Loading request...
Using ChatGPT as a stock research assistant is great, but it lacks real-time data. Users want the ability to integrate real-time stock data for more accurate analysis and decision-making.
I know AI and stock picking is a loaded topic, so let me set expectations: ChatGPT can't predict prices and anyone saying otherwise is selling a course. But as a *research* *assistant*, it's been genuinely useful for me. My workflow: I do the actual data-pulling from Yahoo Finance, Finviz, and SEC filings. Then I use ChatGPT to help me think through the analysis. Stuff like: \- Pasting earnings numbers and asking it to summarize in plain English + flag what's unusual \- Running a "devil's advocate" prompt where it argues the bear case for a stock I'm bullish on \- Analyzing competitive moats using the Morningstar framework (brand, switching costs, network effects, cost advantages, intangible assets) \- Comparing two companies side by side with a structured framework The key realization: it's a thinking partner, not a data source. Every number it gives you needs verification. It will hallucinate financial data with complete confidence. Where it falls short: no real-time data, no technical analysis capability worth using, and it tends to be overly positive about well-known companies unless you explicitly tell it to be critical. Anyone else using LLMs as part of their research process? I'm curious how others are incorporating it. I also used notebooklm to summarize earning reports a few times and had an audio file as output which was coo. However I am not sure if all data was captured and what the LLM considered irrelevant to filter out. Anyone (non-professional) tried Claude to analyze data?