Full-time options trader Erik Smolinski has spent years using data and historical patterns to make trading decisions.
His systems and detailed trade logs have paid off so far — he consistently beats the S&P 500 and considers himself financially independent — but building that kind of research system takes time and money. It requires expensive datasets, coding skills, and the patience to wrangle large amounts of information.
“For most people, in order to do that kind of analysis, it is realistically out of their reach, next to impossible,” he told Business Insider. “I had to buy tons of super expensive data, and then I had to teach myself how to code — and I had no coding background.”
Now, he says AI is changing that.
Smolinski is currently building what he calls Project No Code, an AI-assisted investing research database with more than 700 million lines of data on options, stocks, and the economy, among other information. Thanks to AI tools, he says he has written none of the code himself.
Smolinski knows his setup is extreme. He’s buying massive, 28-terabyte hard drives just to support his options research — “right now, I’m at like 110 terabytes of data, and I need more,” he said — but he believes everyday investors can use the same basic idea at a much simpler level. Their goal should be to use AI to make better-informed decisions. That could mean researching a stock before buying it, understanding how similar companies have performed historically, stress-testing a retirement portfolio, or identifying portfolio risks before they become obvious.
What AI tools to use
Smolinski said he pays for the most expensive version of most AI tools so he can test them all. For his investing research, his favorites as of May 2026 are Claude, Claude Code, Codex, and Gemini.
For someone doing moderate to advanced research, he said the strongest combination is Claude Code and Codex. Unlike standard chatbots, Claude Code and Codex are coding-focused AI tools that can help users build, review, and run technical projects.
“Gemini is nice to have on top for broader data review, but I wouldn’t consider it a requirement,” he said. “Claude Code and Codex can pretty much handle it.”
The typical long-term investor doesn’t need that full stack. Smolinski said they can start with a basic chatbot, like Claude or ChatGPT.
How regular investors can use AI
Smolinski said many people are still using AI for relatively basic tasks, such as summarizing emails or drafting responses. While those uses can be helpful, he thinks they barely scratch the surface.
For investors, he recommends starting in a regular AI chat, such as Claude or ChatGPT, and describing an investing goal in plain English.
For example, an investor could ask: “I’m a long-term investor saving for retirement. Help me understand how I can use AI to analyze my portfolio and make better decisions.”
They could also try: “I want to build a portfolio that does [blank].” The “blank” could be beating the S&P 500, reducing volatility, saving for retirement, generating income, or balancing stocks and bonds.
From there, AI can help investors define what they actually want their portfolios to do. A goal like “beat the market” sounds simple, but it raises other questions: Are you willing to take more risk? Can you handle bigger swings? Are you optimizing for returns, stability, retirement income, or peace of mind?
For investors who want to stress-test a portfolio, Smolinski suggests asking AI to run a Monte Carlo simulation, which projects thousands of possible future outcomes rather than relying on a single forecast.
An investor could prompt: “I want you to run a Monte Carlo simulation, forward projecting what this portfolio could look like in 10,000 different paths.” The point, he said, is to get a clearer idea of what your future may look like.
To be clear, Smolinski isn’t encouraging investors to let AI manage their money. Rather, he said they can use it to clarify their goals, model trade-offs, and better understand risk before making decisions.
Use AI to challenge your thinking
Smolinski said he leans on AI to challenge his own assumptions.
“When I’m building strategies, I’ll say, ‘This is the general market effect that I think I see. Help me disprove my own thesis. Help me identify if I’m seeing something that’s not there.’ And it’ll walk me through a ton of novel ideas that aren’t necessarily occurring to me.”
Regular investors can use the same approach by asking questions such as:
- “What are the biggest risks in this investment thesis?”
- “Am I overweight in one sector?”
- “What assumptions am I making about future returns?”
- “What historical examples contradict my view?”
Investors should never blindly trust AI’s outputs, he emphasized. Models can make mistakes, misread data, or produce faulty code. Treat AI as a research tool, not as a personalized financial advisor.
Importantly, don’t ignore the technology altogether.
“You use AI, or you fall behind,” Smolinski said. “I firmly think that’s going to be the case increasingly.”
