← Everyday AI Podcast – An AI and ChatGPT Podcast
Everyday AI Podcast – An AI and ChatGPT Podcast
Ep 727: 7 Huge AI Feature Updates You Likely Missed: From AI Video and Gmail to Agents
00:33:33
Loading summary
Transcript3 lines
- [00:01]
A
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.
- [00:17]
B
If you follow AI developments, you already know it's next to impossible to keep up. But I see the eyeballs and attention are generally attracted to the same few, few things. The AI news stories with huge implications, the big tech drama and the next big model. But here's the reality. So much of our day to day benefits from using AI don't really hit any of those three categories. Sure, the next big model is impactful, but it's usually the incremental steps, features and updates along the way that make a compounding difference in what we're doing. All able to collectively accomplish. And to be honest, we've kind of been dropping the ball on that part here on the Everyday AI Podcast. But that changes today as we're trying out a new segment going over some of the most impactful AI updates over the past few weeks. So today we're going to be recapping the seven huge AI feature updates that you likely missed. And so let's just get straight into it. Here's what you're going to be learning on today's show. So you're going to learn how AI is moving from chat to actually operating your work apps. How the grab the mouse controls can finally fix AI design's biggest weaknesses. How controlling Gmail and drive from your terminal unlocks real automation for humans and agents. I'm excited for that one. And how an AI council approach can reduce hallucinations using model cross checking. And last but not least, how your docs can become cinematic videos instead of static summaries. Yeah, if any of those things, you're like, wait, yeah, I want that. All right, we're going to go over it. This show is going to be a fast and a furious one. But welcome to Everyday AI. If you're new here, my name is Jordan, I'm the host and we do this thing every day. It's your daily live stream, podcast and free daily newsletter helping everyday business leaders like you and me keep up with the tsunami of updates. I tell you what matters, what doesn't, to grow your company and your career. So it starts here with the podcast, but take, take it to the next level. And to be the smartest person in AI at your company, make sure you go and subscribe to our newsletter. All right, maybe I've let you down. I've been thinking a lot about, you know, this podcast, what we cover, what we don't cover. And I realized something, that the new AI at Work on Wednesdays has been a really popular series, right? Both you all tell me. And it's, you know, usually each week it's one of our most listened to shows and so is our Monday news recap, right? But I notice a lot of the kind of large language models or big AI features don't always make the Monday news recap shows, right? Because a lot of times it's the, you know, the big job implications, the, you know, the, the, the sanctions going against these big companies and you know, the, oh, here's the next big model, right? Those kind of the things that I was talking about at the top of the show and then at the end of the AI news stories, we have the kind of what's news and what's next. So essentially any given day or, sorry, any given week, there's probably at least a dozen noteworthy AI updates. So they either end up as a very short one bullet point in the show or we choose one of those 12 and then we spend a whole episode on it on Wednesdays during our AI at Work on Wednesdays. And I'm like, that stinks. Right? Here I am with a daily podcast and I probably can't keep most of you up to date with the features that actually matter. That's why, you know, for the Wednesday show, especially over the past three months, because there's been so many useful updates, many of them free, many of them simpler than ever to create just extremely valuable output. You know, when it comes to planning my Wednesday show, I am just like, oh my gosh, what am I going to do? Well, maybe this new segment doesn't have a name yet. So if you guys like it, you can help me name it, maybe in our inner circle community. But that's kind of why I'm doing this. So let's jump in and, you know, see if this thing works. All right, this is not going to be an overly visual episode, but I am going to be sharing my screen and going over a couple things here. If you are listening on the podcast, I'm going to do my best. Just kind of describe. So here they are, the seven new AI updates that you probably missed because they weren't grabbing the headlines. All right, first, and this one's like brand, brand new. I haven't even gotten a chance to try this one yet, and I'm stoked about this one. So this is the Google Workspace cli. And if you're not a developer, you're probably even like okay, well, what does that even mean? Google Workspace cli. All right, so that's the command line interface. But essentially what this is is it allows you to interact with Google's workspace just in the terminal. Right? So this is the way that Google describes it. They say it's one CLI for all of Google Workspace, built for humans and AI agents. So it gives you access to drive Gmail, calendar in every workspace, API and 40 agent skills included. So, okay, why does it matter? Right? So obviously if you're a developer, software engineer, what I just said probably instantly resonated with you and you're like, oh my gosh, this is amazing. Yeah, it's been out for not even 24 hours and it's not getting a lot of headlines obviously because it's a command line tool. But here's why that's extremely powerful because of some of the newer updates like OpenAI's Codex, you know, Claude code, even Claude coworker. Well, those three different tools you can talk to in natural language and they can actually power and control your terminal. Right. So I've actually reached out to a couple people at Google to get some clarity on this, just because there's not a lot of info. Right. There wasn't a big blog post. It was essentially a tweet and a link to the GitHub, which I've read. But for the, for from what I understand about this, and again, like I said, this came out just fresh and that's why I'm excited to talk about it. Usually with these things, right? Because from the Wednesday show I spend so many hours. So some of these, it's more like, okay, I've used most of them. Some of them, if they're brand new, I haven't. But as far as I know, this will literally give you the ability once you kind of install this via GitHub. That's very easy, right? If you don't know this, go ask Chad, GPT or Claude or Gemini. Hey, how do I install this, you know this Google Workspace CLI in my terminal, walk me through it like I'm 12, right? And you'll be up and running in no time. But this essentially allows you to automate those repetitive Google workspace tasks. Fast, ideal for anything. Executive ops, analysts, automation, heavy teams, marketing. Right. But yeah, as far as I know, you will now be able to control your Google workspace. Right? If you have the access and ability. All that good stuff from the command line from your terminal in also from third party apps that can control the Terminal. Right. So this opens up just I think a world of possibilities. So you know, just that unified terminal tool for drive, Gmail, calendar sheets, docs, chat and admin. And I think, you know, obviously you know who's going to find value in this? Well, I think a lot of people. I think if you can get over the hump of like oh my gosh, this is living in the terminal, which is not a big deal, especially when you can talk to your terminal in natural language and codex, Claude code, you know, cowork, etc, but I mean it admins, rev ops people, analysts, founders, you know a lot of people I think are going to find some big value here. All right, let's move on to number two. So kind of related, I just said the word Codex, right. This is OpenAI's kind of vibe coding tool. Right. It's more than that, it's a full blown ide but it's really good. All right, so a lot of people are comparing OpenAI's codex to Claude code and I'd say probably not a fair comparison. I've been advocating and you know, even chatting with some people in the OpenAI team and encouraging them to market it this way. I think it actually sits between Claude code and creating Claude Cowork. The majority actually of what I use Codex for, well it's probably now more 50, 50 is things that I would be doing in Claude cowork. Right. People don't understand the skills, automations and schedules that you get inside of OpenAI's codecs is great and it's actually an extremely easy tool to use. Right. It looks just like chat GPT. Right. I kind of wish that OpenAI did something similar to Anthropic and you know how Anthropic the desktop app. So this is the new update here is obviously it's available for Windows, it's been available for, for Mac for about almost five weeks now. Right. I kind of like in Anthropics version in their desktop it three little tabs at the top it says chat cowork code. I kind of wish that it was just ChatGPT1 app and you could click between chat and Codex and I think way more people would use it anyways. It is its own dedicated app but it looks and it functions pretty much like Chat GPT and there's actually a lot of things in there, features that I would Normally go to ChatGPT for even like data analysis. I really just do that in Codex. One of the reasons why. Right. Like I talked about automations skills, things you don't have built in inside ChatGPT. But then also even simple things like, you know, being able to steer the model midway through. Right. You can send 10 prompts if it's. If you know it's a task that's going to take many hours, you can send 10 prompts and you know, click the steer button to steer it along the way, which is really cool. And aside from that, well, it's the only model or the only way that you can get the access to GPT3 codecs. All right, yes, OpenAI released GBT3 instant. And any minute now they're going to Release, you know, GPT3 thinking GPT5 or sorry, GPT53 thinking GPT53 Pro. Etc and you know, we already seeing rumors about GPT54. Regardless, Codex I think is a game changer and now it is available for Windows. So it does run natively with PowerShell support, you know. So what's new in here? Well, it's just a dedicated workspace for parallel HTML. Right. That's the best way to describe it for long running tasks and reviewing results centrally. And the biggest thing obviously is you can give it access, like I already talked about giving it access to your terminal, but you can give it access to any files on your computer, any folders on your computer. Right. One thing I'm always doing, and I've been talking about this a lot over the last couple of weeks, anytime I'm recording a show, I usually have Claude code and codecs going at the same time. And a lot of times they're just auditing each other's work. Right. Which is kind of fun for me. And it's great, I think practice on having a better understanding of, you know, where the, you know, true power is in each one. But I think this just helps people ship things faster. And it's not just shipping software. Shipping vibe coded tools. Yes, but shipping anything. Right. For me it's shipping insights from spreadsheets way faster, you know, in the ability to keep tasks separated and also reduce context switching. So who's going to find this valuable? Codecs and Windows? Well, I mean developers, product managers, tech leads, coordinating multiple work streams. Right. A lot of people are going to find huge value here. All right, next. My gosh, I freaking love Stytch. All right, you might be thinking like, what the heck is Stitch? Well, it's Google's. This is absolutely free, right? The first two. Sorry, I should have mentioned this Google workspace cli, that's free, it's open source as well. Should have mentioned that Codex not free. You do have to have a paid account right now. And then. Now let's get to Stitch. So the new update here is Direct Edits and Stitch is free for anyone with a Google account. And if you've never heard of Stytch in general, general, it's a. I don't even know how to describe it. It is one of those things that I don't understand why this. Why Stitch is not like one of the most talked about tools ever. Maybe it's just because it like really hits me and you know, some different things that I'm working on. I would call it like vibe designing, right? There's certain things in Canva that I wish would actually work. Like you can put in a prompt and it kind of stinks. So this essentially it brings the power think of like Nano Banana, right? And then turning that into a design that you could ultimately ship, right? Code, right? Like a lot of people I know are just. It's so good. Even though it's meant, I think kind of to, you know, ultimately live in a coding environment, right? A lot of people use it just for design. So you can upload, you know, photos, you know, ask for, you know, different edits. So kind of like you would in Nano Banana, but then you get an interactive canvas, right, that you can drag things around. But the new update here is the Direct Edits. So this just gives you manual control over the AI generated screens that pop up. You can edit the text, swap images, you know, target certain element updates. You can annotate things. So this is huge. And one of the reasons why I'm absolutely loving it and I think you will too, right? A lot of what I was doing in Stytch is I was having to do a lot of, you know, prompting, right? Just through dictation, you know, describing things to a T. Because you didn't have these Direct Edits capability. So now you can just, you know, select any screen, you know, because it, you can say, hey, build me an entire app based on this photo based on my website, and It'll give you 5, 6, 7, 8 screens and you can kind of choose that and have a little finite control. But then you can enter this kind of Direct Edit mode on the screen and make precise changes without having to regenerate everything every single time. That's huge. So if you have used inside of AI Studio, they have something similar. Google's AI Studio in Canvas mode, in Google Gemini they have something similar. And now Stytch has it Stitch low key. One of my Most like, like mind blowing, right? Every time I'm using it, I'm like, this is amazing, right? I kind of call it like you're playing prompt, prompt roulette. A lot of times before this, you know, just because you have to regenerate something. If you think of, you know, iterating inside of Nano Banana, inside Google Gemini, you know, before you had the, the ability to update slides in Notebook LM via Google, via Nano Banana. The same thing with ChatGPT's image gen. Right now it's much better. So who's going to find value in this? A lot of people. I mean designers, marketers, founders who are just publishing landing pages. If you're working on an app mockup, right? For me, one thing I use this for now all the time is doing front end for, you know, things that I'm vibe coding, right? I build myself software kind of all day. And you know, before Stitch it took a lot longer. You had to do a lot more prompting, you know, more screenshots, more this and that. The great thing about Stitch is you can export it as code, you know, as a zip file and then you can upload it into, you know, Claude code or Codex or Anti Gravity, whatever it is you use. All right, let's keep going. The next one S might sound like a small update, but I think this is actually a huge quality of life update. And again, unless you read our newsletter top to bottom every day, or unless you read the Chat GPT release notes, you didn't know about this. So this is small. But this new update here is inside of Chat GPT projects and this allows you to use sources from anywhere. So you do have to be on a paid plan to get full advantage of this. But here's what's new. You can add project sources from other apps, save chats, or just pasted text notes. That one I'm like, my gosh, about time, right? Just about everyone else has had that except for OpenAI's chat GPT. But that's not all. Not just the ability to paste in, you know, a wall of notes like you would inside of Notebook LM. But you can paste Slack URLs, you can paste Google Drive links, which is great. And I think, you know, one of the things here is just the ability, like for me anyways, being able to paste in Google Doc links that are dynamically generated, right? Because then you don't even have to worry about it, right? So let's just say that you have a, a, a weekly or maybe a daily standup meeting. Right. If you're use some simple automation zapier, or maybe there's a direct integration, you know, with Google Docs, with whatever, you know. If you're using Google Meet, obviously you can do this almost automatically, right? But you can have that transcript automatically without you having to do anything. Go to a certain Google Doc, as long as you've added that Google Doc to the project, right? It's like being able to talk with your daily meeting without you having to do a single thing, right? You don't have to update it every day, every week, no copy paste, do it once it's there. Uh, so who's going to find this helpful? Well, anyone that uses ChatGPT projects, right? I love ChatGPT projects. This is my kind of ongoing reminder to myself even that, you know, there's really no use. AI moves too fast to follow, but you're expected to keep up. Otherwise your career or company might lag behind while AI native competitors leap ahead. But you don't have 10 hours a day to understand it all. That's what I do for you. But after 700 plus episodes of everyday AI, the most common questions I get is where do I start? That's why we created the Start Here series, an ongoing podcast series of more than a dozen episodes you can listen to in order. It covers the AI basics for beginners and sharpens the skills of AI champions pushing their companies forward. In the ongoing series, we explain complex trends in simple language that you can turn into action. There's three ways to jump in. Number one, go scroll back to the first one in episode 691. Number two, tap the link in your show notes at any time for the Start Here series. Or you can just go to starthereseries.com, which also gives you free access to our inner circle community where you can connect with other business leaders doing the same. The Start Here series will slow down the pace of AI so you can get ahead. No reason anymore not to use projects, right? If you took our updated PPP prime prompt Polish course, you know exactly why. Because there's a lot of new features and functionality even just within projects that have changed a lot. Some things that were restricted three or six months ago are no longer restricted. So like I said, the ability to paste a Slack channel or Google Drive file or folder and add it directly as a source is huge. And just like I said, sometimes you just want to be able to copy and paste a bunch of notes, right? One thing that I'm starting to do now, a little bit, I'm a huge Notebook LM user, right? And I paste in just large, right? I'm talking like 50 pages of text at a time. So now I can go back, grab that out of Notebook LM and my sources and throw that into a chat GPT project pretty quickly because I probably copy and pasted all that content from a lot of different places. So I'm a big fan of this update. All right, next. This one did go viral when they announced it. It actually went more viral than the huge project that it was announced with. Right. Gemini 3.1 Pro. This actually got way more traction online. So this is one that maybe didn't slip under the radar too much, but I think it was more of just a blip on the radar, right. It had all the attention for like an hour. And then I'm just wondering, I'm like, is anyone actually using this? So this new update is P Melli Photo Shoot. So if you don't know Pumeli, it is a straight up powerhouse. If you are a small business owner, if you work in marketing, you know, any SMB, I think this is great for, you know, those smaller companies that maybe don't have the budget. You don't have, you know, 100 people on your marketing team. You don't have a creative agency that anytime you can just go get, you know, good photos. Well, Pomelli Photo Shoot helps you do that. So Pumeli has been out for a while. But this new photo shoot feature is amazing. So essentially it allows you to turn basic product photos into professional studio or lifestyle marketing images. So, you know, kind of showing the, the live stream audience here, an example, right? So in this example, you know, someone's just uploading a photo of a product, right? Not a great photo. Looks like something you just upload, you know, with your iPhone. Nothing special about it. And then it's going to do some automatic, you know, scanning of that photo and then it's going to suggest shot templates. From there you can kind of choose the different shot templates and then instantly you're going to have great looking photos, right? So, you know, being able to instantly take any product, you know, this is great for product photography, obviously, but I think there's a lot of other use cases. But even just thinking about product photography, right? Especially if your company sells products or services, right? So in this example it was face cream and then there's a nice, you know, white studio backdrop. Then there's, you know, one that places the product, you know, kind of on a stone table with some flowers and Then there's a model, you know, applying the face cream and then you can edit and iterate with those in natural language as well. So this is not available worldwide, but it is available in a lot of beta regions like the U.S. canada, Australia and New Zealand. So these Google ones, you do sometimes have to opt in via Google Labs and they are, you know, for the most part things in Labs are free. But the other thing, and this is not new in the Pict photo shoot, but the way P works in general is you can just upload your website and it essentially imports your business DNA, right? So you do have to, you know, give it access, Right. But you can just enter your website, it'll go scrape all the photos, your brand identity, everything. Right? So that's Pumeli in general. And you can make a lot of different, you know, valuable assets. But the new photo shoot is a little bit different and it's amazing, right? So the ability to get those studio quality creative without the expensive shoots. So who's going to find value? A lot of people, like I already said, I think small teams doing E commerce, dtc, local business marketing that have limited creative resources. This is one of those. I think it's kind of like the. The only limit is your imagination. All right, next one that I might do a show on soon because I personally am loving this one because before Model Council came out from Perplexity, I was doing this manually, right? I was actually trying to build something like this. And so, you know, luckily, you know, Perplexity did it and don't have to use my janky version. But what is Model Council? So in short, it just runs multiple Frontier models in parallel and then it synthesizes those different results into a single comparative answer. So you send one query to several models, then it combines the outputs and highlights agreements and disagreements. Right? Pretty good kind of comparison here. So, you know, Perplexity on their announcement here saying, you know, your query runs across three of the models available on Perplexity at once, such as Claude Opus 4.6GPT 5. 2 and Gemini 3. I think they have Gemini 3 one now. And then a synthesizer model reviews the outputs, resolves conflicts where possible, and gives you one answer that shows where the models agree and where they differ. So that's cool. It literally gives you a chart, you know, it gives you a finding. And then was that finding confirmed in this example on ChatGPT, on Claude and on Gemini? And then it gives you evidence for each. And then it also gives you where the models disagree, which is great because then that's where the expert driven loop comes in. Right. You need to go in and find those things. And then it also gives you unique discovery. So if you think of kind of like, you know, Venn diagrams and the crossover, it's kind of like that. Here's where everything agrees, here's where some of them agree. Here's where some unique findings. The downside right now, access. Yeah, you got to be on the Perplexity Max plan, which is crazy expensive. 200amonth. But you also get access to the Perplexity computer, which we reviewed on yesterday's show. One of the more impressive AI tools I've ever used. The downside? Yeah, it eats through credits like I eat cereal for dinner. Just ridiculous. So, I mean, why is this helpful? Well, it reduces that single model blind spot. It's better for research, strategy and high stakes decisions and just, you know, it's, it's not a way to completely get rid of hallucinations, but it is an easier way to reduce hallucinations, you know, kind of out of the box. So who's going to find value in this? Well, just about anyone. Executives, analysts, consultants, you know, anyone that needs to do that is generally using a lot of models, multiple models, thinking models. Specifically, if you need high accuracy model console is one that you probably should check out. All right, last but not least, this is one, you know, live stream audience will be able to see it a little bit more. But the new feature, this one is also hot off the presses. It is Notebook LM Cinematic Video Overviews. So the video overviews from Notebook LM are great. But now at least on the upper tier, there is the cinematic video option. So yeah, unfortunately this is only for people who have Google AI Ultra. So that is the $200 a month plan. But here's what new. What's new? So it generates cinematic video overviews from your sources, expanding beyond text summaries. So the general video overviews are still really good. Right, but there's not a lot of, you know, motion, motion graphics. And that's really what this is. So let me just kind of read this one paragraph that'll hopefully explain it a little bit better. So this is from Google's the keyword, they said. Using a combination of our advanced AI models, including Gemini 3, Nano, Banana Pro and VO3, cinematic video overviews generate fluid animations and rich detailed visuals to help you learn and engage with the topics you care about. Gemini now acts as a creative director, making hundreds of structural and stylistic decisions to best tell the story with your sources. It determines the best narrative visual style and format and even refines its its own work to ensure consistency. So looking at some of the examples of this that Google shared, I mean, it's really good, right? Some of these graphics, right? I'll just kind of play a, you know, to be able to see things animate in real time that are AI generated, like data in a graph. Really good, right? So the general video overviews, you know, you get some of that, but you don't get that motion, you don't get that high quality. And now apparently you have Gemini acting as a creative director, making hundreds of structural and stylistic decisions to get you a better version. All right, y', all, that's a wrap. This was a quicker one. So do me a favor, let me know, do you like this kind of show? If so, maybe I'll make it an ongoing thing. So maybe leave a comment if you're still listening on the live stream, just say. I don't say, what are we gonna do? Hot or not? We'll do hot. Hot or not. There you go. So just, you know, say hot if you love this. If you want something like this, maybe once a week, just say not if this is not your thing. Like I said, I, I do think, and this is just an example, seven, I think, huge AI features, right? And today we kind of COVID the gauntlet. You know, whether you're a designer, whether you're software engineer, whether you're a non technical person, whether you're, you know, in front of spreadsheets all day. I think some of these updates, like I said, some of these updates are more important than whatever the next model is going to be. And, you know, like I said, our new show on Monday, a lot of these things are just literally one bullet point. No explanation, no nothing. And then, you know, Wednesdays we choose one of them. But there's so many now, right? The pace of AI and the useful, helpful features that can greatly expand what you and your team can do. It's like every day there's multiple of them. Whereas, you know, two or three years ago, you know, it's like maybe you get one a month. Now there's multiple a day. So it's hard to keep up. So if you do like this show, I don't know, say hot, I should have came up with something better, right? This is unscripted, unedited. So, you know, let me know if this is something that you think we should do. If you have ideas, let me know. I hope this was helpful. If it was, make sure you also go check out our 2026 AI prediction and roadmap series, episode 712713, then go sign up for the free daily newsletter. We're going to be recapping today's show and a whole lot more. Thank you for tuning in. Hope to see you back tomorrow and every day for more Everyday AI. Thanks y'. All.
- [32:15]
A
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.