Advantages:

• Source-grounded: Unlike general AI chatbots, NotebookLM focuses specifically on the content you provide, which means it will not pull irrelevant information, terminology or "ghost answers".

• Interaction with sources: NotebookLM allows you to ask questions, create outlines, generate content and even generate ideas (i.e. projects) based on the content provided.

• Audio overviews: A notable feature is the transformation of sources into podcast-style audio output which is great for reviewing information.

• Summarisation and insight: Because it is source-grounded, you can specify specific moderation tools and criteria, unlike other AI chatbots.

• Number of sources: NotebookLM can take up to 50 sources, unlike other free products.

• Ease of use: Google products are incredibly intuitive to the end-user.

Limitations:

• Document type: NotebookLM accepts a whole range of document types, including pdfs, jpegs and google docs.  It however, does not read Office Suite sources, such as Word, Excel or Powerpoint documents - to use these you need to convert it to pdf first.

• Formatting: While it is great at creating content, it cannot format or put it out in a document in the required way for printing - you need to copy the information to your preferred document type and edit it yourself.

• Image generation: NotebookLM cannot generate images - this means that to set questions for things like geometry, visual comprehensions or graphs you need to make the image in a different program and add it as a source.

• Prompt word count: There is a limit to how much information you can put into a single query - e.g. you cannot paste the content of an entire paper into the prompt box and ask it to edit it. To do this, its better to add the information as a source and ask it to refer to it.

• Referencing within prompts: NotebookLM currently performs poorly at referencing previous prompts; i.e., if you prompt it to do something the same way it did two prompts ago, it will not be able to do it. You need to repaste the previous prompt and then suggest changes.