Which one do you prefer? I'll do a comparison between different llms in the future, but I don't think there's one that is an order of magnitude better than the rest
The LLMs consistently miss the forest for the trees, today. The uber bear case on Wise is Crypto. Did Grok highlight that anywhere? The LLM (or maybe SLMs) will get better, getting to 90% will make them great. That last 10% is where the analyst has access to proprietary data, the public LLM can never get.
I’m not using Grok myself, but this was a great walkthrough on how to get the most out of any LLM for stock research. Your emphasis on crafting good prompts—and then verifying the output—really resonates. Too many people just take whatever the model spits out at face value.
Also appreciate the honest look at what these tools can and can’t do. Fact-checking with a second model is a smart move. At the end of the day, nothing replaces real due diligence, but using AI as a fast “first pass” makes sense.
Thanks for sharing the specific prompts—super helpful for anyone looking to speed up their research funnel.
Excellent post! Thanks for such a detailed article.
Thank you very much!
If you convert pdf to word docs it makes it easier for AI to read and you get more accurate results
Thanks. Yeah, I noticed that. All of them have trouble parsing through pdfs
I wouldn’t trust Grok with a lunch order. There are plenty of other AI’s you can do this with.
Which one do you prefer? I'll do a comparison between different llms in the future, but I don't think there's one that is an order of magnitude better than the rest
This is really helpful, thank you. 😊
My pleasure!
That's a comprehensive study. Thanks man!!
Thank you. Please let me know how I can make them better in the future!
Grok: "Say nothing, you understand? Get an AI!"
😀
The LLMs consistently miss the forest for the trees, today. The uber bear case on Wise is Crypto. Did Grok highlight that anywhere? The LLM (or maybe SLMs) will get better, getting to 90% will make them great. That last 10% is where the analyst has access to proprietary data, the public LLM can never get.
Fantastic article. I’ve officially lost my Substack commentary virginity.
I’m not using Grok myself, but this was a great walkthrough on how to get the most out of any LLM for stock research. Your emphasis on crafting good prompts—and then verifying the output—really resonates. Too many people just take whatever the model spits out at face value.
Also appreciate the honest look at what these tools can and can’t do. Fact-checking with a second model is a smart move. At the end of the day, nothing replaces real due diligence, but using AI as a fast “first pass” makes sense.
Thanks for sharing the specific prompts—super helpful for anyone looking to speed up their research funnel.