Everyday AI Podcast – An AI and ChatGPT Podcast - Gemini 3 Deep Dive and 3 Upgraded use cases Anyone can use

Episode Date: November 19, 2025

You ever see a new AI model drop and be like.... it's so good OMG how do I use it? 🤔Same. And yeah.... Gemini 3 is THAT good. So if you're wondering what's new, why it matters and how to use it, this episode is for you. AI at Work on Wednesdays: let's get it with the world's most poweful model in Gemini 3. Gemini 3 Deep Dive and 3 Upgraded Use Cases Anyone Can Use -- An Everyday AI Chat with Jordan WilsonNewsletter: Sign up for our free daily newsletterMore on this Episode: Episode PageJoin the discussion:Thoughts on this? Join the convo and connect with other AI leaders on LinkedIn.Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineupWebsite: YourEverydayAI.comEmail The Show: [e-mail] with Jordan on LinkedInTopics Covered in This Episode:Gemini 3 Release Overview & ContextGemini 3 AI Model Benchmark ScoresGemini 3 Advanced Multimodal CapabilitiesGemini 3 Agentic Research & Tool UseGemini Canvas Interactive Data DashboardsAI Studio Vibe Coding Experience UpgradeAnti Gravity Platform for Developer AgentsGemini 3 Pro Features in Google SearchTimestamps:00:00 "Gemini 3: Unveiling the Upgrade"05:43 "Google Gemini 3 Capabilities"08:19 "Gemini 3 Powers Business Workflows"11:16 "Gemini 3 Pro Dominates Benchmarks"16:30 "Immersive Generative UI Advances"18:09 "AI-Powered Dynamic Calculators"22:30 Interactive Trend Dashboard Creation26:03 "Streamlined AI News Workflow"29:31 "Everyday AI and Gemini Updates"31:54 Interactive Team Rankings Summary36:04 "AI Studio and Business Tools"37:20 "Evergreen AI & M&A Topics"Keywords:Gemini 3, Google Gemini, Google, AI update, Large Language Model, Multimodal AI, AI benchmarks, AI Studio, Vibe Coding, DeepMind, Agentic Research, AI Search Mode, Gemini Canvas, Data Visualization, Reasoning capabilities, PhD-level performance, Blind benchmark tests, LM arena, GPQA Diamond, Math Arena, Apex, Multimodal understanding, Video analysis, Spatial reasoning, Million token context window, Agentic encoding, Tool use, Web dev arena, Design arena, Learning tools, Flashcards, Visual aids, Long horizon planning, Workflow automation, Enterprise AI, Anti Gravity platform, Developer tools, IDE, AI mode in search, Google Vertex, Google CLI, Function calling, Real-time interactive tools, Simulation widgets, On-demand calculators, Dynamic generative UI, Send Everyday AI and Jordan a text message. (We can't reply back unless you leave contact info)

This is the Everyday AI Show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business, and everyday life. Meet Firefly AI Assistant, now live and Adobe Firefly, the All In One Creative AI Studio. Just describe what you want to create and the assistant handles the rest, orchestrating multi-step workflows across Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome. The assistant accelerates execution. Google didn't need to release Gemini 3.

Think about it. Even though they launched Gemini 2.5 Pro back about eight months ago in March, it's still widely considered one of the best multimodal models in the world. And even though in AI years, it's like 90 years old now, Google Gemini 2.5 Pro was still an overly capable and probably underappreciated model. Yet, here we are with a brand new Google Gemini 3 that was released to a lot of fanfare. So even though we maybe didn't need it, we get it, which is amazing. So on today's episode up Every Day AI, we're going to do a deep-ish dive on Gemini 3 and also

dish on three upgraded use cases that you can start using immediately today, even if you don't have a lot of experience with Google Gemini. All right, I'm excited for today's show. I hope you are too. Welcome to Everyday AI. My name is Jordan Moulson. We do this every day. It's your daily live stream podcast and free daily newsletter,

helping everyday business leaders like you and me make sense of all these AI updates. My gosh, they're raining down on us this week. So here we help you make sense of them and use them and leverage them to grow your company and your career. So if that's what you're trying to do, it starts here. But make sure to take it to the next level. Go to our website at your website. EverydayAI.com. Sign up for the free daily newsletter. We're going to be recapping the highlights

from today's show, as well as giving you all of the other AI news because, yeah, like dog years, AI years go quickly. So you got to keep up. So let's get straight into it. We're going to do a deep-ish dive on Google Gemini 3. But today is put AI at work on Wednesdays. It's our weekly segment where we go hands on with usually a new model or mode from one of the big four. Google, Microsoft, OpenAI, or Anthropics. So obviously, Google, Gemini, three drops. My gosh, we're doing that. And I think that there's some great use cases that, again, like I said, you need no

experience. We're going to do them live. All right. So on today's show, here is the rundown. Here's what we're going to do. We're going to highlight Gemini 3's groundbreaking benchmarks, and my gosh, they're good. we're going to showcase a handful of new features and modes outside of Gemini 3 that kind of slid under the radar. And then we're going to explore three new and upgraded use cases for Gemini 3, one in Google Gemini, one in AI Studio, and one in Google's upgraded and really underappreciated AI search mode.

All right. So those three demos are going to be at the end, but let me just wet the appetite a little. So we're going to go over smarter searching using Gemini. 3 thinking in AI mode for more robust AI search. We're going to do a little bit of agentic research in AI Studio with their improved tool usability. And then we're going to do a little data visualization fun. One thing I love doing in the Gemini app with Gemini 3 showcasing the improved multi-turn

and coding capabilities and using Gemini Canvas. All right. So what the heck is new in Gemini 3? Well, there's a ton. So let's go over at bullet point style. First, state of the art. Everything is state of the art. I mean, when it comes to reasoning, it tops just about any important benchmark.

So LM Arena, which blind test, two different models. You put in a prompt. You vote for which one's better. You don't know which model is which. So it is crushing it on LM Arena across just about any category. All that the dorky benchmarks, GPQA diamond, math arena, apex, I mean, everything. Also, just the nuance and depth.

That's something even though I've only been using it now for a day. It really excels at understanding context, intent, and also complexity. According to Google, we're talking PhD level performance with high scores on even advanced evaluation, like the newer and more popular humanities last exam and the ARC-AGI 2. On ARC-AGI, it's like in its own hemisphere. compared to all the other models.

So pretty impressive results on the ARC AGI test. Multimodal capabilities. Again, from Google saying best in class multimodal understanding, but also that does come across in the benchmarks as well. And it does work seamlessly across text, images, video, audio, and code. Yeah, one thing, one area on multi-modality that Google Gemini really stands out is just video. Right. So obviously you can use, you can output video and create video with VO31 inside Google Gemini,

but it also understands video, right? Not just transcripts, but if you go in AI studio, we did an episode probably like six months ago, and that's something that blew everyone's mind. Yeah, Google Gemini can see and understand video, even if there's no audio. It just understands it and is extremely impressive in doing so. Just the visualizations and the spatial reasoning has improved. the Million token context window. It's Gemini III's ability to handle extensive input for richer outputs,

agentic encoding abilities. One thing I've experienced so far in my limited use is just the ability for the agentic model to call tool use across multiple of its tool sets, right? And that's something that if you didn't look, listen to the episode yesterday. Make sure you go do that. Right after the Gemini 3 was released, we had Google's Logan Kilpatrick on the show. So you can just click back if you're listening on the podcast or go listen to episode 656. And that's something he called out. He's like, hey, it wasn't really good, right? Kind of multi, multi turn agentic, you know, on Gemini 2.5

pro. And it's really, really good now. Also, it is the vibe coding leader. That That's something that Google has leaned heavily in, especially on AI Studio. And they have the new feature, which is just the build feature, right? You literally just click build and you can vibe code to your heart's content. And you don't even, right, like unlike other vibe coding platforms, maybe IDEs, right? So actually editors, right, where you're working with an entire code base. It's not like that in Google's AI studio. It will just do it for you.

build an entire code base if needed, and you don't have to do a thing. And you can one-click deploy it to a Google property. So on the agenic side, really cool. It does top the web dev arena. It topped a couple of design arenas that I saw. So yeah, just essentially any category, Gemini 3 is topping it. All right, let's keep going here. Learning and planning.

So Google is really highlighting the interactive learning. learning tools. So yeah, all the capabilities that you probably knew and loved in Gemini 2.5, it's not like they stripped any of it away. So, you know, the flashcards, the visual aids, all that good stuff. It's there. Long horizon planning. So Google says that Gemini 3 manages complex workflows in business simulations over extended periods and shows off increased capabilities in consistent multi-step execution. And that's ultimately what we're going to want, right? As models do better, bringing in your own personal context, that's extremely important, right, from prompt to prompt, eventually with memory, chat to chat, to remember and

improve the reliability in completing multi-step tasks. I probably should have also mentioned availability in the beginning, but Gemini 3 is available on Google Gemini. So, you know, Gemini.google.com if you're using it on the front end. It's also available on the enterprise and business versions of Google Gemini. It's available in Google's AI Studio inside Google Vertex and a new and a first inside Google's AI mode in search. All right. So yeah, also it is immediate, which is great.

So you don't have to wait for it to come out. The only thing we're waiting on, I believe, is notebook LM, at least in, you know, some of the more popular, you know, programs where you may be using Google Gemini. So as far as I know, we're still using Gemini 2.5 inside Notebook LM, but in all of the other ones, like I said, Gemini app, search AI Studio, Vertex AI. Oh, also Google's command line, CLI, and also their new anti-gravity, their new kind of vibe coding IDE, in anti-gravity. So let's talk benchmarks. All right. So if you're on the live stream, I'm showing it on my screen here.

It's silly. So I'll say this. I have a screen of more than a dozen benchmarks. And it tops all of them except for one, except for sui bench verified, which I believe only tests Python. So is it the best metric? maybe not, but Claude Sonnet 4.5 is just barely ahead of a Gemini 3 Pro. But otherwise, Gemini 3 Pro is wiping the competition. And they did, yes, they did put GBT 5.1, OpenAI's new model that

was just released last week. But it's not close. That's the thing that kind of shocked me about some of these benchmarks. So humanity's last exam had a 37 percent. Whereas Sonnet 4-5 had a 13% GPT-5-1 had a 26%. Some other kind of more popular or more well-known benchmarks where it did really well as well. MMMU Pro, the multi-modal understanding and reasoning, is this kind of like the advanced ACT or SAT for large language models in 81% there compared to a 68% from Claude Sonnet and AC. 76%. So that's a huge, huge gap. And then there's this other benchmarks where I'm like,

how did they even do it? Vending bench, a newer one, you know, outperformed everyone almost by 40% MMLU similarly ahead of everyone else. So if you care about the benchmarks, we'll have them linked in our newsletter. But let's just say it wasn't close, right? Almost every single benchmark, they are at the top, and in many of them, it's not close. And actually, in many of them, Gemini 2.5 Pro was still leading or essentially in the lead, even though it was model from eight months ago. All right. Let's highlight some of the more under the radar features and updates that came along alongside

Gemini 3, but kind of were just different tools or features. So, I mean, the first one has to be. the new anti-gravity platform. All right. Kind of a cool name, right? I wasn't expecting Google to come out with, you know, an edgy, you know, name like this anti-gravity. I'm here for it.

So we'll see if it's actually something that picks up a lot of steam. But this is a new agent-first developer experience. So it is kind of a vibe coding IDEE. So the agents act as active partners autonomously planning, coding, coding, and executing end-to-end software tests. So it takes almost like a different approach than most IDEs like cursor. It actually, all the agents go off and kind of work on their own from the videos that I watch. And it kind of brings like an inbox, right?

And it's almost like these, this multi-agentic workflow. They go off and they build these things in their editor. Multiple of them can work and collaborate at the same time. And it just sends you like inbox updates, more or less. inside the anti-gravity platform. It is a program that you download for your Mac or your Windows PC. And then you can kind of review and approve different things as needed. So it does have direct access to the editor, terminal, and what's really cool is the browser.

Obviously, Google Chrome. So anti-gravity, you know, multiple agents can go in testings in a local host browser in Chrome. It can record things, take screenshots, really, really cool. So it's interesting because now Google has a lot of kind of similar tools. So there's overlap with their tools. A lot of them can do code editing in the terminal, right? So if you are a developer, if you are more on the software engineering side,

you have a lot to choose from now. So between anti-gravity, Gemini, C-LI, Jules, and then for more casual people, right, you have Opal, you obviously have AI Studio the build feature. But I mean, anti-gravity could be a serious contender if Google continues to stick resources in it. All right. Some other things. Just the extended tool use and workflow automation. So like I said, you can now with Google Gemini 3, it can actually take actions across different services.

It can organize your emails. I didn't even mention the new agent mode because right now it's only out to ultra-euvre. subscribers that are on that $200 a month plan. So even though I'm on that plan, I didn't want to highlight that because I know the overwhelming majority of our audience, audience won't have access to it. But in this new extended tool use and workflow automation kind of upgrade, you have increased and improved consistency and reliability and tool usage, which I've seen a lot in my limited use of it so far.

And then, like I said, it does have way, way more robust. agentic features in the Gemini app by using the Gemini agent. Yeah, the Gemini agent went through my mailbox and it like marked all of these or it asked me if I wanted to essentially archive all these messages and it, you know, went through years of emails. And it's like, hey, you should delete these. These aren't needed anymore. It was pretty impressive how it, how it was able to kind of keep that in mind over a long haul. I think that I'm really liking is the more immersive generative UI. So you will see kind of different versions of this, both in AI mode in search and in Gemini. So, you know, as an example in Gemini, you have a dynamic generative UI where it will

automatically build dynamic visual layouts with images, table, grids, and interactive tools. So it's not something that you can necessarily click a button and get, but depending on what you're asking for and what Gemini 3 outputs, you might see the... these kind of dynamic generative UIs, which are really, really cool. I'll probably over time figure out how to just prompt them. It shouldn't be too hard. But again, haven't had enough time to play with that in order to let you all in on that one. Next is just the real-time interactive tools and simulations in search.

So this is something I think actually perplexity has done a pretty good job on. And now we're seeing Google roll out with some similar features. So as an example, for certain topics like physics or finance, you might see live simulations in the AI mode inside Google search. So Gemini 3 can create simulations that you can interact with. So, you know, manipulating variables in a physics simulation as an example. So some things that you were maybe vibe coding before in Google AI studio, right, with sliders and interactive elements. Yeah, some of those things are actually rolling out to AI mode in search. these more real-time interactive tools.

Another example is just kind of on-demand calculators and widgets. So for practical queries like loan, budgeting, conversions, et cetera, Gemini 3 can custom code on the fly different calculators or comparison tools instantly, and you don't even have to visit external sites. Like, to me, this is wild, right? Have you ever been on one of those, like, random websites that you're like, this was definitely created like 20 years ago and hasn't been updated since? And it's just like some weird conversion calculator or something.

And you're like, what are all these pop-up ads? Yeah. So now Google is just going to dynamically generate some of these things inside AI mode in search, which is really, really cool. All right. So let's go ahead and take a quick break before we jump into some live demos. Quick word from our sponsors, timely sponsored today. This podcast is sponsored by Google.

Hey, folks. I'm Amar. product and design lead at Google DeepMind. We just launched a revamped vibe coding experience in AI Studio that lets you mix and match AI capabilities to turn your ideas into reality faster than ever. Just describe your app and Gemini will automatically wire up the right models and APIs for you.

And if you need a spark, hit I'm feeling lucky and we'll help you get started. Head to AI.studio slash build to create your first app. All right. So we went over the bullet point specs of Gemini 3. I gave you some of the more noteworthy under the radar features and new modes that kind of slipped through the cracks. So now let's get live. Let's go over some use cases. So again, we're going to be going over smarter searching using Gemini 3 thinking in AI mode, some agentic

research improvements and improved tool use ability in AI studio, as well as some multi-turn encoding capabilities in Gemini using Gemini. I'm a nice canvas for data visualizations. All right. Hopefully, live stream audience, you can see my screen here. We're doing this live, but don't worry, things could break, right? And if you are listening on the podcast, if you ever want to see the video version, go to your everyday AI.com.

Click episodes. Yeah, we have a library of now like 660 episodes. You can watch them all, listen to them all on our website, read about them all on our website. All right, but I'm going to do my best, even for our podcast audience to explain what's going on. So first, I am in Google Gemini, the app, right? So Gemini.com in your browser, the front end of Google Gemini. One thing to call out before we start, a little confusing.

All right. So you might be wondering, wait, where's Gemini 3? Because you used to be able to toggle the model, or at least the versions of it in the upper left hand corner where it says Gemini. So this, at least for now, is a little different. So you need to look in the bottom right hand corner. So by default, it's going to say fast, which is Gemini 2.5. All right?

So I hope that Google updates that and labels it, but just know if it says fast, which is the default, that's Gemini 2.5. So you are going to want to click and click thinking. So that's actually not thinking to deep think mode. That's different. So thinking just means that it's using. uh, Gemini 3 pro, a little confusing. All right, let's go ahead and get this prompt started. Uh, again, we're doing it live. I'm going to tell you what's happening here, um, as Gemini 3 goes to work. So I uploaded my podcast stats. All right. So I want you first to think, what data do you have?

What is the biggest mountain of data that you have that you hate, but is probably maybe useful. All right. So if you're a data analyst or data scientists, you probably love all the data. You're like, give me the numbers. Give me more numbers. I like data, but I have too much of it. I kid you not. I use Gemini Canvas literally all the time just to help me better. Number one, understand.

Number two, clean. Number three, visualize data because I have way too much of it. So in this prompt, I just uploaded some of my podcast stats. I said analyze my podcast stats and then build me a dashboard in Canvas mode. So I went, I made sure to select Gemini 3 in the lower right hand corner. And then in the kind of options toggle, I made sure to click Canvas mode. So I said in Canvas mode, show me the top 10 obvious trends, top 10 hidden trends, 10 biggest growth opportunities and 10 episodes I should plan for in December 2020,

based on recent trends. And then I said in Canvas, make the dashboard. Whoops, it's already building the dashboard. So kick me off there. So I said in Canvas, make the dashboard extremely interactive, sortable, filterable, and useful, as if it was a full-stack, high-price SaaS application. All right.

So now Google Gemini is cooking. So while we allow it to cook like any good cook, let's move on to the next dish. All right. So now we are in Google AI Studio. All right. So in AI Studio, a little different and a little easier. to understand, oh, I'm using Gemini 3.

So in the upper right hand corner, you have your model selector. So go ahead, choose Gemini 3 Pro preview. Or if you want to have some fun, go select Nanobanana. It's a ton of fun. All right. Anyways, let's go to Gemini 3. And I'm going to go ahead and run this prompt. And this one probably won't take long.

Let me explain what I'm doing and some of the toggles that I have toggled on. which are important to know. So if you don't know AI Studio, like, where are you been, y'all? Like, I kid you not, I was actually interacting and chatting with an OpenA.I. employee on Twitter today who said that they had never used Jevina's Google AI Studio. And I thought it was satire. And they're like, no, I never have. So if you are out there and you have not used Google AI Studio, yes, they were technically

sponsor for today's podcast. But y'all, I've been using AI Studio and talking about it and ranting and raving about it for like a year way before Google ever sponsored this podcast. So FYI. If you haven't used AI Studio, you're going to look at it and be like, wow, there's a lot of options, right? Don't worry about it. It was originally built for more developers, but it is extremely easy. So like I said, in the upper right hand corner, you're just going to make sure that you choose Gemini 3 preview, and then you just type a prompt. You don't really need to mess with any of the other settings.

However, in this case, I am, because I did want to show off some of these new, more tool use and function calling improvements that have happened. So actually what I did is there's an option to ground with Google Search, so I have that turned on, and then I also have the URL context turned on. on as well. That's because let's go look at my web at my prompt. So my prompt here, this is something I wouldn't normally do. So in the same way I said, you might not have

podcast data. Think of the data that you have. Okay, think of a research project that you have. That's maybe multi-step and might require a little bit of reasoning. So in this case, I said, read the last five newsletters at read.orgyour EverydayAI.com. Yeah, if you didn't know, you can go read every single newsletter that we've ever put out. We essentially have a dedicated website for that. So I said, go read the last five newsletters at that website. Then give me 10 recent AI news stories that we haven't covered in those last five newsletters, then research them and carefully outline what each of those 10 recent AI news stories

might look like in a podcast format with a suggested title, main points, context, etc. So because I have the URL context button toggled on, it's not going to make anything up. It's not going to hallucinate. It's actually going to go to that website and I can go check the sources. But it's going to go. It's going to read them. And then it's going to use a little bit of reasoning because it's not only going to have to go out and make a tool call or a call out to Google search, but it's also going to

have to check it against those five different pages. So even me, even though. though I put the newsletter together. This would take me a long time. I forget. I don't remember what was in the newsletter four days ago. Do you know how much news happens every single day in AI? A ton. So this is not an easy task. This is an simple agentic research task. So if I scroll down, if this one's already done, and it says based on the archives of everyday AI, here are the last five newsletters as of yesterday. So today's newsletter obviously isn't out yet, followed by 10 recent stories not covered in those specific issues. Okay. So it did a good job. Here we go. It has our

last five newsletters right there. Did a good job. And then it says 10 recent AI news stories, not in the last five newsletters. And so not only that, it has links out to all of these. So one, All right. Technically, I had this in like seven newsletters ago. So maybe I should have said the last 10. All right. So it has the Apple's $1 billion bailout deal with Google Gemini. So correct, wasn't in the last five. It was further back. So technically Google AI Studio did a good job. Here's talking about Google's project sun catcher, the no bailout policy from the White House, some other things. And then it outlined all of those as well. It even gave a suggestion. title for a podcast. It gave a little bit of context, and then I gave some main points. So this is running something like this inside of Google's AI Studio with the new Gemini 3.0 preview, better instruction following, better tool calling, and just overall better job

at research and depth. So a good example, quick example. I think anyone can start to use that right now. All right. Let's get to our last one. And then we will go back and check on our our visual dashboard from our podcast stats. But let me just go into AI mode. All right. So if you don't know AI mode, you should. You can always just go to Google.com slash AI. Or if you're ever on Google's homepage, all right, you can just click AI mode.

So that's over there on the right hand side. So it's going to recommend a couple of things based on my recent search history, right? So here's a little known fact that I don't talk about a lot. I love the UNC TAR heels basketball team. So that's one of the suggested queries in Google's inside AI mode. And I'll go ahead and do that. I've been a little busy with this whole everyday AI thing. I need to catch up on my UNC men's basketball.

So by default, again, this is one of those little kind of user interface things you might not know. So you remember when I was going over the specs of Gemini 3, this is the first time. that Google has rolled this out across their entire main front-end platforms, including Google's AI mode. But it's not on my default. So after I did this, I got a very, you know, generic but good little result. It's mainly text-based, right? Created one little table, but, you know, nothing overly visual.

And then it has the sources on the right-hand side. So it gave me a quick background. You know, the University of North Carolina men's basketball team is a historically successful program, right, blah, blah, blah. Give me a current team overview, season summary so far, historical accomplishments. Great. All right, but all very typical AI mode. Here's the thing, y'all. Go ahead.

There's a little drop down. It's so little in the upper left hand corner. I'm going to zoom in on it. So you have to click thinking. All right. So even though thinking is it's not. not using technically deep think is just aligning Jevonai 3 Pro as the thinking model. So I'm just going to click thinking again.

So you'll see here it is taking a little longer. When I did the default mode, which I believe is still 2.5, it went fairly quick. It was done like a almost like a normal Google search. It maybe took an extra second. So here, it's taking much longer. We're at about 10, maybe 15 seconds so far. it says creating layout. I don't see kind of like what's happening in the background here.

So I'll take a sip. I'm guessing it should be done here in a second. I already ran this test once and it did it in probably 12 seconds. So it's taken a little longer this time. So we'll see what information we get. All right. So you'll see right away a much, well, first of all, a much better overview. So right away, it gives me kind of everything in a snapshot,

Whereas the first one, I had to read on a little bit more to see what's their current record, right? What's the team ranked? So it gave me a better, richer overall summary in that first paragraph. Also, it created this little carousel, right? It has all these visuals. It says home court, the Dean Dome, 2025, 2025, 26 action, you know, head coach, right? So it has this kind of interactive carousel with visuals right there. It still has the sources off to the right hand.

inside. But you'll see here, it's almost like it made a little, not like a website, but it did a really good job, right? It has these breakout boxes with their current record, their AP ranking, the conference with little icons. It has a timeline broken up, right? So much more visual, you know, championship history, the greatest rivalry in sports, you know, Duke versus North Carolina. although is it really the greatest rivalry if North Carolina has 145 wins and Duke has 120? And then another kind of more visual breakdown at the end. So the difference is pretty big. And this is not maybe the best example.

If you're asking, if you remember, about something like financial, right, let's go ahead and do that. Let's just go ahead and do one more kind of AI mode. So let's go back. Let's go to AI mode. And I'm just going to say Google stock price. So I want to see if I can trigger this new kind of more interactive elements. So I don't know if it's going to do it for stock prices. So the same thing.

I just went to Google AI mode. I clicked. I went to Thinking, which chooses Gemini 3 Pro. We'll give it a second here and see what it cooks up. And while we're doing that, let's actually see. Oh, nice. Our dashboard's done. I can't wait to dive into that.

Here we go. All right, cool. It made a pretty nice graph. Okay. So yeah, it did give us an intraday price trend where I can kind of hover my cursor over, has some interactive elements kind of throughout the course of the day. Again, some key statistics broken out in some nice boxes. You know, it's saying why Google's stock is moving.

So pretty cool. All right. So that was all three. go back and check on use case one all right this looks pretty good pretty impressive let's go ahead uh make it full screen so this was our uh dashboard for podcast stats my gosh this is y'all gemini two jemini two five pro was great creating visual dashboards in canvas mode this is this is better this okay this is really really good okay how can i describe this um to our audience i mean it looks like a full-on dashboard let me just say this this looks like literally enterprise software that has

a i baked into it right so it here it has a main overview trend with downloads recent episodes It gave me top performing topics. It gave me December growth target. I have an overview tab. So there's different tabs. Everything in here is interactive. It has like my top three episodes over the last 30 days. It gave me a December focus.

It said I should have more update episodes. I have a trends with the obvious top 10 trends, hidden top 10 trends. And then I have an opportunities tab. Everything's interact. active, great visuals, and then I have kind of a raw data tab as well. That is so, so good, y'all. And I use these, I use the Gemini canvas every day. I use Chad Chippy T canvas every day.

I use Claude artifacts almost every day because I hit rate limit so quickly in Claude. And I'll tell you this, wow, right? Even just for visual dashboards, I know that Google, you know, did a really good job like promoting and highlighting more what you can do on the vibe coding end, on the web design. And there's great examples of that in AI studio. We went over some of those yesterday with Logan on the show. You can create like literally games, interactive worlds, all those things. But I'm like, what do my business leaders need? Right.

I don't know. Do you want to go vibe code a website? maybe, but what if you just had a piece of enterprise software based on the 20 random tasks that you do every month? That's how I want you to think. That's what I'm showing you right here. I can go back in here, and I am going to go back in here because this ones are really good. And I've done this demo probably at least, I don't know, 10 times over the past year and a half. this one is by far the best, both from a visuals perspective, the layout, it is super clean, but also I'm just briefly looking at the insights.

I'm like, these are really good, right? Not to bore you all, but even looking at the, you know, the hidden trends. It says the long tail of security, you know, so it's showing me that in super old episode is still getting a decent number of download. So it's like, hey, this, if you talk, if you have an episode about AI and security, it can last for a very long time. It's not going to die off after a week or two like some of your other episodes are. It says M&A interest is steady. So it says that there's some M&A episodes that are performing great over the long run. So I'll probably have, you know, more people talking mergers and acquisition or investing in startups or something like that on the AI side pretty soon.

So this is really good. It did a great job. All right. So maybe, you know what? Should I share this in the newsletter? Yeah. Should I send a link to this? Maybe.

We'll see. But that's a wrap for today's episode. So I hope this was helpful. So we did the bullet points backs of Gemini 3. Just know this. They're good. All right. We showed off some of the under the radar features and new modes.

Yeah, they're good. I didn't even get into agent mode. I'll maybe do a dedicated episode on that in the future. And then you just saw some three upgraded use cases powered by Gemini 3, that Gemini 3 Pro that anyone can start using today. So if this was helpful, don't be greedy. Tell someone about it. Make sure to share this, but also please go to your everyday AI.com. Sign up for the free daily newsletter.

We're going to be recapping the highlights. So if you miss something in a today show, don't worry. It's all going to be there. Thank you for tuning in. Hope to see you back tomorrow. and every day for more, everyday AI. Thanks y'all. Meet Firefly AI Assistant.

Now live in Adobe Firefly, the Allman One Creative AI Studio. Just describe what you want to create in your own words and the assistant handles the rest, orchestrating multi-step workflows across Adobe Creative Cloud apps, including Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome while the assistant accelerates execution. Stand control with the ability to step in and refine at any time. See it today. at firefly.adobie.com.

And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going. For a little more AI magic, visit Your EverydayAI.com and sign up to our daily newsletter

so you don't get left behind. Go break some barriers, and we'll see you next time.