexis's AI platform includes a "citator agent" designed to verify that legal citations are not fabricated and represent current, valid law.
According to LexisNexis surveys, two-thirds of attorneys are using AI in their work, and the company believes the actual number is likely higher.