AI-Enhanced Venture Capital: Crafting Superior Investment Theses with ChatGPT

Introduction: Beyond Email Drafting - Leveraging AI for Decision-Making

At this point everyone knows about ChatGPT and the impact it is having on the knowledge economy. So I want to scream when I hear people talking about using it to “draft an email”. How inefficient - if you need ChatGPT to write emails, you should be writing shorter emails! ChatGPT should be used to help you make better decisions. And what is VC, but being paid to make good decisions, quickly and efficiently?

Investment Thesis: The Key Tool for Better Investing Decisions

The investment thesis is one of the key tools VCs have to make good investing decisions. Investment theses are a series of statements about an investment, that if true, means it’s a good investment. I use them to: 1) think through an investment, 2) structure my diligence, and 3) persuade my fellow investors to vote to invest.

Harnessing AI for Investment Theses: The Basics

Here’s how I use chatGPT to write a better investment thesis. If you’re familiar with “prompts”, you’ll know that providing some background info or asking chatGPT leading questions will help it provide more specific output. You can use the prompts I suggest, but that’s not the point of the article. I want you to be more creative with how you use chatGPT, not give you a step-by-step list of prompts. I use this investment thesis approach all the time with startups I invest in. But to share the output, I’m considering an investment in Mistral AI, the open source AI company.

The Initial Interaction: Establishing a Foundation with AI

First, I ask chatGPT what is knows about VC investment theses. It gives a good summary - it will cover things like market, the startup itself, what to consider, and why you would use it. So I’ve confirmed that we’re both aligned and it knows what we’re driving at. Next, I ask it to help me draft a comprehensive investment thesis by asking me questions. I specify the requirements I have for my investment thesis (proveable statements, convincing case to invest etc.).

For me, the point on questions is absolutely crucial. I’m using chatGPT to make better decisions. The process of answering the questions is what drives better decisions. It often highlights something I have missed and even the tone of the question can expose my own biases. (Pro tip - by answering questions you can stay privacy compliant - just redact or don’t answer anything you don’t want to.) The first set of questions chatGPT provides are good but generic - I probably could have pulled this off Google.

These are OK, but generic, questions to ask when crafting an investment thesis

Tailoring the Process: Getting Specific with ChatGPT

I then feed chatGPT a few sentences background on the company and ask for more tailored questions. For startups I’m looking at I would just give it a few pieces of information on the startup. This is often my own totally biased description of the startup, e.g. huge market, does this, love the CEO - she has performed brain surgery on sharks while saving the planet etc etc. Again, I find the bias helpful, because now chatGPT will ask me questions around what I like or dislike too about the company. You can ask it to be provocative too - in this case it will probably ask if the CEO has any entrepreneurial experience! For this article, I just asked chatGPT what it knows about Mistral and then fed that to it.

chatGPT uses the background questions to ask more specific questions

Delving Deeper: The Iterative Questioning Process

Now it gets interesting, I’ve got a much more tailored list of questions to dig into on Mistral. It’s much more specific on the info I fed it about Mistral and remains generic if no info was provided. Now, I just answer these questions. Some of them I can answer straight away. But those I need to research are great - it shows chatGPT is forcing me to think differently. Again, for this article I just fed the questions to chatGPT to the research but obviously I normally do the research on the startup myself. Again, no need to share any nuclear secrets here.

Here are the answers chatGPT found to the first few questions for Mistral.

Iterative Refinement: Perfecting the Investment Thesis

This, I think is the mistake that most people make with chatGPT - they use the first draft it produces. That’s not the point when you’re using it to make decisions. Go back and forth and ask it to dig in or refine something. Finally, I take it offline and write up the final draft in my own words. But use the opportunity to examine the tone and phrasing chatGPT is using. Often it will overemphasise something I disagree with, but thinking this through helps me query my own assumptions. Occasionally, I’ll ask it why it thinks something I disagree with and the answer can spark some thinking on my part.

A good first draft of the investment thesis for Mistral

The Final Step: Transforming Thesis into Diligence Scope

Now for the magic. Once I have my final investment thesis ready, I feed it back into chatGPT and ask it to draft a diligence scope. This scope is a list of key questions I must answer in diligence, tailored to my investment thesis. For me, this is a huge time saver.

chatGPT turned the investment thesis into a diligence scope. The better the thesis, the better the scope.

Conclusion: AI as a Personalized Sounding Board in VC

In harnessing ChatGPT's capability to challenge and refine my thinking, I've transformed it into an invaluable tool for venture capital decision-making. For me, this is not just about saving time or outsourcing tasks. It's about leveraging AI to enhance the depth and quality of my investment strategies. By engaging with my personalised sounding board, I've sharpened my insights, challenged my preconceptions, and uncovered new angles in my investment approach. This iterative, thought-provoking process, empowered by AI, is not just a step forward. it's a leap towards smarter, more informed decision-making in venture capital.

#VentureCapital #AIinVC #ChatGPT #InvestmentThesis #TechInvestment #VCInsights #DeepTech #StartupFunding #InnovativeInvesting