Google's AI Note-Taking App Is Now Available 26
NotebookLM, Google's experimental AI-powered note-taking app unveiled at Google I/O earlier this year (formerly referred to as Project Tailwind), is now widely available in the United States. "It's also getting several new features and is 'starting' to use Google's Gemini Pro AI model 'to help with document understanding and reasoning,'" reports The Verge. From the report: NotebookLM can already do things like summarize the documents you import into the app, come up with key points, and even answer questions about your note-taking sources. But now, Google is adding a way to transform your notes into other types of documents, too. Once you select all the notes you want to include, NotebookLM will automatically suggest formats, such as an outline or study guide. However, Google notes that you can also tell NotebookLM to transform your notes into a format of your choosing, like an email, script outline, newsletter, and more.
Additionally, NotebookLM will now start providing suggested actions based on what you're doing in the app. As an example, Google says if you're writing a note, NotebookLM may automatically "offer tools to polish or refine your prose, or suggest related ideas from your sources based on what you've just written." Some other handy features coming to the app include a way to save helpful responses from NotebookLM as notes, share your notes with others, and focus NotebookLM's AI on select sources when chatting with it. Google is expanding some of NotebookLM's limitations as well. You can now include up to 20 sources in your notebook, each with up to 200,000 words.
Additionally, NotebookLM will now start providing suggested actions based on what you're doing in the app. As an example, Google says if you're writing a note, NotebookLM may automatically "offer tools to polish or refine your prose, or suggest related ideas from your sources based on what you've just written." Some other handy features coming to the app include a way to save helpful responses from NotebookLM as notes, share your notes with others, and focus NotebookLM's AI on select sources when chatting with it. Google is expanding some of NotebookLM's limitations as well. You can now include up to 20 sources in your notebook, each with up to 200,000 words.
Clippy 2.0 (Score:3)
It looks like you're writing a shopping list. Would you like help with that? Last time you went shopping you bought 50 pounds of sugar. Would you like to add 50 pounds of sugar to your list?
Re: (Score:2)
Damn! Beat me to it. "NotebookLM will now start providing suggested actions based on what you're doing in the app" sure as hell does sound like Clippy on steroids. Or Badgey.
Clippy on Steroids (Score:2)
With AI, imagine all the new and fun possibilities Clippy 2.0 can provide us!
"I see that you're planning for a budget. Looks like you don't have enough money to support a child this year. Can I recommend to you a good contraceptive?"
"I see that you're writing a novel. I have already assimilated your draft into my memory banks to contribute to other aspiring authors who lack any ounce of creativity but write prompts with such precision that I am required to plagiarize your entire novel for their benefit.
Re: (Score:2)
Huh so humons invented anticompression (Score:2)
Seems like we've invented anticompression. You write down a few notes, it's autounsummarized (https://www.bbspot.com/News/2003/12/autounsummarize.html) to a colossal length which then goes over the wire. Once received, it's then recompressed with AI.
Countdown (Score:1)
To this being deprecated and retired
The betting pool is open (Score:2)
Since Google released a new gimmick, it's time to get the betting pool going: How long 'til Google sends it to the graveyard?
I say 1 year, 2 months.
Re: (Score:2)
Re: (Score:2)
You can of course bet on 10+ years, but I guess your odds of winning are rather slim, given their track record. And there were quite a few things that looked pretty promising too that are now littered on the graveyard of has-beens.
Lying is wrong (Score:2)
Only available in the US of A? (Score:3)
Regionalism rears its ugly head once again. Another reason for Canadians to abstain from buying US products or using US services. Costa Rica is less expensive than Hawaii too. Typical from a corporation that rips off content creators and advertisers, and through them, all of the rest of us, google has been sucking the life blood out of our economy for years
Re: (Score:2)
Re: (Score:2)
but in Canada
turn on AC commenting pls.
Use case: insurance and contract documents (Score:2)
My first reaction was to put this tool to the test. I uploaded my homeowner's and auto insurance policy documents, then asked it some questions:
- What are the coverages in the auto policy?
- What are the exclusions?
- What is the premium for each of the coverages?
It was able to neatly summarize answers to each of the questions, a lot faster than I could have done so by reading the documents.
I can see this tool being useful for summarizing any kind of contract, like leases, mortgages, title deeds, terms of ser
Re: (Score:2)
Re: (Score:2)
Like all LLMs, it's not always correct. But it does link to the source document when it provides answers, so you can confirm. And it's a lot better at scanning for specific things than a human.
Re: (Score:2)
Yep, I can see why you posted as anonymous coward. You are indeed too cowardly to own up to your response. I suggest you actually try NotebookLM and see how it performs, before you say things about it that are completely unfounded.
Whole cloth (Score:1)
Re: (Score:2)
So your resume might now match the job description. You now have 40 years of AI experanciance
um.... (Score:2)
...do people not understand that the point of note-taking is to consume something, codify it in your head, and boil it down to bullet points as a supplement to that understanding?
It's not in lieu of understanding something yourself.
Thus defeating the whole purpose of note taking (Score:2)
It's the act of taking notes that solidifies an idea in the mind. Having your robot do it simply increases your dependency on that product, and leaves those ideas inaccessible unless you are clued to specifically search for them.
Re: (Score:2)
That's assuming that the robot can even take notes reliably. (Spoiler: it can't.)
Integrating AI into tools like this implies a reliability that modes like this are simply not capable of providing. Blind trust in AI is already destroying lives. From false arrests to job loss, it's becoming a real problem.
Maybe we've misunderstood those old "warnings" from OpenAI. It's not that the technology is dangerous because of what it can do, it's dangerous because of what people think it can do.