Learn the Math Before You Touch the Calculator

Most of us grew up with math teachers who wouldn't let us use a calculator until we could show our work on paper. We hated them for it... Years later, the kids who'd put in the hand work were the ones who could tell when the calculator was spitting out nonsense, or better yet even know what needed to be typed into it to get it to spit out an answer. But those of us who were forced to do the work had an idea of what the answer was supposed to look like before we hit the equals button.

New angels using AI tools right now are mostly reaching for the calculator before they've shown their work. They're trusting whatever the AI spits out cause they don't have the judgment to know when it's wrong. That's gonna cost them and that's bad, but what's worse is that they don't even know enough to know it's gonna cost them. 

The math is the slow boring work that builds judgment. You have to put the reps in, and there's no AI shortcut that gets you there faster.

Reading bad pitch decks teaches you to spot the good ones. The first 50 decks you read all look the same; good team, big market, smart-sounding problem solve. After a few hundred you start seeing through the templates and noticing the small things; founders hiding something in how they word their traction slide, market sizing slides doing magic tricks with numbers. You can't get there by having AI summarize decks for you, because you’ve never seen the patterns yourself. And thats what this game is all about.. PATTERN MATCHING.

Same with founder intro calls. For your first few calls, you're mostly just trying not to feel or look dumb as you try to keep up..lol. You’re not sure what to ask, and you’ll be tempted to buy into whatever they’re saying. But after you’ve gotten your feet wet with about 30-40 calls you may start to hit your stride and get something asking the dumb question. You also start noticing things, little signs and tells you never saw before. Things like the founder who subtly deflects the harder questions, or the one who seems to know his numbers, but can't answer an adhoc numbers question. The ability to catch those reads are the edge in early-stage investing, and you can't outsource the reps that build them.

This is the way. It's slow, repetitive, and looks like wasted time at first. It's the part that makes everything else you do as an angel investor work. AI doesn’t get you there.

Once you've done the reps, AI tools become an accelerator. The work itself doesn't change; but the time it takes to do the work decreases a lot. I'm nine years in at this point, and the work I do on a deal today looks pretty similar to what I did three years ago. It just takes about a third of the time, and its a much crisper process.

Founder background checks used to be a two-hour exercise for me. Now I can pull a profile in a few minutes and it's better than what I used to do manually. I know what I'm looking for in that profile cause I've been reading founder backgrounds for years. When the AI summary glosses over a red flag or misses a pattern, I can catch it. A new angel using the same tool wouldn't.

For market maps, AI can give me a competitive landscape (the breakdown of who's playing in a market and how they compare with each other) in 15 minutes that used to take hours. I can dig into it faster cause I already know what a market map should show me and what's missing from it. Without that base knowledge, the map just looks like a list of companies but you still don't know how to tell which ones are the better ones.

Document review has been the biggest unlock for me. NotebookLM lets me upload a full data room (everything a founder shares with investors for review) and ask specific questions with page references. I can get through cap tables (previous investors what of the company and how much equity each has), term sheets (the deal terms in writing), and financial models in a fraction of the time it used to take. But I know which questions to ask cause I've reviewed hundreds of these documents already. A new angel asking AI to "review this term sheet" gets a generic summary that misses the things that matter for their specific deal.

Every tool fits the same pattern. AI accelerates the work I already know how to do; it doesn't put the years in for you.

AI doesn't give you judgment. It magnifies whatever judgment you built, and if you haven't built any or it sucks.. Well, you’re gonna have some sad times ahead as an investor.. So a smart investor with smart use of AI tools can be a much more efficient..not a better investor. Because they cover more ground than they used to, but it doesn't necessarily mean they cover that ground better. A new angel who hasn't built that judgment yet ends up with AI helping them make bad decisions faster.

The dangerous part is the speed. A few years ago a new angel might have looked at 20 deals and made two investments. Now with AI tools that same new angel can look at 100 deals and make 10 investments in the same window. If they don't know what they're doing, they're just digging themselves in a hole much faster.

I had a conversation with a newer investor that exclusively uses AI to generate first look deal summaries. When I asked him to tell me what he liked about the companies he was thinking of investing in, he had to dig through his notes to answer. He couldn't tell me from memory what he liked about them because if hadn;t actually interacted with them or their data enough to have a real feel. Thats fucking madness if you ask me. He had a stack of AI-generated info but now real gut feel or judgements about the companies.

If you're early in your angel investing journey, do the slow boring work. Read decks without AI summaries, sit through founder calls and take your own notes, not rely on AI note takers, dissect market maps yourself if it takes more than than you think you have, form your own opinion on a deal before you check what AI thinks. In fact, I would suggest not even touching AI for this until you’ve done at least 5 angel investments. Because as a newbie, you don't even know what questions to ask about AI or what to look for when you’re looking at AI output. 

Trust me when I say that time isn't wasted; that's the showing your work part.. You're building the foundation that allows AI to become a useful tool later on.

Once you've put the reps in, lean on AI for the parts of the work that aren't building skill or judgement. You can move faster, cover more deals in less time. But check the AI's work against your own judgment. When the model tells you a market looks great and your gut says something's off, trust your gut, cause that gut feeling is what dozens or reps of doing the math gave you.

One last thing, I feel like I shouldn’t have to say this but I know some lazy folks will do it. But DON'T EVER EVAA ask AI to make the call for you. Asking AI should I invest in this is a disaster waiting to happen and if you’re prone to doing this then maybe investing isn't for you. Use AI to pull information, summarize documents, and maybe give you insight to patterns you see. The actual decision has to come from you, period.

Electric calculators have been around for 60 years. The people who took their long division reps got the most out of them. AI in angel investing isn't gonna play out any different. The shortcut works way better when you don't need it.

Cheers,

~Abdul

Hey Hey… I’m Abdul I’m the chaiman of Ajo Angels and Shujaa Capital and I’m on a mission to introduce angel investing to 25,000 black folks over the next five years. I’m doing this with the goal of narrowing the racial wealth gap as well as trying to close the billion dollar funding gap for black founders.

This article reflects personal perspective and experience, not financial advice. Every career and investment path involves different risks and opportunities. Make decisions based on your own circumstances and goals.