EP 645: OpenAI’s $1 trillion IPO, Google Gemini closing ground on ChatGPT, Microsoft’s AI app builder & more AI News That Matters

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.

OpenAI finished restructuring to a for profit corporation and Microsoft's equity share got much smaller. But did Microsoft win or lose on this new OpenAI deal? More big AI questions. Will Google Gemini realistically be able to catch ChatGPT? And is OpenAI actually going to IPO at a 1 trillion dollar market cap? A lot of big questions in AI this week and we'll try to answer those and more. Because although we didn't get any huge model releases from the big AI labs this week, what we did get was some of the biggest partnership and company news in a long. Oh, and some big news if you're a vibe coder or an Excel user. And bad news if you work at Amazon. All right, let's tackle it all. Welcome to Everyday AI. What's going on y'?

All?

My name is Jordan Wilson and this is Everyday AI. We do this every single day. It's your daily live stream, podcast and free daily newsletter helping everyday business leaders like you and me keep up with everything that's happening in the AI world and, and help us all make sense of it to grow our companies and our career. So if that's what you're trying to do, awesome. Me too. Starts here with the unedited unscripted live stream podcast. But if you want to take it to the next level, make sure you go to our website at your everyday AI dot com. There you can sign up for our free daily newsletter as well as keeping up with all the other AI news happening today. So if you're brand new here, thanks for tuning in. So on Mondays we generally do this AI news that matters segment. It's where, well, we take the hundreds of big AI news stories that happen each week. Yeah, there's hundreds. And we boil it down to usually the top 10 AI news stories that matter. So let's get straight into it and let's start with probably the biggest one, something that has been rumored to be happening for more than a year. That is OpenAI's restructure. So OpenAI has completed their major recapitalization, shaping its governance and also technically deepening its partnership with Microsoft. So OpenAI announced that its non profit, now called the OpenAI foundation, holds a 130 billion dollar equity stake in the new organization, the for profit business called OpenAI Group PBC or the Public Benefits Corporation. So the foundation side, the Nonprofit now controls 26% of the for profit arm, while employees and investors hold 47%. And Microsoft's stake now previously was 49%, I think it went down to 33% and now it's 27%, which holds a value of 135 billion, at least according to the latest OpenAI reported financials. So, yeah, Microsoft stake started at 49% and then after some recent funding rounds, it went down to 32.5. But now with this new restructure, it goes to 27%. So the recapitalization cements OpenAI as a non profit with control over its commercial operations, reversing earlier plans to fully convert the organization to a for profits only. So kind of a confusing structure. But what this ultimately clears the way for, well, OpenAI had to do this and get this on the books before the end of the year in order to properly fulfill their requirements for some of their earlier funding, as they just raised $50 billion, I think, in the last two quarters. So right now OpenAI's current valuation stands at 4,500 billion dollars, making it the world's largest private company. So on the foundation side, according to reports, the foundation will commit $25 billion to accelerate health breakthroughs in AI safety research. So that's kind of how they're justifying, I guess, that this nonprofit arm is still a big part of the mission. So when OpenAI was initially founded in 2015, it was founded as a nonprofit technically, although the, the structure has gotten a little more convoluted and complicated since it's always been a non profit, which is why you've seen all of these lawsuits in the last year or two from Elon Musk, one of the original funders and founders in OpenAI. So now they're one of the more, you know, one of the bigger companies in the world and one that I think probably has the highest ceiling out of anyone, if I'm being honest. So what did Microsoft get out of this? Because I think that's what people are overlooking. Well, the partnership itself has changed. So one little footnote that didn't get a ton of coverage. Well, OpenAI will buy an additional $250 billion in Azure Cloud services, though Microsoft loses its exclusive right to be OpenAI's computer provider. So kind of a give and take there. But hey, the downside is you get a guaranteed $250 billion in future revenue from OpenAI. Also, claims of achieving artificial general intelligence, or AGI, must now be independently verified before any change to the company's revenue sharing agreements. So both companies can now work with third parties on AGI and product development, broadening their collaboration opportunities. So previously there was a lot of language with this, their current partnership on AGI and what that meant for revenue and their partnership. So now a lot of that is kind of out the window and this is actually a big deal because previously Microsoft had some exclusive rights to some OpenAI technology that OpenAI really couldn't shop around to other people. So this does expand what OpenAI is able to do do with other partners as well as cloud providers, which as you can probably tell by that 250 billion dollar figure there is a huge part of their business. So a lot of people were looking at just that one piece and saying, oh well, Microsoft lost. Right. OpenAI can go pay and partner with other cloud providers, which they are if you cover the show. We've been talking about some of their huge technically a trillion dollars in deals just between September and October. But I'll actually say Microsoft obviously won this thing. If you look back at their original investment was reportedly a little bit more than $13 billion. So if you look at this recapitalization effort and what it ultimately means, well, Microsoft technically has a larger stake than any other single entity. Right. So the foundation controls 26% of the new for profit army. Microsoft controls 27%. So 1% more, which at today's value is 135 billion. So if you look at just that, they 10x their investment. Right. But obviously I think OpenAI is going to be worth a lot more and very quickly. But aside from that 250 billion like a straight up eye watering deal there with Azure, 250 billion in future Azure cloud services. So yeah, a lot of people are saying, oh, Microsoft lost out by not holding on to their exclusivity longer. You know, OpenAI is able to go work with other cloud providers. Yeah, I don't know, I'm not going to be too Sad if Microsoft CEO Satya Nadella and saying we 10x our our investment and that investment is going to be worth way, way more in the near future. So a little bit more on Open what this means for OpenAI later in the show. But it's not just some of these obligations that they needed to secure their full round of funding. As an example from SoftBank, I believe that was $30 billion. There's actually bigger implications. All right, next piece of AI news. Amazon has already confirmed 14,000 corporate job cuts tied to AI. And that number is widely reported to be 30,000 people losing their jobs working at Amazon, marking its Largest AI driven reduction to date. So a big shift here as Amazon focuses not just on generative AI but also its cloud services to achieve generative AI. And also a kind of shift from going from OPX to capex or in other words laying off tens of thousands of employees and investing more in AI infrastructure, which it looks like is one of Amazon's short and medium term goals. So a lot of big cuts across Amazon and you know we did an entire episode on this last week, episode 643, so if you want to know more, make sure to go listen to that episode. But essentially Amazon's corporate reductions focused on mid level and admin roles across a lot of different departments at their hq. So these are white collar roles across HR devices, advertising, prime video and one of the hardest hit divisions was their hr, also called the PXT or People Experience technology, where that department is expected to get cut by 15% globally. So this is not an over hiring correction from the pandemic because through a couple of rounds Amazon had already reportedly addressed this and they called this out. Well, they said this is because of AI. And one thing that not a lot of people were kind of paying attention to this. This isn't a surprise because last year Amazon announced that they had saved through their own generative AI tool, Amazon Q just in one use case, more than 4,500 developer years annually on one use case. So when that came out, not a lot of people were talking about it. In summer of 2024 I said yeah, this is going to lead to some widespread job cuts and they're going to attribute that to AI. And yes, here we go. Beyond that, there was also some reporting out of the New York Times that Amazon also plans to hire 600,000 fewer warehouse workers over the next eight years with a heavier focus on AI robotics. So yeah, aside from just cutting, already confirmed 14,000 cuts to AI, that's expected to get to 30,000 by next year. Aside from that, Amazon, according to the New York Times report, is looking to hire 600,000 fewer workers over the next eight years in their warehouse. So Amazon admittedly saying they want to operate like the world's largest startup. So yeah, getting that phase of the company kicked off by, well, axing 30,000 people. All right, on to a company that can't get AI right. So according to reports, Apple is actively considering even more additional AI app providers for system level integration, according to reports from CNBC and Bloomberg. So Apple CEO Tim Cook told CNBC that quote, our intention is to integrate with more people over time and by people talking about AI companies, underscoring Apple's openness to bringing more third party AI tools into Apple's operating systems. Because Apple Intelligence doesn't work, I kid you not. 10 minutes before I was getting ready to do my show here, I always have Text edit open, so I'm on a Mac, right? And I very you can't use Apple Intelligence because it's not intelligent. It's nowhere. I'm like, oh yeah, let me try to use Apple Intelligence for this. And it just said, oh, I have to use chat GPT for this. Anyways, Apple has already embedded chat GPT into Siri and Google Gemini support is reportedly in progress with a new Bloomberg report talking about that. So that's going to be in today's newsletter, so make sure you check that out. So what other providers are going to be coming to Apple's operating system and why and how? Well, the how part and what that might actually look like is still being developed. But some recent rumors also point to potential integrations with Anthropics, Claude or Perplexity, which would give users more of a choice of models and providers directly within Siri and Apple apps. So executives did signal that this strategy could work, saying Apple may look forward to doing integrations with different models like Google Gemini in the future. So Cook did also say that Apple is making quote unquote good progress on an AI upgraded Siri targeted for next year. So I'm sure next year when I'm still doing this show, I'll say that next year got pushed back another year. Yeah, Apple just can't get this right in. The only real Apple reports on AI are either lawsuits, delays or they're having to work with even more third party providers. So yeah, they've already had a active partnership with OpenAI where essentially Apple Intelligence just kicks queries to ChatGPT. We've seen a lot of reporting recently on essentially Apple having to rely more on Google Gemini for some certain features of Apple Intelligence. And now this new report saying it may even also be partnering with Anthropomorphic and Perplexity to bring in even more help. All right, now going to a company that I didn't expect to be developing their own AI models yet here we are. So at its CANVA World Tour conference, Canva just recently announced a new foundational AI design model that generates editable layered designs and not just flat images across posts, presentations, whiteboards and websites. So Canva said that the new AI model understands design layers and formats, letting users start with a prompt and then directly tweak objects, aiming to reduce prompt to perfection friction for non designers and pros alike. So Canva AI, which is the platform's assistant, is now available throughout the entire interface. So they're kind of bringing an assistant that you can talk to whether you're working in a standard design, a video, you know, whatever. So you can just use the at mentioned command and talk to the new Canva AI to make changes. The assistant gains new creation features including generating 3D objects and copying the art style of any design, which could speed up brand consistent content production for teams. All right, now over to Anthropic, Making some pretty big news with bringing Claude to Excel. So according to, well, the company's announcements, Anthropic is making one of its biggest moves yet into financial services, rolling out Claude AI tools that integrate directly with Microsoft Excel. Not only that, but can also connect to real time market data. So Anthropic is slowly starting to release Claude for Excel, letting financial analysts interact with its AI assistant inside their spreadsheets, which remain the core tool for Wall street and finance professionals. However, Claude for Excel is currently available to hardly no one. So yeah, it's in a super small beta which has a wait list and the plan is to only roll this out to a thousand users in the first round and those users have to either be on a maximum team or enterprise plan. So yeah, you already got to be coughing up a ton of money just to get on the wait list, but looks pretty nice. All right. Especially if this is your line of work, you know, you should be keeping an eye on and hanging, you know, tuning into everyday AI. Because once this does roll out to generally available or more broadly available, we'll obviously be talking about it. So the new Claude Sidebar Inside Excel so this is living inside of Excel. So a little different, right? This isn't a. Are you still running in circles trying to figure out how to actually grow your business with AI? Maybe your company has been tinkering with large language models for a year or more, but can't really get traction to find ROI on Gen AI. Hey, this is Jordan Wilson, host of this very podcast. Companies like Adobe, Microsoft and Nvidia have partnered with us because they trust our expertise in educating the masses around generative AI to get ahead. And some of the most innovative companies in the country hire us to help with their AI strategy and to train hundreds of their employees on how to use Gen AI. So whether you're looking for ChatGPT training for thousands or just need help building your front end AI strategy, you can partner with us too, just like some of the biggest companies in the world do. Go to your everyday AI.com partner to get in contact with our team or you can just click on the partner section of our website. We'll help you stop running in those AI circles and help get your team ahead and build a straight path to ROI on gen. Kind of the more of the agents that use Claude's models Right now you're not inside Claude AI, so this is inside Excel. It's essentially a Claude plugin and it can pop out as a sidebar. It can read, analyze and modify Excel workbooks, explain every change at the cell level, and help build financial models. So beyond Excel, select Claude models will be available in Microsoft Copilot Studio and Researcher Agent as well. So those are separate releases. We covered those a couple of weeks ago. So the company with this new release is introducing six new agent skill that are pre built workflows for tasks like discounted cash flow models, comparable company analysis and processing data room documents, and earning call analysis. So like I said, very, very limited rollout, but I think this is going to be wildly popular. So let's be honest, Microsoft's I'd say first year and a half of Copilot even in Excel did not go well. I'd say their last three months they've been crushing it. And one of the reasons might be well they're starting to partner outside of just OpenAI. Right? Their agent release about six weeks ago was the first time that they officially started using in their core Copilot products any other models aside from OpenAI. So it seems to be going well. Right? Because they're technically when you think about it like this, they're having to pay for that, right? Whereas when they're quote unquote paying for OpenAI models, they're just paying themselves because they're the largest single shareholder in OpenAI. So it's pretty telling now with this direct integration in Excel. And hey, if you are one of the 1,000 people that have access to this, reach out to me. I'd love to see a demo or maybe talk about it on the show and demo it. All right. Speaking of demoing things. Well, all you need is a text prompt to start Vibe coding to the max in Google. That's because Google announced some new updates in a complete redesign of its Vibe coding products inside Google AI Studio. So there's a new Vibe coding workflow that turns a single sentence into a working AI powered apps in minutes. So Google AI Studio users can describe multimodal Apps in plain language, such as a VO powered video generator or a Nano Banana powered image editor, whatever you can think of. You can use different Google products and build an entire app inside of Google's AI studio just with text prompts. So the system uses Gemini models to handle core code wiring, UI scaffolding and core logic, reducing the time to go from idea to prototypes. So there's also a revamped app gallery that provides a visual library of Gemini powered examples where you can not just browse ideas. Well, you can click and preview them, or you can remix them and build off them just on natural language. There's also a new and improved, and I love this, the new annotation mode inside Google AI Studio lets users click on interface elements and request changes in natural language. So, yeah, there's essentially a way that you can click and mark up your designs, multiple markups, and just say, hey, fix this. Move this left, you know, change, you know, let a user upload a photo here and use Nano Banana to, you know, do A, B and C. So very cool updates and a complete refresh in really putting the Vibe coding kind of momentum at the forefront inside AI Studio. So it's not technically any new functionality. There's a lot of newer bells and shinier whistles, but what they're doing is they're really just highlighting some core functionality that was there before, making it easier to find and adding some new features. So a lot of this was already technically available, but it was kind of hidden, it was kind of buried. So now inside AI Studio, Google is really pushing the vibe coding in the build platform form to the forefront. All right, speaking of big vibe coding expansions, Microsoft is getting in on the action. That's because they just announced a new app builder and workflow agent for Microsoft 365 copilot customers in the Frontier program, letting employees build apps, automate workflows, and create agents using natural language inside Microsoft. Copilot it. All right, let's talk a little bit about what's new, because it's actually a lot. So the app builder is new and creates working apps in minutes without database setup, grounding content in Microsoft 365 files and storing new data in Microsoft lists with share by link distribution and iterative previews on the workflows side. So workflows is new and it turns plain language instructions into automated steps across the different platforms. This one sounds pretty amazing. So simple text prompt and you can build automated workflows in natural language across Teams, SharePoint, Outlook, Planner, etc. And it shows each step in real time and allows edits in the same conversation. Also Copilot Studio inside Copilot now builds personalized work grounded. That's the big key there. Work grounded agents that tap SharePoint meeting transcripts, team chats, emails and external systems like ServiceNow and Jira. Jira, sorry. With the full studio available for advanced workflows in multi agent systems. And then as an example, in a launch scenario, Copilot could spin up an app to track milestones and campaign progress, send Monday teams updates with planar deadlines and deploy an agent to answer what's next using only company resources. So yeah, Microsoft's just aside, right? We covered their, their big winter release or no the fall release I guess about 10 days ago. But they've just been low key just putting out new updates almost every single week. It's actually been hard to keep up with Microsoft and what their putting out there because like Google they're just starting to ship a ton of updates. So no longer just on the big you know, two time, you know, two times a year updates. It's almost becoming multiple times a month where Microsoft is rolling out some pretty big updates. All right, speaking of pretty big, 650 million. Let's talk about that. So during its quarter three earnings, Google announced that the standalone Gemini app now has over 650 million monthly active users, marking some absolute nutty growth and broad consumer adoption across Android, iOS and the web. So this is also helping Google close that gap and get closer to OpenAI's market leading share, still very far ahead in different metrics here. So OpenAI's last reported metric is 800 million weekly active users. Now Google at 650 million monthly active users. But still a ginormous jump from Google. Especially when you consider a year ago they were only at 90 million and back in March they were only at 350 million. So just between March and the end of September they almost doubled their total monthly active user base which is. It's bananas. Well let's just say it's bananas because one of the reasons is they're wildly popular. Nano Banana. We talked about all of the Updates inside of AI Studio. The Gemini 2.5 Pro model, still the best in the world according to many benchmarks. We will probably be getting a Gemini 3.0 soon. Make sure you keep keep your eye on everyday AI because we're going to have some, some interviews coming out soon around that also I mean you just can't overlook the growth and the fact that they've already exceeded what CEO what the CEO wanted out of the company, which was 500 million users by the end of 2025. So, yeah, they already crushed their goal. And I might think they could catch OpenAI by the end of the year, so we'll see. But Google did credit actively credit the Nano Banana image generation and editing model, driving 23 million new Gemini app users in September alone. Also, CEO Sundar Pichai announced that Gemini 3 is coming later this year. So we did get confirmation that it is coming in 2025, signaling another round of upgrades that could further boost engagement and users. Also, Google's AI mode reported reportedly has over a hundred million monthly users in July, is now seeing 75 million daily users. And y' all don't use, don't use the AI overviews. Not good. Always make that extra click and go to AI mode. I do know that Google will eventually be making the AI mode the default mode. It is so good. So if you're not using AI mode, you should probably start using it. And we're probably going to do a dedicated, a dedicated episode just on Google's AI mode. It is that good. And they're rolling everything out into AI mode. They're bringing Nano Banana, they're bringing Canvas, they're bringing Deep Research. Right. It's really turning into almost like a version of Gemini, but also a very fast and extremely accurate version of their search engine. All right, last but, but definitely not least, you want to talk about big numbers, let's go from 650 million to a trillion. Y', all these valuations, I feel like It's. It's like Dr. Evil, right? It's like $1 trillion. Well, that's. OpenAI is reportedly preparing for an initial public offering that could target a staggering $1 trillion market valuation, making it potentially the largest IPO in US history. So according to reports from Reuters in Bloomberg, OpenAI's potential IPO could happen as early as 2026. And the company recently achieved a $500 billion valuation in October through a secondary share sale, signaling rapid growth in OpenAI. But if successful, this IPO could raise $60 billion in outside capital funding, CEO Sam Altman's ambitious plans for advanced AI infrastructure. And if they could raise $60 billion, which it looks like they could, it would triple the previous record for a US IPO set by Alibaba at $21.8 billion. So y', all, this thing, this, like this OpenAI IPO I know people are looking at and they're like, I don't know. I'm believing it. I'M buying it, y'. All. I'm not going to go back and say I told you so, but I told you back in 2023, before Nvidia was a household name, that my mom knew what it was, right? I told you all. I said Nvidia is going to be the, the biggest and they are the most important company in the world. People laughed at me because at the time Nvidia wasn't even a top 10 company in the US by market cap. Now they're obviously the biggest company in the world with a 5 trillion dollar market cap. And I could see in the next three years OpenAI being a top five biggest company in the world. All right, that's not financial advice. That's just if you follow the writing on the wall, it looks like things are headed this way. And it did. Right? We started the today's show by talking about the restructuring, OpenAI's recapitalization and converting from a non profit to a for profit or public benefits company. And that's what lays the groundwork for OpenAI to go public. So aside from reportedly needing to finish that recapitalization to bring in or finish some of its earlier funding rounds from private investors, it also lays the way for this ipo. And in an interview late last week, Sam Altman did essentially acknowledge that an IPO was a very real option just because of the capital intensive work that they're working on. So a ton of big news, especially on the partnership side in AI. But that's not all. We got a ton. So now let's move on to the what's new and what's next. So on the front end of the show, right, usually focus on the top 10 stories. I literally plan this on my laptop every week and I start with like 50 stories and then I like rank and re rank them. So here's everything else that didn't make the top 10 cut. Some of these are new partnerships, new model releases. Some are just rumors too. All right, so here is what's new and what's next. Let's go. Gonna be fast. All right, Like I said, Nvidia became the first ever company to hit a $5 trillion market cap. PayPal is partnering with OpenAI to bring instant checkout to chat GBT for direct in chat purchases in Ask Me Anything live stream. OpenAI essentially set the deadline for AGI to 2028. Grammarly rebranded as Superhuman, a company they acquired and released an AI assistant. Superhuman Go so Grammarly is now just Superhuman. Google dropped homeli for small business marketing and it's actually really good. We might do a dedicated show on that. It's so good. OpenAI unveiled Aardvark, a GPT5 powered security researcher agent. So essentially a GPT5 agent for security just in beta right now. Notebook LM had some huge kind of brain updates. They added a 1 million token context and made chat history on by default. OpenAI also released pay as you go credit options for Codex and Sora xai's Grok released their version of Wikipedia called Grokopedia. Can't wait to not use that. Here's an interesting one. Universal Music Group Group settled with Udio, the AI music generating platform. Not just settled them because UMG was suing Udio, but they're also going to partner with an AI and create an AI streaming service. So that could be absolutely really cool. There's more. Adobe launched a Fire Fire Firefly video editor in beta with AI clips, sound, voice and upgraded image to video using different selectable models. OpenAI rolled out character cameos inside Sora so you can create cameos of your pet or different objects, toys, etc. Microsoft released the researcher with computer use feature in Microsoft 365 copilot. Big deal. Another big deal. Nvidia launched an AI factory for government to help government agencies build secure scalable AI. There is a new Wharton Human AI Research study that finds that nearly 3/4 of enterprise leaders say Gen AI is already delivering a real ROI. Yeah, MIT, this is what a real study looks like, not your fake study. OpenAI rolled out Pulse on the web cursor, dropped their V2 GitHub, launched their Agent HQ, a dashboard to manage coding agents from OpenAI, Google, Anthropic and more and more. Last but not least, OpenAI released GPT OSS Safeguard, a fine tuned model of their GPT OSS Open Source model that is made for safety models for custom content moderation. My gosh, there was a lot happening in AI news this week, but we covered the big stuff, we covered the small stuff and we're not going to stop covering anything. So if this was helpful, if you're listening on the podcast, I'd appreciate it if you click that follow button. If you're on Spotify or Apple Podcasts, we'd appreciate you following the show leaving us a rating. We'd really appreciate that. So on Mondays we go over the AI News that matters. Tuesday I usually give you a hot take on something that's happening in the world of AI. And Wednesdays we do kind of an AI at work. On Wednesdays usually going over a new feature, a new model and we do live use cases as well. And Thursdays and Fridays generally we'll just cover, you know, sometimes the the news of the day or news of the week or bring on super smart experts and interview them. So. So if you're new to the show, thank you for tuning in. Please make sure to follow subscribe to the show. Then when you're done, go to your everyday AI.com go sign up with a free free daily newsletter. We're going to be recapping all the highlights of today's show. So if you're listening on the treadmill and you miss anything, it's all going to be in the newsletter. Make sure you go check it out. Thank you for tuning in. Hope to see you back tomorrow and every day for more Everyday AI. Thanks y'. All.

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 youreverydayai.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.