I think the main problem with using AI is that it is not trained for investing purposes. It is simply trained to generate text on whatever subject is asked, using the most probable words from information that is generally available to the public. This is especially the case when an answer is generated using low-quality sources, and even more so when AI-generated text was used to train the model in the first place. It is hard to argue that this information has any value from a market perspective.
I tried NotebookLM and consider it a powerful search engine for natural language information in various formats. Nothing more, nothing less. Without high-quality information sources and the right queries, it’s basically useless. And if you know the right questions to ask about a particular business, you sort of know the business already. So, it's a chicken-and-egg problem. :)
This is a very interesting article. Maybe I'll give it a go because I, too, struggle with the AI telling me made-up stories and being unable to track down its sources.
You say that because you've experienced the difference. But fair enough, it doesn't look hard to use, I'll probably give it a shot next time I'm frustrated at Gemini's B.S. answers ;)
I understood the benefit thanks the other commenter and your answer to him ("with NotebookLM you force the AI to only use your data instead of making things up.")
I use somewhat the same process feeding NotebookLM deep research from other LLMs. But I also like the audio overview. You can get some high quality "podcast" from it.
I think the main problem with using AI is that it is not trained for investing purposes. It is simply trained to generate text on whatever subject is asked, using the most probable words from information that is generally available to the public. This is especially the case when an answer is generated using low-quality sources, and even more so when AI-generated text was used to train the model in the first place. It is hard to argue that this information has any value from a market perspective.
Agreed. And that’s what makes NotebookLM so valuable because you choose the data to "train" the model on effectively.
I tried NotebookLM and consider it a powerful search engine for natural language information in various formats. Nothing more, nothing less. Without high-quality information sources and the right queries, it’s basically useless. And if you know the right questions to ask about a particular business, you sort of know the business already. So, it's a chicken-and-egg problem. :)
This is a very interesting article. Maybe I'll give it a go because I, too, struggle with the AI telling me made-up stories and being unable to track down its sources.
Why maybe? It’s so valuable in one‘s process!
You say that because you've experienced the difference. But fair enough, it doesn't look hard to use, I'll probably give it a shot next time I'm frustrated at Gemini's B.S. answers ;)
I understood the benefit thanks the other commenter and your answer to him ("with NotebookLM you force the AI to only use your data instead of making things up.")
I use somewhat the same process feeding NotebookLM deep research from other LLMs. But I also like the audio overview. You can get some high quality "podcast" from it.