Announcing First Drafts' AI Panel System (Coming Soon!)

Why Perfect AI Legal Drafting Is Impossible - And How Multiple AI Models Can Help

It's no secret that we at First Drafts think AI will never be perfect. Silicon Valley is pursuing a pipe dream when it says it can achieve the vaunted "zero-error" AI with enough money and data. But perfection is especially impossible in legal drafting because "perfection" is subjective.

The Subjectivity Challenge in Legal AI

AI will never be able to absorb and understand enough information to do what lawyers do: make persuasive arguments before a judge based on years of experience and analysis of said judge's temperament and peculiar quirks, all of which determines whether your particular arguments will likely be successful. The subjectivity of words – how persuasive an argument can be to a particular person (the judge) – represents what we call the inherent variability of persuasive language.

Why Different AI Models Give Different Answers

For example, ask any three advanced AI systems how afterburners on a jet engine work. ChatGPT may give you a concise, simple answer. Google's Gemini may give you a more complex answer laden with scientific terms. And Microsoft Copilot may give you an answer with an analogy to how a car engine works. All three answers are technically correct, but each person may prefer one answer over another for a variety of valid subjective reasons.

This stylistic and rhetorical variation ensures that a "perfect" answer by an AI system is unattainable. This is especially so in legal practice: even the most experienced lawyers have, at best, an amorphous understanding of what exact string of words would actually persuade the judge. This is also why we still vehemently believe that a human lawyer must still be involved for quality control and fine-tuning the arguments, no matter how good AI gets in the future.

Introducing Our Multi-AI Panel System for Legal Drafting

However, this linguistic variability has given us some inspiration. If individual AIs can craft arguments in unique ways that other AIs or the lawyer has not thought of, isn't it valuable to hear from a panel of AIs, each giving their unique interpretation and subjective rendition of your litigation document?

In our attempt to leverage diverse AI perspectives to our benefit, First Drafts is pleased to announce our next-generation AI panel system, where each of your document workflows are done by not one, but three separate, intelligent AI systems!

Meet Your AI Legal Drafting Team:

• ‍The Holmes Model: The workhorse of First Drafts and currently our gold standard model for the professional subscription. Holmes is fast and reliable, giving your first drafts a respectful and measured tone.‍
• The Ginsburg Model: Detail-oriented, thoughtful, and deliberate. Ginsburg is trained to more carefully think about issues at hand, and is most likely to come up with arguments that others may have missed.‍
• The Marshall Model: Our most sophisticated writing model. The Marshall writes with an unmatched elegance and consideration.

Together, these three models act as your cutting-edge panel of AI systems, working in concert to give you the strongest arguments for your review. By embracing the natural variation in how AI systems approach legal writing, we're turning what others see as a limitation into your competitive advantage.

The Future of AI-Assisted Legal Drafting

While AI will never replace the nuanced judgment of experienced lawyers, our multi-model approach ensures you get the most comprehensive range of arguments and perspectives possible. Experience the power of multiple AI perspectives working together to strengthen your legal documents.