How to do 10 hours of research in 20 minutes with Ai (Gemini NotebookLM)

Hey, everyone on today's show, I'm gonna blow your mind, because I'm gonna help you become one of the top 5% of AI learners and practitioners in the world, and I'm gonna do that through show you how to stack AI tools together to create super productive workflows. Today we're gonna break down how you can use Gemini deep research with NotebookLM and its sick new interactive voice feature in using this workflow. I truly believe anybody can be an expert on a topic in minutes instead of days.

They can be world class in probably weeks instead of years. This is the opportunity to accelerate your life, accelerate your career, your business. Let's get into this show.

Okay, so this is what I was doing a lot of the holiday break. Was thinking about, how do you put multiple AI tools together to do really interesting work? And I wanted to share today how I've been using a couple of Google's AI tools.

And so one of the tools that I really love is Google Gemini and its new deep research feature. I'm a very curious person. I want to get into the details of a topic, and deep research has been an amazing way to do that.

And so if you go to Google Gemini, what I'm trying to do is I wanted, I had Google Gemini do a research project on AI agents, the use of AI agents, the marketing application of AI agents. You can kind of see my prompt there, which is, I want you to do a deep research project for me on artificial intelligence agents. I want to know what they're capable of automating.

Well today, what they're not capable of automating. Well today, when should I use an agent for versus a traditional web application, and how should I think about how AI agents should be used in marketing? So if you're watching this podcast, which is all about marketing and artificial intelligence, then obviously, like, these are the kind of questions you're going to have, and these are the kind of questions I have.

And so what I would have done before this is I would have gone and done a bunch of manual research and kind of culled it and copied it together, and Gemini is doing that for me. So what you'll see right now is that Gemini is working, and it has researched over 32 websites on this topic. You can see all of them there.

Has everything attributed, and now it's in the in the midst of creating my report. And so it's going to take what I want to know. It went out and did all the work.

This is, like, a great example of AI like I probably first of all only would have gone to like, 10 or 15 of those 32 websites that wouldn't have found all 32 we're also going to assume that Google's artificial intelligence is as good or better at actually, like, doing the search queries and finding the most relevant links and articles as I am. And here it goes, it generates a report for me, and that took about five minutes before that would have taken me five hours, probably, I mean, at least a couple hours for me to go through and do everything. And so what you'll see here is that it has put together a very detailed and very robust report with all these source topics.

And what I love about it is that I can just go and open it in Google Doc, and I can create, create a shareable document to kind of share with anybody, which is very cool. So now I have this, I have this document here, and I could go read through this, I could do everything I wanted with this, and that would be cool, but I don't think it's actually as helpful as we'd want. So I want to do something different.

I want to go to a different one of Google's tools, which is Google's notebook LM tool. Hey everyone. I'm so excited to share something with you that's an absolute game changer for marketers like us.

We've put together the real playbook top marketers are using to turn AI into their secret weapon, and this guide skips all the fluff. It's packed with strategies that are actually making an impact right now. We're talking proven tactics straight from the pros at HubSpot, a 16, z and Asana.

Whether you're looking to scale your marketing operations or future proof your career, these strategies are exactly what you need. Grab your copy at the link in the description. Now let's get back to the show.

So I go over to notebook, lm. google. com, and this is what it looks like when you just come in.

It's blank, it's empty, and what you're the goal is to is to let you to create a notebook of lots of materials that you're trying to learn about or work with. And so what's cool is, let's say I want to go and upload or add sources to my notebook. The first thing I can do is do a Google Doc.

So I want to do these AI agent capabilities, which is the research that Google Jim and I just did for me, so all that deep research that we were just talking about that's now, now a notebook. Lm, like, all right, that's, that's pretty cool. I like that, but I've got some other sources that I want to add.

I'm a big fan of what our friend of the pod, Dharmesh Shah, and hubspo co founder Dharmesh Shah is do. Doing Over agents. ai and so he also has a newsletter called simple.

ai and let's say I want to add some articles from simple. ai so he's got this great email newsletter that gets sent out. Go check it out@simple.

ai to subscribe. And he posts all of his newsletters online. So I can just take the links to these newsletters, and I can go and just add them in to my notebook alum.

So I'm just going in and adding and doing that. That's pretty cool. So I got a few relevant email newsletter articles that I have then posted there.

If you didn't have the links, you could just post just post the raw text in to notebook. Lm, like, that's really, like, Okay, that seems pretty interesting. What I also want to do is, again, I said I'm a big fan of what he's doing on agent.

ai so darmesh was on my first million and he talked a lot about AI on this show. So what I'm going to take is I'm also going to take Dharmesh episode that he did on my first million and I'm going to add that as a source over here, so that my I have that my notebook, because it allows you to easily add YouTube videos as a source. And what's really cool about notebook LM is like, I can have all these sources, or I can interact with just one of these sources.

Okay, so the other thing that is new is notebookLM has come up with a brand new feature where you can interact with the podcast. So it can take all of these sources and it can generate an audio overview. And so I am going to generate an audio overview.

And while that's generating, I'm going to do a little bit of a prompt to get a summary. I would like a detailed summary from all sources that outlines the five best uses of AI agents today, five examples of how to use in your marketing and an executive summary. Okay, that's my prompt.

I'm going to hit that prompt up. It's looking at the deep research I did. It's looking at Dharmesh YouTube video.

It's looking at those, those newsletter examples that I add. And here goes the five best uses of AI agents today, automated, complex workflows. And what's really cool is it, it gives you Sources and Citations.

One of the things I think Google has done a really good job of is connecting like the AI recommendation result back with the source material. It's really good at data collection and analysis. Unsurprising is what we're using AI for right now, personalizing your marketing, customer engagement.

Totally agree, that's the number one use case we have seen in doing our work at HubSpot with AI great at content creation. Agents are great at content creation and great at acting as digital proxies. Agents can understand user preferences and handle basic tasks such as planning travel, evaluating software, Managing Schedules.

Worst you don't want agents to need common sense, real world knowledge. They don't have extensive memory. A lot of them don't have up to fully up to date information.

They don't have the creativity and emotional nuance and emotional intelligence of humans. So it's pretty, pretty clear what's good and what's bad, and what's awesome is that from a massive amount of deep research that Jim and I did and saved me a bunch of time, and then from additional resources that I had failed, that I thought were really good, I got what was a very manageable readout on what to do. Okay, so now what we have is, after a couple minutes of work, it is generated a 23 minute AI over overview where I can listen to a 23 minute kind of conversation.

Let me give you a read of what it looks like Welcome to this deep dive all about AI agents. Oh, cool. We're going to be looking at a bunch of articles today, yeah, and even a YouTube interview we found with calling out the solution.

All right, CTO of HubSpot oh, yeah, for sure, it has the context to know who Dharmesh is Oh, wow. Okay, he's calling AI agents, the next big thing. And so, like, this is like having a personal podcast.

And when notebook LM first launched, this was the feature where we're just like, wow, this is incredible. Like, I can get this amazing summary in a podcast and audio form so that I can consume on go, which is really, really cool. And then it was made cooler because deep, you could go to deep research and Gemini and get real, reach rich insights, and then get that into notebook, LM, and get that in audio form.

Now what I want to show you today is their brand new interactive mode, which is still in beta, but it allows you to interact with the show. Okay, and I can join the conversation with the two hosts. Let's see how this goes.

Hey, there. Want to share your thoughts on this. Yeah, I was hoping that you could explain to me how I should be using AI agents in my marketing strategy, based on on what you two know, oh, that's a great question, and something we can definitely dive into for you.

Yeah, we were just about to get into the nitty gritty of what AI agents are, right, and how they're totally different from things like chat bots, and then we can definitely talk about how you might be able to actually use them in your marketing. Yeah, it seems like Dharmesh, the guy we're talking about, is really excited about AI agents. Yeah, he's the founder of agent.

ai Oh, wow. Okay. He's calling AI agents the next big thing in tech, big statement, and even compared his new company to like the LinkedIn for AI agents.

Well, that's a bold statement for sure. It definitely piques my interest. Yeah, me too.

So before we get too far, let's just break down exactly what an AI agent is, okay? Because I think when a lot of people hear the words AI agent, they just think like, chat bots. Yeah, right.

But this seems like it's something totally different, right? Yeah, you're right. It's way more than just a chat bot.

Okay, so think of it's more like a digital teammate that can handle complex tasks, you know, things that require multiple steps. Gotcha? Oh, yeah, let's hear it.

So when you're talking about AI agents handling complex tasks, can you tell me more like, what are some examples of complex tasks AI agents could handle? That's a great question, and it's helpful to get into some specifics. Yeah, okay, so let's unpack this a bit so it's not just about simple one off tasks like what a chat bot can do, right?

Instead, think about things that require a sequence of steps, like planning a whole business trip, so an AI agent could handle booking flights, finding hotels and making restaurant reservations, wow. Or in marketing, they could manage an entire social media campaign. Like, what would that look like?

It could start with generating content, then scheduling posts and even analyzing engagement. Oh, yeah, uh huh. So I really am interested in that use case around an agent planning a trip that you just talked about, what specific tools or tool could I use to actually go and do that?

Now? Okay, that's a super practical question, and it's great. You want to try this out.

Yeah, let's get into that. So while there isn't one single perfect tool that does everything end to end, yet, there are some platforms that you could start experimenting with today. Okay?

For booking flights and hotels, some people are using tools that have integrated AI like Kayak or Expedia. Okay, these tools can help you find the best deals, and some even offer personalized suggestions based on your past trips. That makes sense.

Then, for things like restaurant reservations, you could try using something like open table or racy which also sometimes have aI features. Okay? So no like single AI agent to handle the entire process quite yet, not quite yet, but the idea is that these tools are already starting to become more interconnected.

So it's like we're seeing the building blocks right now. Exactly. It's like we're seeing the early versions of those multi agent networks that Dharmesh talking about.

Okay, that's a helpful analogy. And while those specific trip planning tools are not fully AI agent based yet. They give you a good idea of how an AI agent could work.

Yeah. It's cool to think that these types of tools are just going to keep evolving. Yeah.

And it is also important to note that these agents will improve as they work with you and learn your particulars, oh, yeah, like learning your preferences over time to get better at their tasks. And that's where this idea of hybrid teams that Dharmesh talks about comes in. He thinks that we're gonna soon be working in these teams where you have humans and AI agents working together, wow, and each is gonna play to their strengths, yeah, yeah.

So I really like the conversation you're having about these hybrid teams, and I'd love to understand what hybrid teams could look more like, Okay, that's great feedback, and we're glad you're finding this so interesting, yeah, so let's dive deeper into this idea of type read teams, right? So you know how we were talking about AI agents taking over those repetitive tasks, yeah? Like scheduling social media posts, exactly.

Well, that frees up human employees. Yep. Go on.

Give me like, the five most repetitive tasks that AI agents could handle today? Absolutely, that's a great question, and it's helpful to think about some of the specific tasks. Yeah, let's get into that.

So first up would be scheduling social media posts. Yeah, that definitely sounds repetitive. AI agents can handle posting across different platforms.

That's a great use case. Second would be data entry and report generation, yeah, that's super time consuming. Agents can pull data and create basic reports automatically, okay, that's a good one.

And third is sending out personalized email sequences, ah, so they can nurture leads automatically, exactly. They can tailor the messages to different users, okay? Fourth, AI, agents can also handle basic customer service inquiries, yeah, like answering FAQs, right?

This frees up human support staff for more complex issues. That makes sense. And lastly, they could do AB testing of ad copy and images.

Oh, yeah, that's super repetitive for a human Exactly. They can try different variations and analyze the results. Oh, back to our listeners.

Could you tell me what tools I could go and use for those five. Five examples that you just gave? Okay, that's a very practical question, and it's great.

You want to get your hands on some tools, yeah, so let's dive into what you could use today. So for scheduling social media posts, a lot of people are using tools like HootSuite or Buffer, yeah, those are popular choices and pretty easy to use. And then for data entry and reporting, you could check out tools like Tableau or Google Data Studio.

Okay, so those are good for the analytics side of things, right? And when it comes to sending out personalized email sequences, MailChimp and Active Campaign are great options. Yeah, those are some of the big players for email marketing.

Oh, yeah. Well, what about HubSpot for email marketing? I've heard they do a good job.

That's a great point, and it's worth exploring that option too. Yeah, we've talked about Dharmesh being the CTO of HubSpot, and HubSpot does have its own market of automation tools. Yeah, it's a pretty popular platform.

So you could definitely use HubSpot for personalized email sequences, just like you could with MailChimp or Active Campaign, and it can handle a lot of other marketing tasks too. So I think what's interesting here is that how they've built this feature is that you are a listener, and you're asking contextual questions in the moment, like you are jumping in, if you'll notice, like the very first thing I tried to do was to use it as basically like a voice chat bot, and just ask the question I wanted to know right off the top, before it got into anything. And it was basically like, nope, nope, nope.

We're gonna go ahead and play our script that we have, and when it comes to the point we're talking about that we can address your question where it really shines. And if you're like me, it's like, I'm often watching a show on YouTube, or I'm watching a podcast, and I'm like, I just have a bunch of questions. I'm like, All right, that's cool, but like, I need the next level of detail, or what about this?

And this experience is basically letting us do that, and I suspect at some point, not too far away, you'll be able to watch like your favorite podcast, and through AI and AI avatars, the hosts are going to be able to basically have an avatar and interact with you and go deeper on those individual questions, which would be pretty amazing. But right now, I think Google has built the best stack of products for research on the planet, like if you are trying to do great research. I think the combination of Google Gemini 1.

5 with deep research and notebook LM is unmatched for the ability to get a lot of information in one place, so deep research and then put that into notebook LM, and take any other sources and put that in notebook LM, query that information, but also learn in the way that works best for you. In this with this new voice beta feature, where you can actually interact with the generated audio overview. And what's really cool is you can still go, take and listen to the podcast, but just while you're working, if you're just listening in, you can just chime in with questions.

And I think, to me, where this shines is you can go from novice to a high level of expertise on a subject exceptionally fast. That's what we're talking about today. We're talking about how you go from knowing very, very little about something to being a top 20% expert in knowledge in a topic in days or weeks, not months or years.

And that is transformational in terms of your ability to find, access and process information. And these two tools have been awesome at it. I highly encourage you to go play with the new notebook LM, interactive audio feature.

I find it to be one of the better voice AI features I've used so far. So I would go check that out. I think it's, I think they've done a really good job.

I would go and use Gemini deep research and a book on together, if you were tackling any of your kind of getting the year kicked off projects where you're like, I'm just trying to learn to get up to speed. This is gonna help you get started so much faster. Thanks everybody for tuning in today's episode.