Ep 793: Apple’s WWDC AI plans, U.S. Gov wants equity in Big Tech, OpenAI’s business moves and more

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.

Hot AI summer continues, and it's probably about to get weirdly interesting and spicy all at the same time. Here's what I mean. I mean, we now have an executive order on AI safety, but it's not anything like the first proposal. Apple is hours away from announcing their actual real Apple intelligence, and we'll see if they actually deliver working AI or if they'll get sued again for $250 million to failing to do so. Microsoft just released everything under the sun when it comes to AI. Codex cooked up some big business moves, and there's a new venture firm on the block looking for a piece of all of the big AI tech labs. Okay, well, it's not actually a VC firm, it's the federal government. And yeah, they're actually looking for equity in all the big AI companies. I mean, blink twice and you'll be 2 months behind where AI is at today, and you'll have no clue where AI is headed. That's why we spend hours every single day keeping up with the AI news and developments that impact you, your company, and your career. Don't get caught spending time trying to do it. That's literally what we do so you don't have to. Let's get into it. Welcome to Everyday AI. If you're new here, well, we do this every day. This is your daily livestream podcast and free daily newsletter helping business leaders like you and me not just keep up, but get ahead of all of the AI news and developments to grow your company and your career. So it starts here with the unedited, unscripted livestream podcast. But please make sure you go to youreverydayai.com, sign up for the free daily newsletter. We're gonna be recapping all of today's stories and a whole lot more. And while you're at it, make sure you please subscribe to our podcast on Apple or Spotify, wherever you listen to it. All right, so on Mondays we bring you the AI news that matters. On Wednesdays we usually go hands-on, uh, with a demo on Working Wednesdays. And then on Friday, we're just talking about new updates to different large language models. That's our, uh, Fresh Friday Features segment. And Tuesday and Thursday, we rotate a little bit. So it is Monday. Let's get into the AI news that matters. And my gosh, there is a lot. First, interesting timing here. Uh, Anthropic is warning that leading AI labs may need the option to slow down or temporarily pause the development of front frontier AI systems because the technology is advancing faster than society and safety researchers may be able to keep up with. So this warning came out late last week in a blog post from the Anthropic Institute, the company's research arm, which said AI is increasingly speeding up to the creation of new AI models and could eventually help build its own successors. So it's important to note that Anthropic did not call for an immediate pause, but it said the world should have a realistic way to slow frontier AI development if the risks begin to outpace safeguards, oversight, and public institutions. The company said any meaningful slowdown would require coordination among multiple leading AI developers and governments because a pause by just one company would do little if competitors kept moving ahead. So Anthropic framed the issue as a timing problem. Powerful technologies have been managed through coordinating agreements before, but building the trust monitoring systems and political will for that takes time. And the company warned, we don't have that long. Doesn't sound doomsday-y at all. So Anthropic's post is pretty notable because it came the same week that Anthropic filed its draft S-1 paperwork. Yeah. So that means that they said, hey, we're officially going public. And then a few days later, they're like, hey, Our models are so good, we think maybe everyone should slow down. So Anthropic pointed to its own internal data as evidence of the speed of change, saying now more than 80% of the code merged into its codebase is now written by Claude, and that its typical engineer was merging 8 times as much code per day in the second quarter of 2026 as 2024. I don't know, this isn't tongue-in-cheek., but maybe that's why their uptime's like 98% and not like OpenAI's at 99.9%, but that's just me. So critics are not taking it easy on Anthropic because they're saying that this whole verifiable slowdown plan is just a power grab to get governments to regulate their competitors, locking in Anthropic's recent momentum in the enterprise while hurting both open source and other faster labs. So critics also point out the timing. Yeah, Anthropic is kind of, according to critics, taking a fearmongering approach while talking about its model's capabilities in another attempt to drum up excitement for its upcoming IPO. So, yeah, I don't know. This one to me, I don't think you get to file your paperwork for IPO and then like a couple days later and be like, ah, you know, maybe everyone, we should pause this. You know, what do you guys think about a pause? Uh, again, just, I think pretty terrible optics, uh, from Anthropic, right? Uh, we've seen over the last year, you know, it's really turned into a, you know, kind of, uh, it's Anthropic versus OpenAI for the enterprise, right? Google, even though they've done great with models, you know, Google and Microsoft aren't as much competing for the enterprise. They're really just competing for everything across the board. And it's become pretty clear over the last, you know, 6 months to a year especially, that Anthropic and OpenAI are fighting tooth and nail to win the enterprise. So pretty interesting here, you know, as Anthropic has momentum, you know, they are probably a week or two away from releasing Mythos, you know, their next model that they've said is so scary no one can use it, even though the main bench that they led with called CyberGym was instantly defeated by Microsoft's M-Dash anyways. To me it seems just more like trying to pump the, uh, the, the IPO, uh, train. That's what it seems like to me. But, uh, regardless, I think it's something that's gonna be grabbing some headlines this week, so it's important for you to know. All right. Uh, next, speaking of AI safety and development, President Trump signed an executive order this week asking leading AI companies to voluntarily submit their most powerful models for federal testing up to 30 days before public release, marking a pretty big shift toward more federal attention on AI security risks. So the order that was just signed directs federal agencies to create benchmarks for measuring AI models' cyber capabilities, including whether systems can identify or exploit software vulnerabilities. So the White House also wants agencies to build an AI cybersecurity clearinghouse to review and share information about vulnerabilities, which could help government teams respond faster when AI-related security problems are found. So the final order is narrower than an earlier version, which reportedly would have given the government up to 90 days to review advanced AI models before the companies would release them. And then obviously, we talked about this last week, that was essentially cut down by David Sachs, among others. But mainly, right, I think a lot of the big tech companies said that they're really only working with a 30 to 60 day timeline. Uh, so, you know, uh, essentially submitting a model to the government 90 days before your intended release doesn't always make sense because half the time companies aren't 100% sure what they're even going to be working on in 90 days. Uh, so it's important to note that this executive order is voluntary. It's not an actual law. It's not legislation. Uh, President Trump signed an executive order, so the order just says that it does not create a mandatory licensing, preclearance, or permitting system for releasing AI models. The move just reflects kind of tension also inside of the Trump administration, which has generally favored a lighter touch on AI regulation but has faced some growing concern that advanced models could be used to strengthen cyberattacks, especially when you— and I've been talking about this and go listen to my, you know, 2026 prediction series that we ran, I think, in late January. I do think whether it's this year or next year, I think most of the world will realize that the new wars are going to be fought in AI, specifically on cyber. So it does make sense that the federal government wants to have some sort of understanding of what's going on across the board. And I do think it's probably a good compromise. This might be one of those instances where the federal government, the AI labs, and the general public may have all won. However, this is voluntary, right? So that's another key thing to keep in mind. So the order affects just the major AI developers such as OpenAI and Tropic and Google. But any binding rules on AI model releases would still likely require action from Congress. So, yeah, this is just more of a, you know, let's all shake hands and play nice and let the federal government know just how powerful these models are because the importance of this is, well, because this is in 3 to 6 months where the United States adversaries are going to be distilling this information. So it's no secret that China's AI models are for the most part distilled from frontier labs here in the US. So I think that this is probably the biggest kind of implication of this is, yes, the federal government can say they want this for safety, and I'm sure that is some aspect of it. But I think if nothing else, they want to know what ultimately models will be distilled from this. And so they essentially want to know what weapons their adversaries are going to be using in 6 to 12 months against them on the cyber end. So I think that's something I haven't heard anyone talk or write about, but I think that's ultimately what the White House is wanting here. And if that is the case, right, obviously I think there's great, uh, you know, again, from a U.S. side, I'm talking about it from the, um, from that aspect. So I think there are great, uh, pieces to this executive order if the big labs follow it. But more than anything else, I do think this is a national security, uh, kind of hope more than anything else. 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+ 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 3 ways to jump in. Number 1, go scroll back to the first one in episode 691. Number 2, 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. All right, Microsoft just did the Oprah and they said, you get a car, you get a car, you get a car. But instead of getting a car, you just got AI. You got AI everywhere. So Microsoft just wrapped up its annual Build developer conference in San Francisco and announced new models, new capabilities in Copilot, AI-powered hardware, and more. So the most noteworthy announcements probably were the new Autopilots category. And so that is essentially Microsoft's version of agents that work for you versus just a Copilot that you would prompt. And they are starting this and kicking it off with an Autopilot called Scout, which is designed to work in the background across Microsoft 365 tools like Teams, Outlook, OneDrive, SharePoint, Calendars, and Files. So Scout is meant to move Copilot from a tool that answers prompts into something closer to a proactive a workplace assistant handling tasks such as meeting prep, scheduling conflicts, follow-up reminders, task handoffs, and work coordination. Or in other words, you know, the majority of the work that knowledge workers do and don't like. So the catch, at least right now, is Scout is a little bit limited just to, uh, private preview customers in Frontier and who also have a GitHub Copilot license. So, you know, if you hear this and you're like, oh my gosh, I want this Autopilot thing that can go out and scout all my information and do half my work for me. Well, you do have to be in the private preview in a frontier organization with a GitHub Copilot license. So that's not all. Microsoft also signaled that the Copilot super app is coming this summer, as Microsoft CEO Satya Nadella mentioned it only briefly in his keynote. Microsoft did spend a little bit more time, though, highlighting their new in-house MAI models, including their new state-of-the-art-ish MAI Thinking One. That's a 35 billion active parameter reasoning model built for tasks like long context work, math reasoning, and code generation. Plus, they have a new MAI CodeOne Flash for coding workflows. So not really frontier level just yet. I'd say it's probably about 6 to 8 months behind, which you could look at a company like Microsoft and be like, okay, you know, they have, you know, Azure server racks and, you know, GPUs and NPUs and TPUs coming out of their armpits. You know, why isn't their brand new model amazing? And, you know, at the actual frontier level, I mean, you have to remember they started from scratch., and I said the, the same thing about Meta, right? So people looked at Meta's Muse Spark and they're like, oh my gosh, Meta spent umpteen billion dollars on this and, you know, couldn't even crack the top 5 on frontier models. And you could say the same thing about Microsoft's MAI models. But again, to go from zero to even a top 10-ish, or to go from zero to a model that's, you know, maybe 6 to 8 months behind, when you're hearing me say that out loud, you might think that's not impressive, but it's actually extremely impressive. So, and obviously companies like Meta and Microsoft, you know, once they do get these models, I think their ability to scale, right, essentially just throw more compute at something that kind of works. Uh, so again, I wouldn't write Meta or Microsoft out. Uh, but that wasn't the only thing Microsoft announced. They also, which we covered briefly on last week's AI News That Matter show because it came out, uh, this was in Computex. Uh, so it came out like an hour before our Monday show, but they did also partner with NVIDIA, uh, to push out some new class of AI hardware, including the NVIDIA-powered RTX Spark laptops and the Surface RTX Spark Dev Box, uh, aimed at developers and companies that want to run AI tests locally instead of just sending everything to the cloud. No pricing on those yet, but again, I think when the pricing comes out, people are going to be like, What the— 'cause yeah, I'm guessing they're gonna be at least $4,000 or $5,000. Yeah, for a laptop, but extremely powerful. So no, and I did do an episode on this on Thursday, I believe. Yes. So if you do wanna know more about Microsoft Build Conference, I was there covering it. So make sure you go listen to episode 791. All right. Speaking of conferences, I will not be covering this one, but we are hours away from Apple into their Worldwide Developer Conference. So yeah, when I said it is hot AI summer, it is humid and sweaty when it comes to AI releases right now. So Apple is kicking off its Worldwide Developer Conference or WWDC here in a couple hours as it stock nears record highs, and investors are watching closely to see if the company can finally deliver anything that actually works in artificial intelligence. And when I say that, that's not an exaggeration. Apple has faced multiple class action lawsuits and had to pay out $250 million in damages because they said 2 years ago that they were developing all these cool AI features that they announced at WWDC, and obviously they never happened at all. So Apple literally got sued, had to pay $250 million because they promised something that they couldn't deliver. So although there's probably rumors about, you know, all these great AI things that Apple is going to release today, I would take it with a $250 million grain of salt in front of it. But the centerpiece of the announcements today is expected to be a major overhaul of Siri. Apple's voice assistant with features like a standalone chatbot app, on-screen awareness, multi-step command handling, and integration with outside AI models such as Google's Gemini. So yeah, more or less, you know, what the reporting that we've been hearing is that Apple wasn't able to get anything done internally. Uh, so they're having to pay, uh, Google to use their models to fix all of their AI shortcomings. And you know what, when I said this, like, I don't know, 2 or 3 years ago, everyone's like, this Jordan guy's dumb. He doesn't know what he's saying. Apple's going to completely dominate at AI. And obviously I was right on that. So we'll see what actually happens. And if this, uh, WWDC will work this time, obviously Google Gemini has fantastic models, right? But it's all about how will Apple implement them? And the eyes are going to be on Tim Cook. So this marks Tim Cook's final conference as CEO, making it a legacy moment as John Ternus prepares to take over. So analysts say that Apple must show its Apple intelligence can actually do more than add 1 1. That's not what they're saying. They're just saying that they need to actually put out some AI that maybe kind of works to see if this will be a true iPhone upgrade driver. Yeah, my thought is Anything that actually works and is actually worthwhile out of this conference in terms of AI will require a brand new piece of hardware. That's how Apple normally does it. So Apple's approach, though, has always focused on privacy and device-level AI and partnerships rather than just massive internal investments in data centers and paying billions of dollars for, you know, or millions of dollars for researchers. So, yeah, instead, they're just looking to partner with others because there are reports aside from the new Siri, just, uh, the new Siri Cloud, uh, or sorry, the new Siri could just partly run on Google Cloud and NVIDIA chips. So yeah, that one's even interesting as well, you know, not running on Apple's private cloud. So, uh, we'll see. Let me know, let me know, uh, Spotify comments, livestream comments, just say Apple yes or Apple no, right? Should, should we cover this on Tuesday? Should we just give you a quick rundown? I'm personally interested, but mainly it's because for me, right, I have an iPhone. I, you know, on this live stream, I'm on my Mac. It does impact my daily work, right? I would say if we could get 30% of what's available inside of ChatGPT or Claude or Gemini, I think the world will become more productive, right? But Apple has showed continually to mess up everything when it comes to their Apple Intelligence. Probably the biggest thing was being, uh, cocky enough to call it Apple Intelligence, right? Trying to rebrand artificial intelligence. I literally said that right after WWDC24, and I'm like, this won't end well. And it obviously didn't end well. All right, something that may actually end well because this was announced and it's out and it's really good. OpenAI made some major money moves on the business side with some Codex announcements. So OpenAI revealed that it will soon also bring Codex capabilities into the ChatGPT app everywhere, making it easier for businesses to access agenda coding and productivity tools in a single platform. So here's essentially what happened. OpenAI had a, big event called their Intelligence at Work. It was a livestream where they went over, uh, kind of 6 new, uh, plugins that they released, 2 big new features, and they kind of started to lay out this super app concept, right? Uh, you know, which will be Codex. Codex is the super app. They've already kind of announced that, right? But they are going to be bringing together, kind of elements of ChatGPT and Codex and starting to fuse that. So they did, uh, talk about that a little bit, but you know, nothing solid, uh, at the Intelligence at Work livestream on what that, uh, merge might actually look like. But what we did get is 6 new Codex plugins. And if you haven't tried these, my gosh, try that. They are scary good. Yeah. So combine the new Codex plugins with the fact that The Codex interface is just really, really good and intuitive, and I think you'll see how your work can start to change. So each one of the plugins is designed to help specific business functions such as sales, data analytics, creative production, uh, product design, public equity investing, and investment banking. So those are kind of the 6 different categories. Uh, so a lot to get to. There's also, you know, a bunch of new plugin partnerships with big companies like Figma, Canva, Shutter, uh, Shutter, Stock, PixArt. There's new, you know, the data analytics plugin I know is going to be a really popular one because it integrates directly with platforms like Snowflake, Databricks, Genie, Hex, Tableau, a lot of other things. The sales plugin, right, that connects to Salesforce, HubSpot, Slack, Outreach, and more. So there was the— those were the 6 new plugins. And then there was kind of 2 new features, I think one bigger than the other., so the smaller one would be annotations and then the bigger one were, uh, would be sites. So annotations kind of allows users inside of Codex to refine specific parts of documents, spreadsheets, and slides by just annotating, right? So there's a new feature, you know, so if you're working on an artifact inside of Codex, you know, if you have it, you know, I don't know, scour the web and create a PowerPoint for you and there's little certain things you wanna tweak in there, you can literally just like draw on it. And tell Codex what to change. Really cool. Uh, but I think the one that is actually going to be bigger is the new Sites feature. Sites feature, I don't know how popular this will be, but I know how popular this could be, right? The potential for this is absolutely enormous. So if you've used something like Lovable or Bolt, that's what Sites is. It is absolutely bonkers. Like, I still can't believe that more people aren't talking about this. I know it's only, you know, been like 5 days since it was announced, but essentially, I mean, this brings database, auth, security, everything, right? This isn't just like a, oh, here's a nice little vibe-coded site that you can share with your coworkers. It is a literal working app, right? Everything you need. Database. Auth, everything is in there, right? And then you can share it securely across your organization. So great tool for teams. And yeah, I don't know, should we cover this a little bit more? I did, you know, two pretty in-depth Codec shows a couple of weeks ago. And it seems like there's some appetite, but you all got to let me know if, you know, we should do something like Sites, if we should talk about Sites or talk about the 6 plugins., you let me know. All right, y'all, I wasn't kidding. More hot AI summer. So SpaceX is making headlines with a very likely record-breaking IPO set for this Friday. So yeah, it will very likely be the largest initial public offering in US history. So SpaceX has filed a fixed IPO price of $135 per share. That one's interesting. And plans to shell sell about 555 million shares, aiming to raise about $75 billion. So in case you missed the news last month and you're like, why are we talking about SpaceX, the rocket company, on an AI show? Well, that's because SpaceX now technically owns xAI and its AI chatbot Grok following that merger on May 6th. Also, SpaceX, and via XAI, has a couple of kind of valuation pending with some pieces of, I guess, some transactions they have that are pending. So let me tell you what that means. So SpaceX's valuation could hit $1.7 trillion if the EchoStar Spectrum and the Cursor transactions are completed. Which if they do happen that way, that would make SpaceX the 7th largest US company by market cap, even ahead of its— I don't know if you'd call Tesla the bigger brother now or the smaller brother, but one of Elon Musk's other companies. So SpaceX will debut on the NASDAQ ticker this Friday with the call symbol SPCX. All right, so make sure you look out for that. But, uh, this approach is a little unusual, offering a fixed price on the, uh, stock instead of a price range, which is what you would typically see, uh, for these tech IPOs. Uh, so Musk also, like I said, merged SpaceX with XAI earlier this year. At that time, the combined entity was $1.2 trillion. So between the Cursor, uh, acquisition and just the pace of AI, right? And obviously, um, SpaceX has signed, uh, some deals with Anthropic for compute. And there's some other reports out there that they're going to be working with some other big companies, essentially selling compute. So, uh, you know, saying they're, they're going to send rockets and, you know, send all their compute to space so it's unlimited, right? We'll see what actually happens there. But regardless, this Friday we likely to see the biggest IPO ever and one of the new big AI tech titans on the block. All right, speaking of companies going public and equity and things like that, here's an interesting one. When I first saw this, I think I saw this breaking on Twitter, and I'm like, is this real? And then I read it. There was some, uh, some kind of leaked audio, and I'm like, oh, this real, but it's worth talking about. So according to reports, the Trump administration is actively deciding whether the US government should take direct financial equity stakes in leading AI companies, a move that could fundamentally reshape the relationship between the government and the tech sector. So President Trump did confirm in an audio interview that his team is exploring ways for the American public to, quote unquote, essentially become a partner in the financial success of AI, signaling a dramatic shift toward public-private partnerships in high-tech industries. So the leading proposal, which was actually championed by OpenAI CEO Sam Altman, would have AI companies voluntarily transfer a portion of their shares to a government-run sovereign wealth fund, with profits potentially distributed to citizens through child-oriented, child-oriented IRAs, sometimes called Trump accounts. So similarly, Senator Bernie Sanders said this past week as well, and he proposed a more aggressive plan that would require a one-time 50% tax paid in stock from major AI firms, given the government voting rights and board representation to directly influence company decisions. So it sounded like President Trump wasn't going to go that far, but he did signal that, yes, they were looking at a partnership with the government, would technically have equity in these companies, but it would technically just be shared with residents in one way, shape, or another. I mean, it sounds kind of like a cool pyramid scheme, but not really at all, right? It's like, okay, well, if the government is signing off on it and the, the big AI tech labs are signing off on it. It's just one of those things. It's like it seems maybe too good to be true. I don't really know. It did also seem like this was something that was being floated out there, maybe by the government as a way to put a more positive spin on the AI industry, because obviously the AI industry, between data centers popping up in local communities and people being, I think personally, extremely ill-informed about data centers' impact on local communities. Save that discussion for another day. And the more obvious and very deserved kind of quote unquote bad rap that AI can get for taking jobs, right? Because I think sometimes CEOs just blame AI and they're like, all right, well, let's just cut 5,000 jobs and blame AI and we'll call it a day, right? So it seems like the federal government is very cognizant that AI might start getting a little bit of a black eye, right? You had all the college graduates booing AI anytime the keynote speaker said the word AI. So it seems like this is like it almost like a PR play from the federal government. But I think ultimately, if people actually somehow benefit from this, I don't think people will care. Even if it's a PR campaign and people see through it, it's like, all right, well, if it puts money in the pockets even if it's a stimulus, an AI stimulus package of sorts, I think people will at least temporarily be like, yes, let's get on board. So I have no clue if this is actually going to work, if it's actually just, you know, some smoke that's being floated out there before upcoming midterms, right? I'm sure that's another piece to, you know, for certain politicians to come up before midterms here in the US with elections around the corner to say that, you know, to have this piece, right? If they get attacked on AI, like, oh, you supported a data center in your community, you know, then they'll have something to fall back on and say, oh, well, what about this? You know, we're going to be putting all this money back in the American people's pockets. So again, I have no— this is, this is fresh. I have no clue if it's actually going to happen, but I mean, President Trump did confirm that they are working on this. So it would be one of the stranger partnerships, I don't know, maybe in the history of the private-public sector. Uh, but that's if it actually happens, but we'll be following it because I'm sure it's going to be a hot topic. All right. That is all the big AI news stories for this week, but we have now what's new and what's next. This is kind of a series of smaller stories, some releases that are out, some rumors, some leaks, and all of the above. So I'm going to take a little sip from my cup here. Yeah, this is live, unedited. Here we go. All right, we're going to go through them quick. First one, first couple, maybe not as quick because I might have to opine on them. So things are going to get hotter because we have rumors here that we are probably going to get in the next maybe 2, 3 weeks all of these models. So buckle up, ready? GPT-5.6, Gemini, not 4.5, that's Gemini 3.5 Pro and Mythos Preview. Yeah, so we might get new flagship models from OpenAI, Anthropic, and Google all in the next like 3 weeks. We'll see if it's actually impacted by this 30-day process or not. But at least before that was confirmed, we did see that we were going to get models from all of the big players. All right. Next, new Cloudflare data said that AI agents are now officially generating more web traffic than humans. That's not surprising. NVIDIA did launch, and it is out now, their beefy 550 billion parameter Nimotron 3 Ultra for cheaper, faster AI agents. So yeah, that's open source. If you have like a trillion dollars, you can run it locally. Microsoft unveiled Solera, an Android-based operating system and hardware for future AI agent devices. Looked pretty cool when I saw it in person. ChatGPT did hit that record, 1 billion weekly active users. Huge. First time that's happened that fast. OpenAI made frontier models and Codex generally available inside AWS via Bedrock. That one's big. Could have made that a main story. But if your organization is AWS Bedrock only and you wanted to use OpenAI models, but you obviously couldn't, well, now you can. Big partnership between Amazon and OpenAI. Speaking of OpenAI, they just rolled out a smarter ChatGPT memory to US paid users. It is based on Dreaming V3. You won't see that Dreaming V3 anywhere, but what you probably will see is a smarter, more efficient, and better memory. Uh, Alphabet is seeking— yeah, so Google parent company seeking $85 billion in capital. Uh, so just going through a, a funding round. Yeah, Google is, as AI spending is testing investor appetite. Appetite. Airbnb CEO Brian Chesky is apparently starting an AI lab that may focus on design and UI. That one's interesting. Reports show that OpenAI's super app transition could start as soon as this month, right? OpenAI did announce, announce and confirm the super app and that it's coming soon, but we see reports that it could be this month. Some leaks show that Mythos may now be called Claude Mythos 5 as a model slug and released along the current Haiku, Sonnet, and Opus lineup. So yeah, obviously there wasn't a Mythos 3 or Mythos 4.6, so it seems like Mythos is going to sit at the high end of the lineup and it's going to start with a Mythos 5 versus like a Mythos 1 so that it falls in line with the rest of their lineup. NotebookLM teased an unreleased feature that would allow users to generate files in dozens of formats. Like PDF, Markdown, document, MP4, and more. I did have the Google team tell me something is coming today. I don't know if it's this or something else, but keep your eyes peeled. All right. Microsoft switched GitHub Copilot to token billing, and people are not happy about that, right? Another one of the big companies moves away from the subsidized billing, right, where you're technically getting thousands of usage for like $100 or $20. Now you're actually having to pay it. So people aren't happy about that, but that's how companies make money, I guess. ChatGPT can now send, not just draft emails from the ChatGPT interface. Yeah, that's pretty wild. You can literally just go through and read and respond to emails in ChatGPT. A new reported Microsoft memo said that the Scout AI assistant plan was aimed to make people addicted. Yeah, obviously the brass at Microsoft have denied, or played down that report. But yeah, apparently they wanted to make people addicted to Scout AI. Uh, Arena AI rolled out agent mode in leaderboard, ranking agents with sessions. That's really cool. Uh, GPT-5.5 is at the top of the leaderboard. OpenAI upgraded Codex to let developers preview and test iOS apps inside the app. Uh, OpenAI reported that Codex tapped topped 5 million weekly users as knowledge worker use surged. Polk became the first AI agent approved for Apple's Messages for Business. Anthropic reported Claude authored 80% of its Merge production code last month. Anthropic did file their S-1 for a possible IPO this year. Uh, Reev, I think it's Reev 2.0 and Ideogram 4.0 were both released. So Reave, in terms of the AI image leaderboard on LMARINA, they were number 2 instantly, and Ideogram is the top open model. So if you're an AI image person, make sure you go check those out. A new report said that AI became the top reason for mass layoffs in the US as tech companies announced more than 38,000 cuts in May. And Google DeepMind launched a new open-source model. So their Gemma 4, but a 12B variation for multimodal laptop with, you know, if you have a laptop with only like 16GB of RAM, this new 12B version will do ya. So that's it. A lot to cover in a little amount of time. So from Apple's WWDC later today, SpaceX going public, This Friday, big updates from Microsoft, OpenAI, and everyone else. And we're probably going to be getting models after models after models in the next 3 weeks. We're going to be covering it all every single day. So if this is helpful, let someone know about it. Please follow this, follow the podcast and subscribe on Spotify and Apple Podcasts when you're done doing that. So many instructions, right? See, you have so much free time. You can go do these things because I'm keeping up with all this. So you can go subscribe to the podcast and then go to youreverydayai.com, sign up for the free daily newsletter, and then make sure you join us 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.