To my fellow Legal friends: if you’re NOT exploring use cases for GenAI (like ChatGPT), you should be FOMO’ing hard right now!
In this post I will showcase some extremely simple legal use cases that I created in under 20 minutes each, tinkering with some of the latest features of ChatGPT: Code Interpreter and Vision. You can find live demo’s of some of these on the website of my Law & Ops community. Screenshots of the prompts used are included in the attached PDF and on the website.
1. NDA Generator: turn a handwritten note with concept art into a functioning legal tech application in minutes, requiring zero coding skills…
2. Patent Infringement & Validity Analysis: analyze pictures of infringing products or images from prior art and assess them against patent claims, automatically generating claim charts…
3. Trademark Infringement Analysis: compare pictures of similarly looking goods and perform a trademark infringement analysis…
4. Antitrust (Economic) Analysis: interpret economic data/graphs and derive arguments/conclusions relevant to antitrust assessments…
5. ESG QuickScan: create a ESG QuickScan app/online feature using just natural language prompts (no coding)…
These examples are not without their flaws. You will find that ChatGPT’s performance is inconsistent across tasks, with its quality heavily influenced by the prompt provided. For instance, in the trademark use case, its visual interpretation skills were noticeably suboptimal. In the antitrust example, ChatGPT failed to identify the ‘Cellophane Fallacy’ in the data on its own, which, while not anticipated, would have been cool.
While these use cases aren’t deployment-ready, they are intended to serve as a source of inspiration. To me, they underscore the potential to automate many tasks central to our roles as lawyers and legal counsels. Sure, we need the right controls and legal & ethical frameworks for deploying this technology responsibly in our practice. But coming up with rules is easy for lawyers. As traditionalists ( or rather: 🦕 ), the true challenge lies in shifting our perspective to genuinely embrace these tools. Such a transformation demands a new mindset and creativity that may not come natural to us.
I hope these basic examples can serve as a trigger for some to get into action and try for themselves. After all, the best way to learn is through experience and experimentation.
And if you believe I’m ahead of the game, think again: we’re all behind!
I’m eager to know your perspectives. If you have other interesting use cases (we’re already exploring many more!), please share!
Leave a Reply