EP 523: OpenAI could go public, Gemini 2.5 continues dominance and 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.

And spend the week of Vibe coding new, powerful AI models and one of the biggest names in the AI game potentially going public. Yeah, that's right. Anthropic and Apple might team up for a Vibe coding duo. Google Gemini, in its new 2.5 Pro I O edition, quietly wiped the competition for the most powerful large language model in the world. And OpenAI may be restructuring its partnership with Microsoft to to potentially go public in the future. Yeah, a lot happened in the world of AI, in the world of AI, just like every other week and maybe you missed it. If so, don't worry. We're going to be keeping you up with everything happening in the AI world and a whole lot more today on Everyday AI. What's going on, y'? All? My name is Jordan Wilson and I'm the host of Everyday AI. This is your daily live stream, podcasts and free daily newsletter helping us all not just keep up with AI, but how we can use it and leverage it to get ahead and grow our companies and our careers. If that's what you're doing, then you're in the right place. If you're joining us on the podcast, like many people are, thank you. Number one. But number two, after subscribing and leaving us a rating, please go to your everyday AI.com so there you can sign up for our free daily newsletter. So we're going to not just be recapping everything we talk about in today's podcast. That's what we do every Monday through Friday in our newsletter. Recap the podcast episode, but also keeping you up to date with everything else that's happening in the world of AI so you can be the smartest person in AI at your company. So yeah, it starts here on the podcast, but you actually finish it by going to our website, reading the newsletter as well as on our website. You can go listen to more than 520 back episodes, all sorted by category. So no matter what you're trying to learn, we've already done it. We've already got the answers. It's available for free on our website. All right, let's get right to it. Enough chit chat. This is the AI news that matters for the week of May 12th. And hey, shout out to our live stream audience. Good to see you, big bogey face. And Charles and Arvin and Michael and Sandra joining on the YouTube machine. Got some some friends joining on LinkedIn Live, Brian and Graham and Marie, Jose and Joe and others. Thanks for tuning in, y'. All. Hola to Dr. Harvey Castro. So yeah, maybe you listen on the podcast. We also do this thing. It's live, it's unedited, unscripted, the realest thing at Artificial intelligence. Here's the AI news that matters for the week of May 12th. All right, first and foremost, two big names might be teaming up for a little Vibe coding. So Apple is reportedly collaborating with AI startup Anthropic to develop an AI powered Vibe coding platform that upgrades its Xcode software, but using Anthropic's Claude Sonnet model to autonomously write, edit and test code for developers. So this new platform is reportedly being tested internally at Apple with no confirmed plans yet for a public release. So the move marks a pretty big shift for Apple, which has historically relied on its own in house AI development, but is now looking like it's open to partnering with outside companies to accelerate innovation and improve internal workflows. So Apple's previous AI coding tool was called Swift Assist and it was announced last year, but it was never actually released, reportedly suffering from issues like hallucinations and slowed app development, prompting the need for external partnerships. So this potential partnership with Anthropic and their very popular Claude tools which are widely used in the software engineering coding Space, both Sonnet 3.5 and 3.7. So this partnership reflects Apple catching up potentially in the AI race following competitors like OpenAI and Google, with Apple also planning to potentially integrate Google's Gemini model alongside ChatGPT and Siri later this year. Apple continues to develop other AI driven features in house, including things that most people aren't going to use like custom emojis and writing tools which aren't that great, and a notific notification management. Yeah, that's I guess the only quote unquote Apple intelligence that I ever use is when it's like here's a group text with 82 messages and here's a two line summary of what you missed and sometimes it's accurate, but most of the times it's not. So pretty pretty big news there. Apple admittedly and reportedly they not being able to figure out kind of AI coding on its own like it had previously announced and previously hoped for with Swift Assist that actually never publicly saw the light of day. So instead we may see Apple go the partnership route with Anthropic, which seems like a smart move, especially considering our second piece of news. Yeah, it's the week of vibe coding. Two of the bigger stories, because OpenAI has reportedly acquired Wind Surf for $3 billion. Yeah. So Wind Surf, the popular AI coding tool. So now OpenAI has reportedly acquired Windsurf. So according to reports, OpenAI has purchased WindsorF, the windsurf, the second largest AI development platform, for $3 billion, signaling a major strategic move to strengthen its position among competitors like Google and also Anthropic. So this reported acquisition because. Yeah, it has not yet officially been confirmed, I believe by OpenAI or by windsurf, but widely reported. So the acquisition is valued at a 75x earnings multiple. Whoa. And it gives OpenAI access to a highly custom coding environment that could accelerate its progress toward artificial general intelligence. So Windsurf's platform is known for enabling efficient code creation, optimization and deployment of right now, different AI models, providing OpenAI with a proprietary tool that enhances AI powered coding and engineering workflows. So coding's binary nature, where code either works or it doesn't, makes it a perfect domain for precise AI feedback and iterative improvements. A fact OpenAI plans to leverage by integrating Windsurf. The deal is expected to pretty much completely disrupt this new but very trending space. Kind of the AI vibe. Coding tools in putting OpenAI on a path to directly compete with Google, which has recently kind of blown the market out of the water with its Gemini 2.5 Pro models. Also, owning windsurf could allow OpenAI to customize the platform to fit its AI models better, gain valuable user and usage data, and accelerate innovation in AI coding agents, all of which could provide a significant Competitive edge for OpenAI. This acquisition highlights the growing importance of proprietary integrated development environments. That's what IDEs are. So if you ever hear me talk about AI IDEs, these are essentially coding platforms, right? AI coding platforms, but integrated development environments. With companies increasingly investing in these platforms to drive innovation and consolidate their market positions. So a lot of people are putting out incredibly misaligned takes on this whole situation. Right? Because there's a couple of things to keep in mind. Number one, OpenAI itself has some incredible models that can code. Specifically O3, it is a world class coding model. So a lot of people have been coming with a pretty tired take that just shows that they don't know what the heck that they're talking about saying, oh, it looks like OpenAI doesn't really have a coding model. If they need to acquire windsurf for $3 billion, which is a giant face palm, it means that person has no clue what they're talking about. That's kind of like Saying, okay, this is the best race car driver in the world. Who needs a race car, right? Yes. A model still needs an IDE in order to accurately perform. And that's something that OpenAI, at least as far as we all know, has not invested anything in is. Is creating an ide. Right? So, yeah, uh, race car driver is the model itself. Race car is the IDE. And right now, OpenAI does not have a race car. So, yeah, if you hear anyone talking about that and saying, oh, looks like OpenAI doesn't really know what they're doing. They don't really have a good coding model because, you know, oh, look at their O3. Yet they can't. No, stop listening to that person. That person doesn't know what one plus one is. All right, so pretty, pretty smart move in my opinion from OpenAI. But what will be pretty interesting to see. And yeah, a lot of people don't really understand this with all these vibe coding tools like Windsurf, like Cursor, there's a lot of them, but those are two of the more popular ones, as well as lovable in Bolt, more on the web app and application side. But essentially you choose your own model. Right. So you can go into Windsurf and right now Cursor and choose your own model. Kind of some of the favorites for the last year or so have been anthropic Claude Sonet 3.5 or 3.7, although recently a lot of people have been using Gemini 2.5 Pro or Gemini 2.5 Flash, as well as a lot of people have also been using OpenAI's new O models. Right. Whether that's O4 mini high, if you don't care about money, you might be using O3. Although it's an expensive model to use, but it's highly capable. Right. So what should be interesting to see if this acquisition goes through, will Windsurf users. Well, number one, will it still be called windsurf, or will OpenAI just absorb it and make it part of their chat GPT platform? But also, will you be able to use. If it does just remain Windsurf and it remains in its current state, will users still be able to use other models? Right, because so many people right now are using Windsurf with anthropic models, with Google models with other models. So it should be interesting to see how that ultimately pans out. Live stream audience, let me know what you think about these first two stories. Yeah, a lot of vibe coding to start off the week. So Joe is saying this Open AI plus Wind Surf is a very exciting prospect. I use Chat GPT plus for Python coding regularly and it's good, but not great. Yeah, you know, you're obviously going to get best results if you're using a dedicated IDE, which is why I think that, you know, OpenAI is obviously making this moves ago or sorry, Eagle, I'm not sure on LinkedIn saying it's great to see further developments in Vibe coding. Yeah, absolutely. So if you thought that this Vibe coding thing was a phase, I mean, our biggest two stories, Apple and Anthropic reportedly teaming up. Google, you know, put out their own kind of vibe coding tool in Firebase at its Cloud Next conference a couple of weeks ago. And here you have OpenAI reportedly acquiring windsurf. So it'll be interesting to see what happens with Cursor. Right. Obviously Microsoft has GitHub, you know, GitHub Co Pilot, which is huge. So it should be interesting to see what OpenAI does. If anything, once they do acquire Windsurf, are they going to allow people to still use their own models and will anyone acquire Cursor? So yeah, we'll have to. Hey, keep keep tuning in to find out because we'll obviously be covering it. All right, our next piece of AI news, probably bad news for Google. So Apple is considering adding AI search engines to Safari, which would challenge Google's dominance. So Apple is reportedly exploring the possibility of integrating AI search like Perplexity or ChatGPT into its Safari browser, signaling a potential challenge to Google's long standing lead in the search engine market, which currently serves 4.9 billion users worldwide. So Eddie Hu, Apple's Senior Vice President of Services, revealed this during the U.S. department of Justice's Antitrust Trust trial against Google that Apple may allow users to choose from a variety of AI powered search engines within Safari in the future. And, well, why is that a big deal? Well, one of the reasons that Google has been able to completely dominate the search market for like forever is because right now they pay Apple an estimated $20 billion annually to retain the default search engine on Apple devices. So yeah, Google gets a huge leg up with billions of users by paying Apple 20 billion, a reported amount of $20 billion to make Google its default search engine inside of Safari. So if Apple did offer an alternative AI search option, that could greatly not disrupt Google's line of business, but it could help these new AI search platforms like ChatGPT, Perplex, or whoever else may end up in Safari's preferred AI search tools. It could be a huge boom for their business so Safari search traffic reportedly declined for the first time in April 2025, a trend that Q attributed to users increasingly turning to AI platforms like ChatGPT for information highlighting, changing user behavior for for search. So potential AI search engines that Apple is considering include OpenAI's chat GPT, known for its voice and memory features, perplexity, AI praise for real time citations and concise summaries, and also anthropics Claude, known for its reasoning and clarity and which also just recently integrated web access after waiting like what seemed like 100 years to do so. So this move suggests that Apple is carefully observing how generative AI is reshaping information access and may be preparing for broader AI integration across the ecosystem. So yeah, I would expect us to see some official either announcement or rumors around Apple's WWDC 2025 which is next month and like I said on the surface, bad news for Google, right? However, we could also see if Apple is going to be offering up a variety of AI powered search engines in instead of just Google Search. I would be interested to see if Google's new AI search modes, whether it's AI mode or AI overviews or even just Google Gemini, may also be offered as an alternative to kind of google.com traditional search live stream audience. What do you guys think? What do you guys think? Are you happy with kind of the default search engine on the iPhone? I don't know. I'd say I'm personally using google.com less and less. I'm using Chat GBT a lot more. I'm using Google Gemini a lot more. Microsoft copilot, right? I'm using AI search way more than I was using traditional kind of non AI search engines. I do think this is the norm, but again, I kind of live in an AI bubble. So livestream audience, let me know. Is this something you think is going to catch on? Is it maybe too soon? 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 jet. 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 of 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 youreverydayai.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 Genai. Joe is asking when is Apple going to make an AI operating system? Not just integrating AI into Mac OS, but when AI is the operating system? I don't know, I'd say that's like a decade away at least. Right. We saw all these quote unquote Apple intelligence announcements over the last couple of years from Apple. Yet we are here we are years later. Apple's AI is either pretty much non existent, right? Yeah, there's a lot of class action lawsuits against Apple for promoting this Apple intelligence that does not exist. Right. In their commercials and their marketing. So yeah, I mean I've talked about this enough. I don't think any company has fumbled the AI back harder than Apple, especially considering the resources that it has devoted to artificial intelligence. And I did think it was pretty cocky of Apple to just go ahead and try and rebrand artificial intelligence, Apple Intelligence. Yeah. And they absolutely been horrendous in terms of bringing artificial intelligence to its billions of users. Right. I, I did get a new Apple iPhone, not just for the built in AI features. I was kind of due for an upgrade anyways, but partially for the AI features and they're pretty much non existent. Yeah, it's just like more coming soon. So yeah, we've heard that the actual smarter Siri might not be until late 2026 or even 2027, so who knows? I'm not holding my breath for this anytime soon. All right. Yeah, big bogey on YouTube. Just saying. Search engines, are they still a thing? Cecilia saying Apple may be the smart holdout though given some recent skepticism. Could be, could be. I'll say this, I think serious users aren't looking at Apple seriously when it comes to AI. Right. I don't think they are. All right, our next piece of AI news. And this one could be serious. The Apple FDA, the Federal Drug Administration and OpenAI are reportedly talking about potentially teaming up to speed up drug approval process. So the FDA is actively exploring the use of AI to accelerate the notoriously slow drug approval process, which currently takes more than a decade for many new drugs, according to FDA Commissioner Marty Makari. So OpenAI has been reportedly meeting with the FDA as part of this effort, discussing projects like integrating chat GBT into FDA's research, aimed at supporting the center for Drug Evaluation in research's review work. So Jeremy Walsh, the FDA's first ever AI officer, is. Sorry, the FDA's first ever AI officer is leading these discussions, which also involve associates from Elon Musk's Department of Government Efficiency and Peter Bowman Davis, the Acting Chief AI Officer at the Department of Health and Human Services. So no formal contract between OpenAI and the FDA has been signed yet, at least not according to reports. But the agency has already completed its first AI assisted scientific review for a product signaling early adoption of AI tools in regulatory science. Former FDA Commissioner Robert Califf noted that AI has been used in review teams for several years and emphasized that AI's potential extends beyond final reviews to many internal FDA operations experts, though, urge caution, emphasizing the need for clear policy guidelines on data usage and model performance to ensure AI systems learn the right information and maintain reliability. So personally, I think this is both extremely exciting and worrisome. Here's why AI literacy is low, right? Even for people trying to keep up with large language model developments, literacy is generally low unless you are devoting multiple hours, right? Unless half of your, you know, full time job is to learn AI. Right? Because this is one of those things when it comes to artificial intelligence, when it comes to large language models, and this also applies to this scenario of the FDA potentially partnering with OpenAI, you have to have a team of people that are just learning AI, that are just there to become more literate in training the rest of the team. This is not right. AI and large language models are not something that you learn once and then apply it to the business, whether your business is a marketing company, a small business, or the fda. Right? So that's why part of it is, yes, I know that sometimes drug discovery takes more than a decade, takes tens of millions or hundreds of millions of dollars, and getting drugs to market faster could obviously save lives, it could eradicate diseases. Yet at the same time, the stakes are extremely high here when it comes to handing off something as important as drug discovery to AI. And maybe not handing off, right? We hope this is one of those cases where when we talk about augmented intelligence, right, like the best of humans working with the best of AI. But the problem I see here is teams are not dedicated or, sorry, are not dedicating employees to simply just learning the technology and then teaching it to others because the developments are too quick, right? Whether you're talking about OpenAI, Microsoft, Copilot, Google, Gemini, Anthropic, Claude Mistral, Meta, whatever. Throw out any company that puts out AI models. A lot of times you're seeing multiple updates a week, whether those are front facing or behind the scenes. Yet I don't think hardly. I'd say less than 1% of companies in the world have devoted an AI team to both learning models, to highly specialized use cases and to training other employees. Right. Everyone just wants to slap the easy button. So this is one of those that is equally exciting and completely terrifying. All right, speaking of absolutely exciting, Google Gemini has released a new version of their world leading Gemini 2.5 Pro preview. This one called the IO Edition. So Google has launched an early update to its most powerful AI model, Gemini 2.5 Pro Preview, now called IO Edition, preparing developers for the upcoming Google IO 2025 conference here in about a week. So this new version, which let me just say this, Google didn't need to do this. They didn't need to like put out a new version of their Gemini 2.5 Pro preview because it was already the most powerful large language model in the world. And Google's like, yeah, let's just go ahead and make it even more powerful, even though no one could catch them. So now this new version, and here's where I think this is important, it officially ranks first on the Web Dev arena leaderboard, outperforming competitors like OpenAI, Anthropic and Deep Seek encoding benchmarks. So yeah, now Gemini essentially has a clean sweep. So it's not only. Right, so when we look at these ELO scores, right, so this is from LM arena, you know, there's millions of votes, people go on there, they put a prompt in, get two outputs, you don't know who's who. You say this one's better. It's the old, you know, Pepsi versus Coke blind taste test. So you know, Google Gemini and their Gemini 2.5 Pro has cleaned up in terms of overall score, but there was a couple of categories that they weren't number one in. Right, There's, I don't know, I think there's something like eight to 12 categories, but now they essentially have the number one spot, not just collectively, but almost in every single category. So the new Gemini 2.5 Pro IO Edition demonstrates advanced coding skills, making smart architectural designs that resemble a senior developer's approach. An exciting one here. The model can also generate fully scalable learning applications from just YouTube videos and excels in front end web development by automating tasks such as matching colors, fonts and Margins. So yes, you can use this new model via the API, but you can also just use it inside Google Gemini. Inside Google's AI Studio or inside Vertex AI for enterprise companies. I'll say this, go use the Canvas mode. All right, so if you have a Google Gemini paid account, also if you're a student, you get like a year and a half of Google Gemini. The free version, Google Gemini Advanced for free with valid edu email. Go get that, go cash in on that and go try the newest version of Gemini 2.5 Pro. Try the Canvas mode. Okay? This is essentially it can render code. You don't need to learn code, you don't need to know anything about code. Just talk to it. Say, here's my company, here's my problems, here's a bunch of data. Make me something useful in Canvas mode. The new Canvas mode Update in Gemini 2.5 Pro I O is even better than it was in the quote UNQUOTE Normal Gemini 2.5 Pro. It's extremely impressive, mind bogglingly robust, I would say. Like I'm enjoying the more and more time that I spend with this new model using Canvas. Just the possibilities are endless. It's extremely impressive on the coding on the software engineering side. But if you are not a super geek, if you just want to, you know, go and try it out, I would tell you to just try it out on the front end. Go inside Google Gemini. Also, if you do want to know, you can use it in Google's AI studio, which is kind of more for developers. But go look at or go watch or listen to episode 514. We broke down Google's new AI studio for non technical people. I think that'll be extremely helpful. All right, yeah, big bogey just says, go Gemini. Sandra says, wow, Michael, good question says, how do I choose Canvas in Gemini? Is it automatic? So yes, good question. If you are using the Gemini app, there is a little Gemini icon or, or, sorry, a little Canvas icon or button at the bottom that you can just toggle on and off. All right, more AI coding, y'. All. You know what I told, I told everyone about this when I laid out the 2025 AI predictions and roadmap. We did this in an entire series and I said non technical people are going to be building up solutions for themselves. I mean, half of the stories this week have been about AI coding and our next story follows suit. That's because Amazon is getting into the AI vibe coding game. So Amazon Web Services is reportedly developing an AI powered code generation tool called Kiro designed to generate code in near real time by leveraging AI agents, according to reports from Business Insider. So I think it's Kiro, maybe it's Cairo, I'm not sure. Right. When these are reports, they never really come out and say the tool name. So I think it's Kiro. So Kiro is expected to support both web and desktop applications and feature multimodal capabilities, allowing it to work with third party AI agents. What could make it highly flexible for various development environments beyond co generation, Kiro could create technical design documents, identify potential issues in the code, and then optimize it, potentially streamlining multiple stages of software development. Amazon already though, offers an AI coding assistant named Q Developer, which is similar to GitHub Copilot. But Kiro appears to be a more advanced and comprehensive agentic AI coding tool with initial launch plans reportedly targeted for late June 2025. Though those plans could shift at any time. So many new AI coders. Yeah, I guess I have no choice but to cover these way more on everyday AI. Right? I. I've talked about this a little bit on the show and in the newsletter as well. I've always struggled with covering things like AI vibe coders. Right. One of the reasons is I started everyday AI for non technical people, right. And I'd say a year ago, you know, 18 months ago, when a lot of these AI vibe coding tools started to really become popular. I still think they were probably for more technical people, but now they've gotten so good and so commonplace. I would say vibe coding is not even necessarily a technical thing anymore. Right. I don't think it's just for developers. I think, and I even predicted that in 2025 non technical, everyday business leaders would be using these tools, whether it's to generate little apps for themselves, Chrome extensions, or just to create business dashboards that they can go in and use. Right. So think of all the enterprise pieces of software that you use on a day to day basis and you're like, oh man, you know, some simple tasks in here take 30 steps and it takes so long. Or you're like, man, I wish my, you know, CRM or my ERP could do blank. Well now you can just go into these tools and you know, it might not be the best version of it, but you can get a working version up, you know, sometimes just in a couple of prompts. So I do think now we're at the point where this AI vibe coding and spinning up your own apps and software, it's for the non technical people. Yeah. Michael from YouTube says, I think we're all going to have to learn to code now, at least to be able to read and debug it. Yeah, that's the good thing too. AI can help you, right? It can tell you, hey, what does this mean? You know, if, if you go and say, hey, make me, you know, something that connects my, you know, CRM data or, you know, create a business dashboard that I can just chat with that helps me, you know, better identify hot leads in my CRM as an example, you know, it will spit out the code and write a program. But you can also say, hey, what's going on here? Right? Why did you write it in Python versus, I don't know, JavaScript. Right? You can still have a natural conversation and you can say, you know, hey, write this code, render the code, let me run it, but then also explain it to a fifth grader, like, what the heck did you just do? Right? Because I do believe the better you understand the basics, the better, the better you can quote, unquote, vibe code. Right? Although you can get something, you know, duct tape up without even necessarily understanding what's going on under the hood. All right, our last two pieces of AI news. Pretty big in terms of open AI and their formation. So kind of two related stories. So first, OpenAI announced that its nonprofit arm will continue to control the for profit business that develops chat, GPT and other AI products, reversing its earlier plan to grant more independence to the for profit division. So OpenAI CEO Sam Altman stated that this decision came after discussions with civic leaders and attorneys general from California and Delaware, highlighting regulatory and public interest concerns. So the for profit segment of OpenAI will transition into a PBC or a public benefit corporation, which is a structure that balances profit making with a mission driven focus. And the nonprofits will remain the major shareholder, ensuring ongoing influence over the company's direction. So OpenAI was originally founded nearly a decade ago as a non profit research lab with the goal of developing artificial general intelligence safely for humanity. But in doing so, its market value has gone from a nonprofit to. So now all of a sudden its market valuation is $300 billion with ChatGPT, one of the most popular websites in the world. So the governance shift and, well, I guess shift back reflects tensions within OpenAI, including the lawsuit from former co founder, now chief competitor Elon Musk, who alleges the company and Altman betrayed its founding principles. Musk obviously has since launched a rival AI company, xai, and has acquired X and has launched competing AI models called Grok So this development signals ongoing challenges in OpenAI's being able to balance AI innovation, profit motives and ethical oversights, which may impact how AI technologies are governed and developed in the future. So according to OpenAI statements, retaining nonprofit control aims to ensure that AI advancements remain aligned with its broader societal interests rather than purely shareholder profits. Speaking of shareholders and OpenAI's company formation, a new and pretty breaking report that could be huge. So according to the Financial Times and our last AI news story that matters, OpenAI and Microsoft are reportedly engaged in high stake talks to revise their multi billion dollar partnership agreement, aiming to enable OpenAI's planned IPO or initial public offering, while also protecting Microsoft's access to OpenAI's advanced AI technologies. So Microsoft, which was one of the first major investors in OpenAI and has reportedly invested more than $13 billion in the company, is negotiating now how much equity it could receive in the restructured deal of OpenAI as the company moves away from its original model to a 4 profits public benefits company. But that could go public. Yeah, a lot of conflicting reports, a lot of new information, but let's talk a little bit about this partnership and why it's important. So yeah, if you didn't know, Microsoft has really benefited from being a reported 49 stakeholder in Open AI. And right now the existing contract between Microsoft and OpenAI, which was signed in 2019 and runs through 2030, governs Microsoft's rights to OpenAI's IP and revenue sharing. What? Microsoft is reportedly willing to reduce its equity stake in OpenAI in exchange for continued access to new AI developments beyond 2030. All right, so essentially they're like, hey, we'll give up less control if we can hold on for longer. And this is even more complex as OpenAI CEO Sam Altman aims to develop AGI that surpasses human capabilities and has been launching some other huge partnerships with other big companies. So as an example, OpenAI does have some other massive investments, including a $40 billion funding round that they closed recently that was led by SoftBank. So this ongoing restructuring plan includes keeping ultimate control with a non profit board, but granting that board significant equity in director nomination power in the new Public Benefit Corporation. But critics, including Elon Musk, argue this moves prioritizes profit over OpenAI's original mission to benefit humanity. So OpenAI faces regulatory scrutiny from California and Delaware authorities to ensure its restructuring complies with charitable laws and maintains its public benefits mission with Delaware's Attorney General, emphasizing the need for appropriate nonprofit control. This evolving relationship between Microsoft and OpenAI shows a little bit of signs of strain, partly due to OpenAI's independent expansion efforts such as building its own computing infrastructure and called Stargate and partnering with other firms like SoftBank and Oracle in that Stargate project, which creates tension given Microsoft's role as a major funder and technology provider. So industry Experts warn that OpenAI's need to balance investor demands with its mission presents risks as raising billions under a capped profit model is challenging. And failure to complete the restructuring could jeopardize Future funding with SoftBank, namely its own IPO prospects and its ability to compete long term with tech giants like Google. So yeah, more extremely interesting developments with OpenAI and its company formation, right? So yeah, even as we see OpenAI expand its partnership with some of the biggest companies in the world, right? Mainly that 40 billion dollar funding round and partnerships with SoftBank and Oracle, yet at the same time they're going to play in other domains as well, right? So we saw OpenAI countries announce OpenAI really trying to get in in the college student market, you know, giving away free access for about a month and a half and then, you know, like we said, reportedly acquiring Windsurfing. So really going after the AI coding market. So a lot of OpenAI news, not just in the moves they're making, but also in its corporate, in its current structuring as well as future partnership plans with Microsoft. So yes, according to this report from the Financial Times, they could be, Microsoft could be renegotiating and saying, hey, we'll take a little bit less equity if we can retain some of that partnership past 2030, so maybe taking a little less to hold on much longer. All right, a lot of news here, y'. All. Let's quickly recap the AI news that matters for the week of May 12th. So first, Apple and Anthropic are reportedly teaming up for a new Vibe coding experiment that could upgrade Apple's X code. After Apple was not able to really kick off the AI coding project on its own, OpenAI has reportedly acquired AI coding tool and IDE windsurf for $3 billion, shaking up the AI development landscape. Next, Apple is considering adding AI search engines to Safari, which would challenge Google's dominance and be a huge gain for AI startups like OpenAI. Perplexity and Claude. The FDA is apparently exploring a potential partnership with OpenAI to speed up the drug discovery and the drug approval process. Google has released a new version of its Gemini 2.5 Pro preview called the IO Edition, which achieves some top benches in coding software development etc and really just extended its lead, you know, at least against the second and third place models both from OpenAI in OpenAI's O3 and GPT4O models. Next OpenAI is reportedly developing Kiro, a new AI code generation tool aiming to keep up with all the other AI coding tools and vibe coding platforms. Next OpenAI is reversing course and saying that its non profit ARM will retain control over ChatGPT and its AI business. And last but not least, OpenAI and Microsoft are reportedly renegotiating partnership terms with that could lead the way for OpenAI to go public in Microsoft to potentially have a smaller equity slice of OpenAI's pie, but extend it for longer. That was a lot happening in the world of AI news. I hope this was helpful. If so, if you're listening on Spotify, do me a huge favor. Spotify changed their little algorithm recently. I think they're just handing over, you know, a lot more power to big tech companies, big media conglomerates. So if you want more people to be able to find unbiased, just real live information like our podcast like Everyday AI, if you could please subscribe, please leave us a review as well. So if you're listening specifically on Spotify, we'd appreciate if you could subscribe and leave us a rating as well as if you're doing that on Apple Podcasts, please share this on social media and go to your everydayai.com. so yes, we do this every single Monday through Friday. But if you need the takeaway, if you need to learn more, that's always going to happen in our newsletter where we recap each day's podcast as well as keeping you up to date with everything else that you need to know to be the smartest person in AI in your company. And while you're on our website, you can go listen to our actual podcast on our website. Watch the video version of the podcast for more than 500 episodes at your everyday AI.com. so thank you all for tuning in. We'll 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.