AI Investment Thesis Review
Drop your thesis in. Four rival AIs make the bear case as rigorously as the bull case and hand back where it breaks — before your money moves.
A bull/bear synthesis with the assumptions the thesis depends on.
Signed off by an independent Final QA audit — with a “verify before you rely on this” list, never a black box.
Why this team
Every investment thesis is a story, and the storyteller is the worst person to check it. Warren Buffett’s seat asks the plain questions about the business behind the ticker; the Steelman is obligated to make the bear case as strong as the bull case — not the strawman version; the Quant Analyst asks what the price already assumes; and the Data Skeptic audits the evidence the thesis quietly leans on. The Risk Officer sizes the downside against the position, and the Contrarian asks the question that hurts: if this is so obvious, who is on the other side of the trade, and what do they know?
Each seat runs on a different frontier model — GPT, Claude, Gemini, Grok — so it’s a genuine cross-model review, not one AI in costumes. How a run works →
What the debate sounds like
An illustrative excerpt — An investor wants to put a fifth of their portfolio into a listed logistics-robotics company on the thesis that warehouse automation demand will double within five years.
Start with what you would need to believe. The thesis is about the industry, but you are not buying the industry — you are buying one company at one price. Warehouse automation can double while this stock halves, if competition compresses margins or the growth arrives at a rival. Can you explain, in one paragraph, why this company keeps the profits the trend creates? That paragraph is the thesis; everything else is weather.
The price has an opinion already: at the current multiple, the market is pricing in growth close to what the thesis calls its bull case. Which means being right about the trend earns roughly nothing — the position only pays if the company beats expectations that are already optimistic. The relevant bet is not “automation grows” but “automation grows faster than a market that agrees with you thinks”.
The strongest bear case, argued properly: robotics capex is the first thing warehouse operators cut in a downturn, the order book is concentrated in a few large customers whose renewals are assumed rather than contracted, and the technology moat is thinner than the deck suggests — the components are increasingly commodity. None of this needs the thesis to be wrong about the trend. That is what makes it dangerous.
Questions
Is this financial advice — will it tell me to buy or not?
No — the deliverable is explicitly educational analysis, not a buy/sell instruction, and it says so. What it commits to is a balanced bull/bear synthesis, the two or three assumptions the thesis actually rests on, and the early signals that would tell you the thesis is breaking. The decision, and any professional advice, remain yours.
What makes this better than reading analyst reports?
Analyst coverage clusters around consensus, and your own research clusters around confirmation — you have the thesis because you like it. The Steelman seat is structurally obligated to argue the other side at full strength, which neither your reading list nor a single chatbot asked to “review my thesis” will reliably do.
Does it work for assets other than stocks — property, crypto, a business?
Yes — the method is asset-agnostic: what must be true, what the price assumes, where the thesis is fragile, what would falsify it. The persona weighting shifts naturally: the cash-flow questions dominate for a business purchase, the what-does-the-price-assume question dominates for anything traded.
Your material is used only to run your review — never to train public models. Encrypted in transit and at rest. Security & privacy →
Want full control — pick your own minds, set the depth? Open the full council →