This post has the same issues as NotebookLM for me -- overdesigned, overengineered for what at its core is a simple and valuable UX.
NotebookLM: obviously useful, but I just wanna select some files and chat w/ them or have them summarized for me. It's got low info density, way too many cards/buttons/sections/icons, and it makes the core UX really difficult for me to navigate.
This post: I wanted to know what cool thoughts he had while designing it. Instead I get some weird scrolljacking, image carousels, unnecessary visual hierarchy, cards galore, etc.
Not trying to be too negative, it's slick and all but it just gets in the way for me instead of disappearing.
Not too negative, I really appreciate this perspective and agree with some of what you said.
IMO if you wanted to simply talk to a file or two Gemini, ChatGPT, and Claude are great for that.
The goal of this experimental product was to think creatively around what a true source grounded tool could be. (Obviously while building to best support the user needs). Our team put in immense work to move quickly while trying to be creative while keeping it simple. I have no doubt the product will continue to evolve and improve based on continued feedback like this!
Re: my website, I personally digest things better visually. I had hoped the additional visual elements would explain my decision making process to others as well.
Thank you for your work. NotebookLM has been invaluable for my learning experience. With the summaries, the mind maps, the multi-source synthesis and dialogue with the material, it adjusts pretty well to my learning style.
Surely there’s a German word for this - framing a weakness as if it contributed to the success.
I’ve seen it in so many talks, especially from people working in big tech. Something is a success in spite of some aspect of it, and those responsible for that aspect go on speaking tours about their journey and what we mere mortals can learn from them.
Came here to say this. Although the UI is clean, it's in no way a great user experience using NotebookLM. It's just such a great product so I go back to it, but the user interface is not my favorite part.
Literally the only thing people cared about was Audio Overviews. NotebookLM had launched months earlier without that feature and was immediately ignored.
The reason people liked audio overviews was because the voice model was amazing. This guy and the labs team had nothing to do with that. It was developed by a research team at Google Research/GDM who got maybe 1% of the credit: https://deepmind.google/discover/blog/pushing-the-frontiers-...
It’s amazing how these people still insist on denying the massive contribution the audio model made, which is apparent by the lack of credits or acknowledgements to the team anywhere public despite these self aggrandizing blogs and postcast appearances.
We found a 32% increase in subculture perception, which indicates that expressive design makes a brand feel more relevant and “in-the-know.” We also saw a 34% boost in modernity, making a brand feel fresh and forward-thinking. On top of that, there was a 30% jump in rebelliousness, suggesting that expressive design positions a brand as bold, innovative, and willing to break from convention.
Fully concur with sibling comments. Maybe I set the wrong expectation for myself: that it would be a blog post with a narrative to bring the reader through some plot points and a resolution. Instead it is a bunch of schematic ideas presented without connection or progression, like a conceptual summary sketched from an interesting lecture that can be read between the lines.
There’s a lot of great things here that I like - adaptive 3 panel structure, source viewer on left.
But I think there are a few things that I noticed about NBLM which was painful for me.
1// The three panels should be toggle based with icons on the top bar. There is no need to occupy real estate for both notes and chat together if they are not being used together.
2// The center of screen or the largest middle section should be focused on outputs and not chat. If you are focusing on creating something why should it be on the side. Especially since chat isn’t all that special a feature compared to the audio overviews, etc.
3// Information density - the buttons and icons are all too large and clunky. You are in fight for real estate because AI is helping you process superhuman amounts of information.
I think the magic of NBlM is the audio overviews, the chat based Q&A is with citation is pretty standard for all LLMs.
Also I think it only uses Gemini flash which feels like a search model - this needs to be paired with a reasoning model instead.
I use NotebookLM everyday. The simplicity of the design is much appreciated. However, there are real issues scaling the design and keeping it user friendly as the team keeps adding new features.
The most recent example of this is with the addition of 2 new capabilities (Flashcards and Quiz), "Artifacts Button Container" now has 6 large buttons, and is 328px in height! There are users who are accessing the site from small screen devices in India and they have been asking for help on Discord forums because they cannot see their notes anymore. So I had to create a Tampermonkey script to let users collapse it.[0] I heard the team is fixing that soon, but they should have done more testing before releasing it.
There are other issues like this that I've fixed with scripts. The strangest one is the "notes." Why force the users read a 2000 word essay in a 360px sidebar? So I wrote a script I wrote to let you pop it into full screen mode.[1]
Another example is the chat input field. The follow up questions are hardly usable at all. And they're not stable after you select them.
I can go on all day, but I think it's better to fix things than to complain.
This is my website! I love seeing everyone’s perspective, design is all about iteration and constant evolution. I have no doubt NLM will continue to evolve and change. I feel immensely lucky to have been there towards the beginning of that evolution.
As I designer I was trying to skate to where the puck was going technically. We’d get some insight on what was coming from the model side and we’d try to build UI around that future before it arrived. The labs team at Google has done a solid job of trying to build with that mentality.
We were in a mad dash for 1.5 years launching early and listening to user feedback then iterating our way to where the product landed. As I mentioned, those close to me knew how all consuming this process became. It was an amazing time taking a new product from 0 to 1 inside Google.
I Definitely never expected my portfolio site to make the rounds like this!
Apparently it advertises to browsers and other extensions that it has dark mode activated when it's actually not, which prevents them from applying the appropriate theme.
Thank you for your hard work. I gripe about it sometimes but it’s still the most compelling learning experience I’ve ever come across. I hope it keeps getting better and better.
I don't care much for NotebookLM's UX. I find the layout confusing, the concepts conveyed in the UI not very intuitive. I feel like the power of what's under the hood doesn't translate very well into the frontend.
That said, this article was a very nice overview. Clean page, interesting to see his perspective. He clearly took tare and thought about the project. I don't agree with his conclusions and results, but that's just one user. Hopefully we see more thoughtful approaches. The space moving so quickly doesn't exactly foster a craftsman-like care in design.
I don’t know that I’ve seen any product due quite the same thing yet. Like you said the space is pretty new and most of the other LLMs that I’ve seen tend to be about doing research. This seems to be the only one that’s really focused on just researching current documents you upload and giving you an interface to explore that. I don’t think you will take too long for new tools to come, but it’s too early.
leshokunin is right. there’s not at all many close app to it. Among few I could find nouswise very close and having a note driven design was a plus for me. Idk how this feels like for you.
As a non-user who just played with the podcast thing a while back, is notebooklm actually better than just using something like Claude Projects to dump all your files into it, and then chatting with it in that project?
Other than the podcast thing of course which is unique.
Mostly curious about perspectives of folks who used both and can compare them.
I came here to say exactly the same thing. Now the I have Claude Code and Codex CLI I just creat a folder with all the documents I need and I work from there.
Love the context size and podcasts but never understood the UX. Why notes? Is there a difference between my notes and AI notes? Guess that's not my mental model of working with papers (Phd in psy) or I'm just spoiled by elicit.
I'm an autism researcher. I also don't get the "notes" thing. Am I supposed to use it as my note taking app (e.g. like I use OneNote?).
Sometimes interactive discussion with the LLM about a paper can speed learning. I've used audio overviews for the drive to the lab, but I haven't been able to get the overview to go *hard* into the science.
Notes help when sharing the notebook with others. By sharing snippets with others, they immediately have conversation starters that they can use to dive deeper into the material.
The idea of multiple files upload and audio creation with it does not go well with me. How does it choose which topics it should concentrate on? What is the idea behind multiple files upload on a single notebooklm project. I am struggling to understand that.
The user interface has been confusing to me. And note, I have built some projects using notebooklm https://asimov.learntosolveit.com
Any reason why when I upload a set of books to NotebookLM and have an interesting conversation about them with NotebookLM, the conversation is not stored? I can't revisit it later, I can't continue from there. I don't understand why they changed the UX/UI from the other AI solutions (Gemini or GPT).
I also don't get why NotebookLM refuses to write things either, I can't make it write an essay based on the information I fed through PDFs or other files.
Somewhat related, I also find it bizarre that they all but prevent you from getting information out of the app. Export to/Open in Google Docs seems like a no-brainer, but copy-paste is the only option.
NotebookLM has been one of the more useful projects I've used. Being able to focus a language processing model directly to the content you're looking for is really useful.
I currently support a data application where the vendor's documentation for it is not particularly well organized. It's spread across dozens of PDFs, sometimes without particularly well reasoned organization. Dumping all those PDFs into NotebookLM has been extremely useful since it allows us to ask questions that either give us answers, or are immediately fruitless requiring us to contact the vendor. Having an LLM capable of processing all that text has been great.
I've also used it with tabletop roleplaying game manuals. It's especially useful for badly organized TTRPGs, or those with poor indexing. Being able to type a question out in the middle of a game and getting an answer without having to dig through the book and find it yourself can be really very useful.
What I've never found a use for is anything in the Studio pane. They're neat tools, but... it's never been anything I've wanted or needed.
Ha this is one of my uses as well! I'm running an FFG SWRPG campaign and being able to ask it questions about overarching connections between plotlines helps quite a bit.
The mind map feature is also handy, although last I used it, it was a bit clunky to export.
Interesting about the TTRPG. I am a fan of 3D6 Down the Line, the (criminally Under-watched) Actual Play series..where they were doing Arden Vul, a massively intricate mega dungeon..I bet it could have been useful to the DM to look up info over plain search
I'm curious from people who use it regularly. What do you use it for? Aside from audio overviews, what does it do better for you compared to vanilla chat interfaces or docs integration?
I use it for asking questions about a specific board games. It is INVALUABLE for learning very complex games. Instead of hunting through a rulebook trying to find an obscure rule, I just ask it and it answers, citing where it finds it.
Claude code is...
1. Not available on my phone in a friendly form (I know there's claude web)
2. Doesn't format its answers with cited links to where what I need is found in the document
3. Doesn't have a quick, easy interface where I can upload multiple resources to a single index. You have chats but it's formatted where you're meant to have a conversation. NotebookLM is centered around "chat about these resources".
I use it to turn an arxiv paper or hacker news comment section + interesting long form OP, plus maybe other sources, into a podcast that I can listen to on my commute.
Yeah, those audio overviews can be useful! The interface has a bunch of other features too, so I am hoping to hear more about those too. I'm not affiliated, just curious.
The ui however has a 1 click button to do so, while it doesn't do that for aistudio and you need to do something more complex. I think he likes how the ui complements the llm, and I tried it these days and I agree
I am a post sales consultant. Meaning I’m the first deeply technical person that a client talks to after they sign the contract that pre-sales has worked on.
Pre-sales has already had a couple of meetings by the time it gets to me. I put the transcripts of their calls and the contract (statement of work) into Notebook LLM to ask the high level questions like the objectives, the challenges, priorities, risks they have already surfaced etc.
That feeds into my first meeting slide deck so the client can expand upon, disagree with etc to make sure everyone is on the same page.
After my set of discovery sessions, I put those transcripts into the notebook too.
I can then use Notebook LLM to give me a first draft of the management style assessment report. I never expose LLM writing to the client. I have a certain style of writing and I hate AI slop.
Yes this is corporate approved, we use GSuite for everything. I like the fact that everything that NotebookLM outputs has curations from the source material.
Wait until you find out that the client who signed the contract, the lawyers who wrote it, the pre-sales who designed it, and the account executive who started it… also all used NotebookLM for their stage of the deal.
(btw, it’s “NotebookLM” – one “L” – not “Notebook LLM”)
That’s why my kickoff meetings always involve validating what the client wants as far as business outcomes instead of blindly trusting account executives and sales.
On the other hand, part of the reason I never desired to go independent is because working for a company, we do have a sales department to get clients, a legal team, and finance to chase payments. I can just concentrate on the client and the project.
Honestly, you’re in the bravest part of the process. We call it “professional services” but same thing… you’re the one who has to deliver on whatever was promised to the customer, which was totally outside your visibility.
Podcast generation is quite interesting as a concept, but if you listen to more than a few hours of podcasts a week, you'll probably find it uncomfortable to hear anything coming out of NotebookLLM -- too unnatural, the pace is off, and the thing feels more like two robots taking turns to read a script with exaggerated reactions than two humans chatting with pauses and interruptions. At least I stopped after 3 minutes of that.
I was putting together an essay/podcast about a news event in another country, and I was able to collect multiple YouTube videos, search and translate their transcripts, and ask questions like 'is this the video where someone talks about _?'
When working with documents, the citations that NotebookLM provide increase the confidence in answers. This allows to use NotebookLM for team-level knowledge bases.
i think the main star of notebooklm is the audio model and the ux gets in the way of that
and if you're not very clear about what drove notebooklm's success you might be at risk of cargo culting ALL of their process instead of just focusing on the thing that worked.
I would've preferred the original design. You can't even resize each panel - why can't I read a summary by itself without the chat taking up 2/3 of the screen?
I have never understood the distinction between "Chat" and "Studio". I can to the same things in both places for the most part. I find I spend all of my time in Chat and then wonder if I'm supposed to be doing things in "Studio". I just find myself hiding it for the most part.
I really wish NotebookLM would hurry up and open a payment tier where they don't harvest your data. We'd love to deploy it across our org but control of sensitive information is a requirement.
NotebookLM is extremely badly engineered: The large downloads don't work in mobile browsers. For the longest time, the downloads were in wav format which doesn't make any sense. The mobile app lacks much of the functionality too. The generated audio is often about an altogether different topic than the one I requested. And don't get me started about the quality of the speech -- it still has an occasional "third voice" problem that just doesn't fit in. Overall, I would say its UI fits the culture of shipping utter crap, and I would never hire any designer of that team.
This post has the same issues as NotebookLM for me -- overdesigned, overengineered for what at its core is a simple and valuable UX.
NotebookLM: obviously useful, but I just wanna select some files and chat w/ them or have them summarized for me. It's got low info density, way too many cards/buttons/sections/icons, and it makes the core UX really difficult for me to navigate.
This post: I wanted to know what cool thoughts he had while designing it. Instead I get some weird scrolljacking, image carousels, unnecessary visual hierarchy, cards galore, etc.
Not trying to be too negative, it's slick and all but it just gets in the way for me instead of disappearing.
Not too negative, I really appreciate this perspective and agree with some of what you said.
IMO if you wanted to simply talk to a file or two Gemini, ChatGPT, and Claude are great for that.
The goal of this experimental product was to think creatively around what a true source grounded tool could be. (Obviously while building to best support the user needs). Our team put in immense work to move quickly while trying to be creative while keeping it simple. I have no doubt the product will continue to evolve and improve based on continued feedback like this!
Re: my website, I personally digest things better visually. I had hoped the additional visual elements would explain my decision making process to others as well.
Thank you for your work. NotebookLM has been invaluable for my learning experience. With the summaries, the mind maps, the multi-source synthesis and dialogue with the material, it adjusts pretty well to my learning style.
Surely there’s a German word for this - framing a weakness as if it contributed to the success.
I’ve seen it in so many talks, especially from people working in big tech. Something is a success in spite of some aspect of it, and those responsible for that aspect go on speaking tours about their journey and what we mere mortals can learn from them.
Try "Lobhudelei"
Well, not German but there is the tale of the fox who lost its tail and framed it as a good thing
Endless scrolling just to describe a three-panel layout, something that's been around since the 80s (?).
Came here to say this. Although the UI is clean, it's in no way a great user experience using NotebookLM. It's just such a great product so I go back to it, but the user interface is not my favorite part.
I read this comment first and thought: "oh come on, how bad can it be". So then seeing this really being so bad(-ly overdesigned) was quite amusing.
It's written for people to see, not to read I guess.
Really disliked the author's tone: "me, me, me, me, me, me. Also somebody promote me."
All this paragraphs telling us what a great job he's done and this charlatan's product (with great potential)...doesn't even save your chats.
So you start discussing about something in your notes, come back and...the messages are gone.
But sure come tell me what a genius you are with your 3 column layout.
Charlatans, and those people become multi millionaires with those projects and such a crap work.
Literally the only thing people cared about was Audio Overviews. NotebookLM had launched months earlier without that feature and was immediately ignored. The reason people liked audio overviews was because the voice model was amazing. This guy and the labs team had nothing to do with that. It was developed by a research team at Google Research/GDM who got maybe 1% of the credit: https://deepmind.google/discover/blog/pushing-the-frontiers-... It’s amazing how these people still insist on denying the massive contribution the audio model made, which is apparent by the lack of credits or acknowledgements to the team anywhere public despite these self aggrandizing blogs and postcast appearances.
Lmao this is my feeling as well.
Why is a 3 panel layout new?
Isn't vscode essentially that?
I can’t even remember an IDE that doesn’t have the 3 panel layout.
Eclipse, xcode, PyCharm, even Visual Studio from way back when…
I mean, this is from a company that judges its designs for "feeling of rebelliousness" and "courage" [1]
The entire text is rationalizing post-factum for a promotion package.
[1] In case you missed it https://design.google/library/expressive-material-design-goo...
--- start quote ---
We found a 32% increase in subculture perception, which indicates that expressive design makes a brand feel more relevant and “in-the-know.” We also saw a 34% boost in modernity, making a brand feel fresh and forward-thinking. On top of that, there was a 30% jump in rebelliousness, suggesting that expressive design positions a brand as bold, innovative, and willing to break from convention.
--- end quote ---
Fully concur with sibling comments. Maybe I set the wrong expectation for myself: that it would be a blog post with a narrative to bring the reader through some plot points and a resolution. Instead it is a bunch of schematic ideas presented without connection or progression, like a conceptual summary sketched from an interesting lecture that can be read between the lines.
There’s a lot of great things here that I like - adaptive 3 panel structure, source viewer on left.
But I think there are a few things that I noticed about NBLM which was painful for me.
1// The three panels should be toggle based with icons on the top bar. There is no need to occupy real estate for both notes and chat together if they are not being used together.
2// The center of screen or the largest middle section should be focused on outputs and not chat. If you are focusing on creating something why should it be on the side. Especially since chat isn’t all that special a feature compared to the audio overviews, etc.
3// Information density - the buttons and icons are all too large and clunky. You are in fight for real estate because AI is helping you process superhuman amounts of information.
I think the magic of NBlM is the audio overviews, the chat based Q&A is with citation is pretty standard for all LLMs.
Also I think it only uses Gemini flash which feels like a search model - this needs to be paired with a reasoning model instead.
I use NotebookLM everyday. The simplicity of the design is much appreciated. However, there are real issues scaling the design and keeping it user friendly as the team keeps adding new features.
The most recent example of this is with the addition of 2 new capabilities (Flashcards and Quiz), "Artifacts Button Container" now has 6 large buttons, and is 328px in height! There are users who are accessing the site from small screen devices in India and they have been asking for help on Discord forums because they cannot see their notes anymore. So I had to create a Tampermonkey script to let users collapse it.[0] I heard the team is fixing that soon, but they should have done more testing before releasing it.
There are other issues like this that I've fixed with scripts. The strangest one is the "notes." Why force the users read a 2000 word essay in a 360px sidebar? So I wrote a script I wrote to let you pop it into full screen mode.[1]
Another example is the chat input field. The follow up questions are hardly usable at all. And they're not stable after you select them.
I can go on all day, but I think it's better to fix things than to complain.
[0] https://gist.github.com/volkanunsal/94db50629cad816eca84c836...
[1] https://gist.github.com/volkanunsal/fded9124d62422c0d2672b8a...
This is my website! I love seeing everyone’s perspective, design is all about iteration and constant evolution. I have no doubt NLM will continue to evolve and change. I feel immensely lucky to have been there towards the beginning of that evolution.
As I designer I was trying to skate to where the puck was going technically. We’d get some insight on what was coming from the model side and we’d try to build UI around that future before it arrived. The labs team at Google has done a solid job of trying to build with that mentality.
We were in a mad dash for 1.5 years launching early and listening to user feedback then iterating our way to where the product landed. As I mentioned, those close to me knew how all consuming this process became. It was an amazing time taking a new product from 0 to 1 inside Google.
I Definitely never expected my portfolio site to make the rounds like this!
Just for my curiosity: are there any limitations to what Google IP you can write about in public like this?
I was wondering about this question as well.
Are u authorized to share all this IP?
Hi, there's a bug on your website.
Apparently it advertises to browsers and other extensions that it has dark mode activated when it's actually not, which prevents them from applying the appropriate theme.
That's a webdev trick for forcing light mode lol
Thank you for your hard work. I gripe about it sometimes but it’s still the most compelling learning experience I’ve ever come across. I hope it keeps getting better and better.
@jayspiel
> As A designer I was trying to skate to where the puck was going technically.
Resonates big time! At the end of the day, this isn't a full-proof science - it's an art. Req-con-fin(Requires continuous finessing).
I also assume, the project revolved around many roles and as you mentioned, the project was iteratively built around user feedback.
NLM disrupted the space and I know just like with the early days of bard/gemini this will only get insanely better; UI/UX especially.
Dey Well
I don't care much for NotebookLM's UX. I find the layout confusing, the concepts conveyed in the UI not very intuitive. I feel like the power of what's under the hood doesn't translate very well into the frontend.
That said, this article was a very nice overview. Clean page, interesting to see his perspective. He clearly took tare and thought about the project. I don't agree with his conclusions and results, but that's just one user. Hopefully we see more thoughtful approaches. The space moving so quickly doesn't exactly foster a craftsman-like care in design.
Do you use an alternative? I haven't explored the space much.
I don’t know that I’ve seen any product due quite the same thing yet. Like you said the space is pretty new and most of the other LLMs that I’ve seen tend to be about doing research. This seems to be the only one that’s really focused on just researching current documents you upload and giving you an interface to explore that. I don’t think you will take too long for new tools to come, but it’s too early.
leshokunin is right. there’s not at all many close app to it. Among few I could find nouswise very close and having a note driven design was a plus for me. Idk how this feels like for you.
As a non-user who just played with the podcast thing a while back, is notebooklm actually better than just using something like Claude Projects to dump all your files into it, and then chatting with it in that project?
Other than the podcast thing of course which is unique.
Mostly curious about perspectives of folks who used both and can compare them.
The podcasts feature is primarily why I love NLM!
I try starting my morning with learning, lately having a podcast to listen to while I start my day has been awesome.
I came here to say exactly the same thing. Now the I have Claude Code and Codex CLI I just creat a folder with all the documents I need and I work from there.
Then NotebookLM has one point in it's favor: you can use it on your phone with ease too.
Love the context size and podcasts but never understood the UX. Why notes? Is there a difference between my notes and AI notes? Guess that's not my mental model of working with papers (Phd in psy) or I'm just spoiled by elicit.
I'm an autism researcher. I also don't get the "notes" thing. Am I supposed to use it as my note taking app (e.g. like I use OneNote?). Sometimes interactive discussion with the LLM about a paper can speed learning. I've used audio overviews for the drive to the lab, but I haven't been able to get the overview to go *hard* into the science.
Notes help when sharing the notebook with others. By sharing snippets with others, they immediately have conversation starters that they can use to dive deeper into the material.
The idea of multiple files upload and audio creation with it does not go well with me. How does it choose which topics it should concentrate on? What is the idea behind multiple files upload on a single notebooklm project. I am struggling to understand that.
The user interface has been confusing to me. And note, I have built some projects using notebooklm https://asimov.learntosolveit.com
Any reason why when I upload a set of books to NotebookLM and have an interesting conversation about them with NotebookLM, the conversation is not stored? I can't revisit it later, I can't continue from there. I don't understand why they changed the UX/UI from the other AI solutions (Gemini or GPT).
I also don't get why NotebookLM refuses to write things either, I can't make it write an essay based on the information I fed through PDFs or other files.
Somewhat related, I also find it bizarre that they all but prevent you from getting information out of the app. Export to/Open in Google Docs seems like a no-brainer, but copy-paste is the only option.
NotebookLM has been one of the more useful projects I've used. Being able to focus a language processing model directly to the content you're looking for is really useful.
I currently support a data application where the vendor's documentation for it is not particularly well organized. It's spread across dozens of PDFs, sometimes without particularly well reasoned organization. Dumping all those PDFs into NotebookLM has been extremely useful since it allows us to ask questions that either give us answers, or are immediately fruitless requiring us to contact the vendor. Having an LLM capable of processing all that text has been great.
I've also used it with tabletop roleplaying game manuals. It's especially useful for badly organized TTRPGs, or those with poor indexing. Being able to type a question out in the middle of a game and getting an answer without having to dig through the book and find it yourself can be really very useful.
What I've never found a use for is anything in the Studio pane. They're neat tools, but... it's never been anything I've wanted or needed.
Ha this is one of my uses as well! I'm running an FFG SWRPG campaign and being able to ask it questions about overarching connections between plotlines helps quite a bit.
The mind map feature is also handy, although last I used it, it was a bit clunky to export.
Interesting about the TTRPG. I am a fan of 3D6 Down the Line, the (criminally Under-watched) Actual Play series..where they were doing Arden Vul, a massively intricate mega dungeon..I bet it could have been useful to the DM to look up info over plain search
I'm curious from people who use it regularly. What do you use it for? Aside from audio overviews, what does it do better for you compared to vanilla chat interfaces or docs integration?
I use it for asking questions about a specific board games. It is INVALUABLE for learning very complex games. Instead of hunting through a rulebook trying to find an obscure rule, I just ask it and it answers, citing where it finds it.
Are you telling me Claude code or something couldn't do this easily too?
Claude code is... 1. Not available on my phone in a friendly form (I know there's claude web) 2. Doesn't format its answers with cited links to where what I need is found in the document 3. Doesn't have a quick, easy interface where I can upload multiple resources to a single index. You have chats but it's formatted where you're meant to have a conversation. NotebookLM is centered around "chat about these resources".
I use it to turn an arxiv paper or hacker news comment section + interesting long form OP, plus maybe other sources, into a podcast that I can listen to on my commute.
Yeah, those audio overviews can be useful! The interface has a bunch of other features too, so I am hoping to hear more about those too. I'm not affiliated, just curious.
Video explainers, for me, are better than the audio overviews.
A friend uses it to create quizzes and flashcards when studying for uni
Any vanilla chat LLM can do this.
The ui however has a 1 click button to do so, while it doesn't do that for aistudio and you need to do something more complex. I think he likes how the ui complements the llm, and I tried it these days and I agree
NotebookLM gives you sources across powerpoints, pdfs, etc.
I like the mindmap tool
I am a post sales consultant. Meaning I’m the first deeply technical person that a client talks to after they sign the contract that pre-sales has worked on.
Pre-sales has already had a couple of meetings by the time it gets to me. I put the transcripts of their calls and the contract (statement of work) into Notebook LLM to ask the high level questions like the objectives, the challenges, priorities, risks they have already surfaced etc.
That feeds into my first meeting slide deck so the client can expand upon, disagree with etc to make sure everyone is on the same page.
After my set of discovery sessions, I put those transcripts into the notebook too.
I can then use Notebook LLM to give me a first draft of the management style assessment report. I never expose LLM writing to the client. I have a certain style of writing and I hate AI slop.
Yes this is corporate approved, we use GSuite for everything. I like the fact that everything that NotebookLM outputs has curations from the source material.
Wait until you find out that the client who signed the contract, the lawyers who wrote it, the pre-sales who designed it, and the account executive who started it… also all used NotebookLM for their stage of the deal.
(btw, it’s “NotebookLM” – one “L” – not “Notebook LLM”)
That’s why my kickoff meetings always involve validating what the client wants as far as business outcomes instead of blindly trusting account executives and sales.
On the other hand, part of the reason I never desired to go independent is because working for a company, we do have a sales department to get clients, a legal team, and finance to chase payments. I can just concentrate on the client and the project.
Honestly, you’re in the bravest part of the process. We call it “professional services” but same thing… you’re the one who has to deliver on whatever was promised to the customer, which was totally outside your visibility.
I stopped using NotebookLM after one use. Not sure what’s the fuss aside from the podcast generation gimmick
Podcast generation is quite interesting as a concept, but if you listen to more than a few hours of podcasts a week, you'll probably find it uncomfortable to hear anything coming out of NotebookLLM -- too unnatural, the pace is off, and the thing feels more like two robots taking turns to read a script with exaggerated reactions than two humans chatting with pauses and interruptions. At least I stopped after 3 minutes of that.
I was putting together an essay/podcast about a news event in another country, and I was able to collect multiple YouTube videos, search and translate their transcripts, and ask questions like 'is this the video where someone talks about _?'
Copilot built into Ms edge can do this?
When working with documents, the citations that NotebookLM provide increase the confidence in answers. This allows to use NotebookLM for team-level knowledge bases.
U can do that with chatgpt which preceded notebooklm
I replay liked the design of NLM, and thought it was a great implementation of using LLMs beyond a simple chat. Well done.
i think the main star of notebooklm is the audio model and the ux gets in the way of that
and if you're not very clear about what drove notebooklm's success you might be at risk of cargo culting ALL of their process instead of just focusing on the thing that worked.
I would've preferred the original design. You can't even resize each panel - why can't I read a summary by itself without the chat taking up 2/3 of the screen?
I have never understood the distinction between "Chat" and "Studio". I can to the same things in both places for the most part. I find I spend all of my time in Chat and then wonder if I'm supposed to be doing things in "Studio". I just find myself hiding it for the most part.
I really wish NotebookLM would hurry up and open a payment tier where they don't harvest your data. We'd love to deploy it across our org but control of sensitive information is a requirement.
I get it as part of my gemini pro subscription.
NotebookLM is extremely badly engineered: The large downloads don't work in mobile browsers. For the longest time, the downloads were in wav format which doesn't make any sense. The mobile app lacks much of the functionality too. The generated audio is often about an altogether different topic than the one I requested. And don't get me started about the quality of the speech -- it still has an occasional "third voice" problem that just doesn't fit in. Overall, I would say its UI fits the culture of shipping utter crap, and I would never hire any designer of that team.