#43 Chat GPT o1-Preview vs. Signature Pages
This week, I test whether OpenAI’s newest model, o1-Preview, can create signature pages for M&A agreements.
AI will change the world, but how will it change M&A? I want to focus on AI’s impact on M&A in this newsletter. I am not an expert on either M&A or AI, but I want to learn about both topics and how they intersect. I thought there might be others in my situation (or people who are experts in one field or the other) who would find information on M&A and AI helpful in their careers, so I created this newsletter to track and share what I learn.
AI Generated Signature Pages
A few weeks ago I created a GPT tool on ChatGPT that creates signature pages. After some fine-tuning, the model performed very well and produced signature pages ready for use in an M&A agreement. I wanted to test whether Open AI’s newest model, o1-preview, can create signature pages.
Junior lawyers typically create signature pages, so it is not super complex or high-level. That said, there is a particular format that signature pages take and I wanted to see if ChatGPT could pick up on it. To test it, I decided to ask it to make a basic signature page (one entity) and an advanced signature page (entity within an entity). I also changed the level of detail in my prompts to see if the model performed with little context.
Basic signature page request without fine-tuning
I started at a very basic level, with no specialized prompt.
I asked it why the signatories’ titles were not underneath their names, as (I think) is custom for US agreements. Here is what it said:
I apologize for the confusion earlier.
You're absolutely correct about the placement of the signatories' titles. In typical US-style agreements, the title should appear directly beneath the signatory's name. Here is the revised signature page.
It then properly formatted the page. I was not super impressed with its performance, as this seems like something it should get down immediately.
Advanced signature page request without fine-tuning
Next, I wanted to up the complexity to further test o1-preview. Here is my prompt for an advanced signature pages with no fine-tuning:
Prompt: Please regenerate the template in the form necessary for the following parties to sign. The first party is a limited partnership, whose general partner is a limited partnership, whose general partner is an LLC. The second party is a corporation. Please give me a signature page formatted so that each party can sign the agreement and it be legally enforceable against the entities.
Outcome:
I think that is close to correct. I don’t think it is great form to put the dates beneath the signatures, but I wouldn’t say that it is wrong. I am not sure whether the indentions of the first three consecutive lines are correct, it may be up to stylistic preference.
Advanced signature page with fine-tuning
Now I will input my prompt from the previous signature page generator tool. The prompt is long, so I won’t share it here (reach out if you’re interested). Safe to say that the prompt adequately gives the model everything it needs to know about signature pages. The prompt tells the model to ask questions about the parties. Here is my description of the signature page and the output:
I would say that is very good. It still has the indention on the first signature block, which again is possibly a stylistic choice for some lawyers (I did not see it in any examples).
I asked it to fill in the blanks with party names and titles:
Here is what it gave me:
IN WITNESS WHEREOF, the undersigned have executed this agreement as of the date first written above.
High Growth LP
By: General Partner LLC, its general partner
By:______________________________________________
Name: John Smith
Title: Member
ABC Corp
By:______________________________________________
Name: Jane Doe
Title: Chief Executive Officer
I would say that is pretty good!
I’m not sure how much demand there is for AI to take over this task, but I think it is fair to say that AI is now fully capable of doing so. I think most people jump to huge conclusions about AI’s capabilities. “It’s going to replace us” or “What will my job look like if I do not have X task to do?” are now common refrains. I like to think about AI from a bottom-up approach, using it to do smaller, more concrete tasks and then systematically working up the complexity. It’s fair to say that this is the bottom of the barrel when it comes to M&A tasks, but it’s good to know that AI can handle it.
About me
My name is Parker Lawter, and I am a law student pursuing a career as an M&A lawyer. I am in my last semester of law school, and with some extra time on my hands, I decided to create this newsletter. I hope it is informative and helpful to anyone who reads it! I am not an expert at either M&A or AI, but I am actively pursuing knowledge in both areas, and this newsletter is a part of that pursuit. I hope you’ll join me!
Follow me on LinkedIn: www.linkedin.com/in/parker-w-lawter-58a6a41b
All views expressed are my own!