Trading With Claude: A 30-Day Experiment in Learning to Invest with AI

Dushyant Remivasan March 9, 2026_

I want to be upfront about something: I don’t really know how to trade.

I know enough to be dangerous — I’ve read headlines, followed markets loosely, opened a Robinhood account. But I’ve never had a real system. No methodology. No framework for reading what’s happening in the market and translating that into a decision. I’ve mostly watched from the sidelines, vaguely convinced that investing was something other people did — people with MBAs, Bloomberg terminals, and time they didn’t have.

Then I started thinking about AI differently.

The Idea

I work in technology. I think about AI constantly — what it can do, where it falls short, how it changes the way people work and learn. And somewhere along the way, a question started nagging at me: what if I used Claude as a thinking partner to actually learn how to trade?

Not to automate it. Not to hand the wheel over to a machine and walk away. But to use AI the way a smart person uses a knowledgeable friend — someone who can explain what’s happening in the market in plain language, help me think through a thesis, push back when my instincts are wrong, and teach me the vocabulary and logic of investing in real time.

So that’s what I’m doing. For the next 30 days, I’m going to trade with Claude. Live. With real money.

The Rules

The account: $25 in Robinhood. That’s it. I’m not adding to it. Whatever I make or lose comes from that starting balance.

The goal: Learn. The profit targets I set each day are secondary to understanding why a trade makes sense — the macro context, the sector rotation, the risk factors. If I end the 30 days up $50 and genuinely understand how to read a market day, that’s a bigger win than doubling my money without knowing how I did it.

The transparency: I’m documenting everything. The reasoning, the trades, the mistakes, the moments where I disagreed with Claude’s read and what happened next. If I lose money, I’m writing about that too. This isn’t a highlight reel.

The AI role: Claude doesn’t execute trades — I do. It researches the market in real time, explains the context, proposes a plan, and helps me think through decisions as they evolve during the session. I can override it. I can ignore it. The judgment calls are mine.

Why $25?

A few reasons.

First, I can afford to lose it. That matters. One of the worst things a beginner can do is trade with money that has emotional weight attached to it. Twenty-five dollars is real enough to feel the stakes but small enough that a bad day won’t derail anything.

Second, constraints are educational. Figuring out how to generate meaningful returns from a small account teaches you things that unlimited capital hides. You learn about fractional shares, leveraged ETFs, position sizing, and the real cost of every decision when you can’t just throw more money at a problem.

Third — honestly? If I can learn to grow $25 consistently, the skills scale. The logic of a good trade doesn’t change whether you’re working with $25 or $25,000.

What I’m Not Doing

I want to be clear about what this experiment isn’t.

This is not financial advice. I’m a beginner writing about my own learning process in public. Nothing I post here should be read as a recommendation for what you should do with your money.

This is also not a claim that AI is a reliable investment tool. I genuinely don’t know how this turns out. Claude might help me make better decisions. It might also be wrong sometimes — and part of what I want to learn is how to recognize when to trust it and when to push back.

Day One

Today was my first live session. March 9, 2026 — which turned out to be one of the most volatile market days of the year so far. Crude oil briefly spiked above $100 a barrel following the escalation of the Iran war. The Dow dropped nearly 900 points at its low. Almost every sector was red.

I put $20 into SOXL — a 3x leveraged semiconductor ETF — based on Claude’s read that chips were the one sector showing real resilience amid the chaos. The position went underwater almost immediately. I held. Oil pulled back. The thesis played out.

I sold at $53.31. Profit: +$1.44.

Not life-changing. But it worked. And more importantly, I understood why it worked — which is the whole point.

You can read the full breakdown of Day One right here on the blog.

Follow Along

I’ll be posting after every session for the next 30 days. Some days will be winners. Some won’t. I’ll be honest about both.

If you’re someone who’s been curious about trading but intimidated by where to start, or if you’re interested in how AI actually performs as a real-time research and reasoning tool — this series is for you.

Let’s see where $25 goes.

Disclaimer: This blog documents a personal learning experiment and is not financial advice. Trading involves real risk of loss. Never invest more than you can afford to lose.