Most people use AI as a better version of Google. I use it as a junior analyst who never sleeps, doesn’t get a bonus, and never judges my stock picks.
If you’re still asking LLMs to “summarize this article,” you’re leaving 90% of the value on the table. When it comes to company research or stock analysis, the goal isn't just to get the facts—it’s to find the signal in the noise.
I spent years in the startup world obsessed with trends, consumer sentiment, and enterprise transitions. But I’ve learned that data is cheap; synthesis is expensive. The real "alpha" comes from combining a unique perspective with high-fidelity data to build conviction.
The market is irrational and I can’t predict the future. Even with high conviction, I could be dead wrong. But the real work is in the research—and that’s the part I’ve successfully outsourced to AI.
Relying on Friends, Family, and Fools
I don’t spend my Sundays digging through 100-page earnings transcripts. Life is too short for that kind of friction.
Instead, I used to do what most people do—I read other people’s summaries. I’d scroll through X, read a few newsletters on Seeking Alpha, and check the "Analyst Take" on The Motley Fool. Or worse, I’d get those "guaranteed" stock tips from friends and neighbors, usually from an uncle who thinks "the cloud" is where the rain comes from.
I still value the "tips." I still want the ideas. But where I usually fall short is the actual execution of the research. I have the curiosity, but I don't always have the ten-hour block required to verify if a company is actually healthy or just wearing a lot of makeup.
I needed a way to take a raw idea and immediately pressure-test it against the ground truth, without losing a whole weekend to the manual labor of data collection.
Picking the Right Tool
I didn’t start with Gemini. Like any builder, I tried everything to find an edge.
ChatGPT: It was the first mover, but for a long time, it felt like it was stuck in a time capsule. Without native, reliable real-time data integration, it was useless for checking how a stock reacted to an earnings call and latest news.
Perplexity: This was my go-to for a while. It was great because it used live data and gave citations. It felt like a search engine that could “do research”.
Gemini: Recently, Google closed the gap, and is leading the pack. Because Gemini is natively plugged into the live web, it’s accurate and doesn’t hallucinate.
I’ve reached the point where I’m consolidating. I don’t need four subscriptions; I need one tool that actually understands context and has the most recent data. Gemini is the clear winner. And Gemini nicely organized my research in Google Drive.
From Summarizing to Auditing
I’ve never been the guy who feeds the AI raw data or uploads PDFs, and ask to summarize. I don’t have the patience to download transcripts and copy-paste them like a data entry clerk. Because Gemini is plugged directly into the live web, it goes out and does the legwork for me—pulling the latest filings, scanning the morning’s news, and finding the transcripts on its own.
But the "lightbulb moment" was when I tell Gemini it has to find at least 10 different sources before it’s allowed to give me an opinion.
Instead of reading a "Top 5 Takeaways" blog post written by some random person, I’m using a prompt that forces the AI to cross-reference multiple data points and pressure-test the official company narrative. I’m no longer looking for a summary of what they said; I’m looking for a verification of what they’re actually doing.
The "Cynical Analyst" Prompt
You can copy and paste this into Gemini right now. The beauty is that you can customize it—tell it your risk tolerance or your specific investment horizon.
Create a Gem. Make sure “Deep Research” is selected.
Inside “Instructions”, copy and paste my prompt. Feel free to edit to it based on your risk profile and preference.