#Gate广场AI测评官

Investing Enters the AI Age! Why Can't Information "Democratization" Still Make Money? Beware of Being Deceived by AI

AI is increasingly becoming part of ordinary people's lives. In the AI age, investment information flows like springs, gushing forth abundantly, but this does not mean wealth follows suit. Remember, the soul of investing has always been yourself.

1. In Graham's era, the financial conditions of listed companies were almost unknown to the public. They only released one line of "income statement" information annually, showing how much profit they made that year. When Graham reviewed the annual reports of state-level business commissions, he discovered that 8 pipeline companies held large quantities of high-quality railroad bonds. Graham then drove to Washington, D.C., and in the archives of the state-level business commission, reviewed the annual reports of the 8 pipeline companies. Taking Northern Pipeline Company as an example, its stock price was $65, but each share contained $95 in implicit cash assets that could be almost entirely distributed to investors without affecting operations. This is the so-called cheap securities. Graham became famous overnight through his dealings with Northern Pipeline.

2. In Buffett's early investment career, he obtained information by reading Moody's Manual. Mohnish Pabrai, founder of Pabrai Fund, specifically bought several used copies of Moody's Manual on eBay. They were as thick as bricks with text as small as ants, with two or three company profiles per page. Buffett read through thousands of pages of the manual this way, and read it twice! He was searching for companies like GEICO: stock price $15, earnings per share of $25 the previous year. This was clearly like free money dropping from the sky.

3. In the internet age, the process of obtaining investment information has become very user-friendly for ordinary investors. Investors only need to input keywords to search, then flip through webpages one by one, or access company information through stock software F10.

4. However, in the AI age, large language models can extract the core of 100 pages of documents per second, and their intelligent summarization capabilities are also remarkable. After inputting the name of a listed company, large models provide information on stock price fluctuations, financial conditions, industry development, and company dynamics, and continuously monitor the company's movements according to investor needs. AI is an extremely powerful tool that has increased investors' information acquisition capabilities several times over, revealing information previously invisible to them relying on themselves; it can deeply process financial data, making a company's financial risks, growth potential, profitability, valuation comparisons, and more crystal clear; it can work continuously 24/7. If ordinary investors can use AI effectively, there is almost no "information gap" between them and institutional investors. This is information "democratization."

However, AI's shortcomings are equally fatal, with two core issues: hallucinations and data poisoning. Hallucinations refer to AI confidently spouting nonsense. Data poisoning refers to artificially injected "toxic data." How can we enjoy AI's efficiency dividend while avoiding traps?

First, the framework needs to be built by yourself and data sources must be limited; second, establish a "trusted knowledge base"; third, don't let AI provide answers, but rather have it organize facts.

"What must always be remembered is that for value investors, AI is merely a tool. The soul of investing will always be yourself."