Everyday AI Podcast – An AI and ChatGPT Podcast - Ep 615: OpenAI going after Apple hardware, Google brings Gemini to Chrome, Microsoft CEO’s dire warning on AI and more AI News That Matters

Episode Date: September 22, 2025

OpenAI is now going after .... Apple? 🍎Google is finally bringing more AI to its browser. 🖥️And Microsoft's CEO has cast a kinda gloomy warning on AI. ⚠️And that's just the beginning. Don't waste hours each day trying to keep up with AI. That's what we do. OpenAI going after Apple hardware, Google brings Gemini to Chrome, Microsoft CEO’s dire warning on AI and more AI News That Matters -- An Everyday AI ChatNewsletter: Sign up for our free daily newsletterMore on this Episode: Episode PageJoin the discussion on LinkedIn: Thoughts on this? Join the convo on LinkedIn and connect with other AI leaders.Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineupWebsite: YourEverydayAI.comEmail The Show: [e-mail] with Jordan on LinkedInTopics Covered in This Episode:OpenAI Hiring Apple Hardware Talent and AI Hardware PushUS Government Approves Meta's Llama AI ModelsGoogle Gemini Features Rolled Out in Chrome BrowserOpenAI and Anthropic User Studies—Differences in Chatbot UsageGoogle and Partners Announce Open Source Agent Payments Protocol (AP2)Major US Tech Investments in UK AI Data Centers and InfrastructureXai Appoints Recent High School Graduate to Lead Data Annotation After LayoffsNvidia Invests $5 Billion in Intel to Co-Develop AI ChipsMicrosoft CEO Warns AI Could Make Microsoft IrrelevantTimestamps:00:00 OpenAI Challenges Apple with Talent Shift04:27 OpenAI Explores Multiple AI Devices11:11 Google Adds AI Omnibox Mode14:07 OpenAI vs Anthropic Chatbot Usage15:56 Claude Dominates Enterprise; New AI Purchase Protocol22:07 Microsoft's $30B UK AI Investment23:25 Musk's Xai Puts Teen in Charge29:52 AI Threatens Tech Giants' Relevance32:46 "Unlearn to Innovate Fast"35:07 AI Advancements and Updates Roundup37:41 "Tech Giants Drive AI Expansion"Keywords:OpenAI, Apple, AI hardware, AI-powered devices, hardware startup, IO Products, Johnny Ive, Tang Tan, chief hardware officer, Apple veterans, interface design, wearables, cameras, audio manufacturing, supply chain, end-to-end hardware capability, Luxshare, Gore Tech, iPhone assembler, smart speaker, smart glasses, wearable pin, digital voice recorder, AI form factor, recruitment wave, Apple Watch hardware team, Llama AI, Meta, U.S. General Services Administration, federal agenciesSend Everyday AI and Jordan a text messaSend Everyday AI and Jordan a text message. (We can't reply back unless you leave contact info)Send Everyday AI and Jordan a text message. (We can't reply back unless you leave contact info)

This is the Everyday AI Show, the Everyday Podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business, and everyday life. Meet Firefly AI Assistant, now live and Adobe Firefly, the All In One Creative AI Studio. Just describe what you want to create and the assistant handles the rest, orchestrating multi-step workflows across Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome. The assistant accelerates execution. Open AI is making a serious bid to be a contender and competitor in hardware, specifically against Apple.

Google has brought more AI to millions of people via its Chrome browser, and Microsoft CEO had kind of a dire warning about AI and said it could actually make Microsoft irrelevant. Yeah, a lot of big news happening this week in the world of AI. And if you missed it or if you can't keep up, don't worry, we're going to be bringing you today, the AI news that matters. What's going on, y'all? My name's Jordan Wilson and welcome to Everyday AI. This is your daily live stream podcast and free daily newsletter

helping everyday business leaders like you and me, not just keep up with the AI news, but how we can cut through the fluff, the marketing, the BS, and to actually leverage all of this information, grow our companies in our career. So it starts here with the unedited, unscripted, live stream podcast. But if you want to take it to the next level, you got to go to our website at your everyday AI.com. So there, you can not only go listen and watch to 600 plus, you know, back episodes all sorted by category, but also you should go sign up for our free daily newsletter.

We're going to be recapping today's show and everything else that you need to know happening in the world of AI. And yet we do this Monday through Friday, but most Mondays, we bring you the AI news that matters. So there's a ton this week. Without further ado, let's get into it. No, actually, shout out live stream audience. It's good to see you. Brian, joining from Minnesota. Michelle saying, good morning from the YouTube machine. Joe hanging out in Fort Lauderdale, Dennis, saying hello from New York City. Randy joining us from Louisiana. Thanks for joining us, y'all. Yeah, podcast people, come to a live stream, hang out, meet other people, ask questions. But, There's a lot of questions that I have about our first news story.

Like, how the heck did this happen? Because, yes, Open AI is now, could be a serious competitor to Apple. Let's talk about it. So, according to reports, Apple has hired more than two dozen Apple veterans since early this year, a significant talent shift that signals a serious push into AI-powered hardware. So according to new reports from the information, OpenAI has hired different Apple researchers, designers that span interface design, wearables, cameras, audio, manufacturing design, and even supply chain, suggesting that Open AI is building an end-to-end hardware capability rather than just a single product bet. So the recruiting wave follows OpenAIs May 2025 acquisition of IO products, the hardware startup co-founded by Apple,

by former Apple Design Chief Johnny I and Tang Tan, with Tan now serving as OpenAI's chief hardware officer. So yeah, notable recruits include Apple's former Apple employees, Cyrus Daniel Irani, who has 15 years at Apple, Mac Theobald, 17 years at Apple, and Eric DeJong from the Apple Watch hardware team. So, yeah, Apple literally, sorry, opening I just scooping up reportedly more than two dozen people from Apple. But specifically, we're not talking about just AI researchers. We're talking about hardware leads.

We're talking about people in supply chain. So, yeah, according to reports from the information, this signals that Open AI may be working on more than just one hardware play. So according to reports, Open AI is exploring a multiple device concept, including a screenless smart speaker, smart glasses, a wearable pin, and a digital voice recorder indicating a strategy to find the right, always available AI form factor. So we'll see if all of those actually come to fruition or if any of them do, but Luckshare, a major iPhone assembler, has also reportedly signed on.

on to handle assembly with Open AI and Gore-Tech has been tapped for components, tying Open AI's hardware supply chain directly into Apple's long-standing partners. So yeah, Open AI essentially partnering with some of Apple's biggest suppliers, according to the information, and scooping up some of their most tenured talent on the hardware side. So the first Open AI hardware products could arrive as early or as early as late 2026 or early 2027, offering a rough timeline for when consumers and developers might expect a new category of AI-first devices. And Apple is reportedly alarmed by the defections and canceled even its annual offsite meeting in China this past August to reduce poaching risk, reflecting internal concern amid

frustration over incumbent product updates and stagnant stock performance. I would not want to be Apple. I've been saying this and, you know, people can say I'm biased or whatever, but this is my interpretation of the facts. The facts are Apple has multiple class action lawsuits against it for literally promoting AI that it could not deliver. Right. And so now now not only are they not able to deliver anything on the software side when it comes to generative AI in large language models or nothing that they've actually been

at least hyping up. But now even on the hardware side, they're losing some of their, some of their more senior leaders and. hardware. We've been covering it here for the last couple of months, some of their most senior AI researchers, just people leading LLM development internally just seems like almost everyone is jumping ship. Not almost everyone, right? That might be an exaggeration, but countless people. Apple is losing high, like a high amount of its current AI talent. And yes, there's always, you know, people and talented employees jumping from company to company, right?

Like, Open AI will lose some people to DeepMind, you know, and then DeepMind will lose some people to meta, right? And it's this cycle. But Apple is losing significantly more and more than seemingly the rest of the big tech companies combined. So this specific play here from Open AI going after the AI hardware, pretty, pretty interesting. So, yeah, Randy here saying, Tim Cook better do something quick.

Yeah, something is definitely happening here, Cupertino. All right, our next piece of AI news. The U.S. government has a new AI provider, apparently, and I was not expecting this one. So according to Reuters, the U.S. General Services Administration will add meta, Lama AI to its approved tools list for federal agencies, clearing the way for government-wide

experimentation and deployment with META's models. So this signals a broader federal push under the Trump administration to imitate rate more commercial AI offerings into the government's operations with security and legal vetting handled centrally by the GSA. So the GSA procurement lead, Josh Grunb, BOM set agencies can now test Lama, which is free under GSA's assurance that it meets government security and legal standards. So Lama obviously is a multi-modal model.

So it should be interesting to see how ultimately the government uses it. So GSA has also recently approved AI tools from Amazon Web Services, Microsoft, Google, Anthropic, and Open AI. So essentially, everyone accepts X, right, which you would have thought that would have happen sooner. You know, obviously President, U.S. President Donald Trump and XAI founder, owner,

whatever you want to call him, Elon Musk, had a very entire relationship, but then they had a very public breaking up. So it looks like now the government is literally using every single AI model or it has, you know, give it a green light for all of these AI models to be used except GROC.

Oh, sad to see. Google. More AI. Like Google could have shipped anything more. Well, they did. That's because Google is now releasing Gemini in Chrome. So Google is rolling out Gemini powered features directly inside Chrome on desktop right now in the U.S. Making a bunch of new kind of agentic and AI tools available directly in the browser,

such as page summaries, cross tab synthesis, which I love, and a built-in Gemini chatbot integration available to users. So a new button. So yeah, check your Chrome browser. Make sure it's up to date. But if you are in the U.S. And if you have permission, so it's a little trickier if you're on a workspace account or a business account.

But if you are on a Gmail account, check your browser because you probably have it. There is a new button in the top right hand of Chrome that now launches the Gemini chatbot. And users can unpit it if they don't. want the feature, which addresses concerns from those who prefer the more traditional browser experience. Google also plans to add an AI mode. So what this is called, it's called the Omnibox. So in the upper right hand corner, when you launch it, it's this little Omni box that kind of does a little of everything. But Google does plan to add their new, very popular AI mode before the end of this month, giving users a button and keyboard shortcuts that triggered Gemini suggestions based on the current page without replacing regular Google.

So Gemini in Chrome can answer questions about the page you're on and pull context from across multiple tabs. That's the one I really like, which can cut research time for students, knowledge workers, and just small business teams juggling too many sources at once. Also, there are some new agenic style features that will be coming soon with Gemini expected to perform web tasks like adding items to a cart. you know, if you're on a certain web page based on what you tell it to in the background, and also presenting results for approval in early tests of similar tools elsewhere have been slow and error prone. So you might want to curb, you know, any expectations for the, you know, kind of Google Chrome and Gemini being able to agentically complete everything for you, right?

I think most of the today's agents are a little slower. But for the agentic browsers, I think they're a little better. So, I mean, we'll see once all of these are rolled out. But they were initially offered to paying subscribers in May. So I've been using the Gemini in Chrome since May on my Ultra account, on my personal Gmail. So I'm excited to get this going kind of on my work Gmail account. So live stream audience, podcast, people.

Let me just say Gemini Chrome. If you want a deeper dive into this this week. I've been thinking about it because so many people use Chrome, right? So even if you're not a Gemini first person, I think now you have some agentic and AI features baked into the browser, right? I think so much of getting the most out of generative AI for the average business leader. It's about taking advantage of the context that you currently have, right? And still, I think until we kind of unlearn a little more,

So much of that is still just what websites you're on. So yeah, let me know if you want a deeper dive. Maybe we can do the AI working Wednesday this week on Gemini in Chrome because now that this is being rolled out to tens of millions of people in the U.S., it might be time that we take a look under the hood. Yeah, I've been using this since May. I have really enjoyed it, but now it's a broader rollout to all free users in the U.S. All right.

Let's keep it going. some new studies this week that were really interesting, even if you're not a dork like me and read AI studies. So according to new usage studies from Open AI and Anthropic just released, kind of showing how users use their models. So it says that people are using chat TPT mainly to assist in their work, while Anthropics Claude is increasingly being used to hand off entire tasks, marking a pretty clear split in how two of the,

leading chatbots fit into work and school. So if we look at OpenAI's study, so it reports that asking conversations where users seek guidance or decisions without direct output have recently outpaced more doing and expressing kind of queries indicating many rely on ChadGPT as a thinking partner versus a task finiture. So yeah, Open AI's kind of study, which we shared about in our newsletter this past week, looked at, you know, kind of categorize all of these queries anonymously, of course, into different categories.

So yeah, it seems like people are looking for more answers in chat Chb-T compared to how they were previously versus just actually finishing tasks. So OpenAI says Chad ChbT's most common uses right now are writing, editing, summarization in brainstorming with strong adoption by students, marketers, knowledge workers, and educators for drafting essays, prepping reports, creating quizzes, tutoring,

and simplifying complex topics. So OpenAI's framework, classified chats, like I said, into those three major bucket, which we're doing, or generating an output for production, asking,

which is decision support without direct output, and expressing, which is kind of everything else. Anthropics study showed some different results. So their, latest study was the third in their annual series, and it found that Claude is used heavily for

coding and math-based tasks with software development and other quantitative work accounting for more than a third of activity. So Anthropic reports that for the first time in its tracking, automation usage exceeded augmentation as companies embed Claude in workflows for document processing, report generation, and mainly bulk coding. So the studies suggest a practical split between the new, uh, between the two AI startup giants with chat GPT currently thriving in consumer productivity, where users iterate alongside it or just augmenting intelligence, while Claude is getting ground in enterprise settings where teams delegate complete tasks for speed and scale, mainly those tasks having to do with software coding and development.

We got some agent news here. And this one is pretty interesting. It might seem small and maybe even a little dorky. But when it comes to AI agents, I think this actually might be one of the bigger pieces of news we've had in years that wasn't tied directly to a new AI agent becoming available. That's because Google and more than 60 partners announced the open source agent's payment protocol or what they're calling AP2 to let agents complete purchases with clear authorization, authenticity, and accountability. So this is a pretty big deal as this is the first common standard that allows any AI agent to safely complete purchases on their own with clear proof of human.

user permission, right? That's the big thing. So yeah, there's different, you know, AI agents that if you want to roll the dice and if you're feeling spicy, you can go put in all your information, your credit card information and, you know, go push them to complete purchases for you. I wouldn't want to do this just yet. But maybe once this new AP2 protocol comes out, that might change it. So right now, it's available to developers now via a public GitHub repo.

But right now, there's no consumer. Facing implementations yet. So, you know, even if you're ready to go give an agent supporting AP to your credit card and see if it can go do some of your purchases for you. You can't do that just yet. But with the backers supporting this, I'm guessing it's going to be rolling out soon in production to everyday consumers. That's because right now backers include some of the biggest players, including American Express, Coinbase, Etsy, Intuit, Mastercard, PayPal, Salesforce, service now, union pay, world pay, and more indicating broad industry buy-in across payments, commerce, and enterprise software.

So the core mechanism here and what makes AP2 pretty unique is the mandates, which is a tamper-proof cryptographically signed digital contract using verifiable credentials that prove a user's intent and create an auditable chain from instruction, from human instruction to agent check. And then it allows that information to also be shared across agent platforms. So this is kind of, you know, kind of how we've talked about Google's A2A protocol, Anthropics MCP that allows kind of AI agents in large language models to connect to each other and the open web. This is very similar. It's except it is specifically for payments.

And it does right now support multiple payment types, including credit and debit cards, real time. bank transfers in stable coins. All right. Let's take a quick little pause for a word from our sponsors. This podcast is supported by Google. Hey folks, Stephen Johnson here, co-founder of Notebook LM. As an author, I've always been obsessed with how software could help organize ideas and make connections. So we built Notebook LM as an AI-first tool for anyone trying to make sense of complex information. Upload your documents and notebook L.M instantly becomes your personal expert, uncovering insights and helping you brainstorm.

Try it at notebooklm.gooklm.com. All right. Let's keep the AI news train rolling to chew. So this one pretty big for the UK as according to reports, many U.S. tech firms have recently announced a big push for more AI in the UK, specifically as it comes to data centers. So, according to reports, more than $31 billion has been recently invested into the UK by major U.S. tech firms tied to data centers, AI infrastructure, and also quantum computing. So the UK government will speed up data center approvals and ease grid connections.

aiming to cut costs and timelines for power-hungry AI projects. So some of the bigger players, OpenAI, NVIDIA, and UK operator N-scale, we'll build the Stargate UK. All right. So this is the UK version of the large Stargate project here in the U.S., which will start up with 8,000 Nvidia Blackwell GPUs and eventually scaling up to 31,000, with sites including the new AI growth zone at Cobalt Park. So Open AI will also launch the Open AI Academy to provide AI training and reskilling for up to 7.5 million UK workers by 2030, positioning AI literacy as a nationwide workforce priority.

Google, Microsoft and others are getting in on the action with their own investments, but one of the bigger ones is Microsoft, apparently planning to invest more than $30 billion over four years in the UK, including a supercomputer, near London with more than 23,000 GPUs in partnership with N-scale. So pretty interesting here. You know, U.S. President Trump has been easing restrictions to allow more data centers to get up and running domestically here in the U.S. So it's pretty telling that also the big tech companies here in the U.S. are also investing tens of millions of dollars in the UK. So people that want to sit back and say, oh, you know, AI scaling has hit a wall or, you know,

AI is going to go away or the AI bubble is going to burst. Well, the biggest tech companies who are creating the technology are continuing to invest tens of billions of dollars worldwide on future projects. So I don't know if the people making the technology had any hesitation or indication that this technology was slowing down, they probably wouldn't be investing all their pennies in building it out bigger. I don't know.

That's just me. Here's an interesting AI news story that I had to read like five times to make sure that this was true, but apparently it is. So, according to Business Insider, XAI, owned by Elon Musk, put a recent high school graduate

in charge of its grids. rock data annotation team amid sweeping layoffs in leadership turmoil, raising questions about stability and experience at a core function training the company's AI models. So this is according to a business insider report. And it said that more than 600 people were cut from XAI's data annotation unit recently, like as of last week, with headcount dropping from a. dropping from about 1,500 in that division to about 900 after multiple rounds of counts. So again, this is all according to a business insider report, but it said that Diego Puccini,

a 2023 high school graduate. And yeah, 23, I said high school graduate. And university, a Pennsylvania student who is currently on leave, reportedly, began leading XAIs team. in early September following deactivations of at least nine senior staff slack accounts. So workers told the business insider that Puccini said on September 15th that there were no further layoffs plan, but more than a hundred additional staff were cut shortly after that September 15th announcement that no one else was getting fired, signaling ongoing churn and uncertainty in the data annotation division.

So this team is the one that handles labeling, categorizing, and contextualizing raw data used to train GROC, XAI's chatbot, meaning those staffing shifts touch a mission critical pipeline for model quality and safely for millions of people using it. So employees reportedly said that Pissini directed rapid one-on-one reviews, in Slack-based tests to determine people's future roles, indicating a fast metrics-driven reset of the team. So Pissini joined X-A-I in January after winning an XAI hackathon, and his network and interests include robotics in computer visions, according to Business Insider. So two employees who questioned Puccini's appointment in Slack had their accounts deactivated within hours,

according to screenshots viewed by business insiders. Oh, I don't know what to say here, y'all. People sometimes, you know, they accuse me of being, you know, harsh on Apple or harsh on GROC, but I don't know. If I'm an enterprise company, well, first of all, I don't think any large, you know, Fortune 100 companies are seriously considering using XAI's grok in production. But if any of them were,

which I haven't talked to any serious enterprise company here in the U.S., and I talked to a lot of them that are actually considering this, but I don't know. How would you feel if a, again, I'm sure he's brilliant, right? He has to be, but I would not have a ton of confidence if someone that graduated high school in 2023 was the one leading data labeling,

specifically when you have to look at, well, what's one of the main sources of GROC's data? It's Twitter or X. So obviously, data labeling and data annotating is wildly important for any AI company. But when you look at who the other big players have in similar positions, it's usually people with multiple PhDs and decades of experience in machine learning. So I don't have a ton of confidence that this is going to work out very well. Yeah, Dennis here saying this headline feels like clickbait, but then again, who knows any more given Elon's management style?

Yeah, I literally couldn't believe this. But this is, I mean, Business Insider is a reputable publication. Yes, sometimes the media gets things wrong, but this has been reported elsewhere. So yeah, it was one of those things I couldn't actually believe, but I don't know. What's even believable anymore, right? In the world of AI. Well, here's another one that I had to do a double take in. Our next piece of AI news, Nvidia has bought a $5 billion stake in its, well, rival Intel.

It is launching a joint effort to develop chips for PCs in data centers, a rare collab that, that could reshape the competitive landscape in AI hardware. So yeah, Intel had been struggling and one of its biggest competitors, Invidia said, yeah, we're going to go ahead and invest $5 billion and work together. So on that news, Intel shares jumped 23%, massive jumps for Intel lifting its market value by about $143 billion. And Nvidia's shares also rose 3.5% on the news. So this new partnership aims to co-develop chips targeting both consumer PCs and cloud data centers, signaling an effort to speed AI computing performance and diversify supply

for major customers. So this move follows fresh outside backing for Intel as the US government, yeah, we talked about that a couple of months ago, recently took a 10% stake in Intel worth about $9 billion and also Japan's soft bank invested $2 billion. So Intel has struggled in recent years amid leadership turnover, technical missteps, and tough competition from Nvidia and AMD and Broadcom, but in which we've talked about recently that Open AI is apparently developing chips with Broadcom. And Intel's margins and sales have suffered as it has lost dominance in key markets. This could be one of those pieces that turns it around.

All right. And our last big piece. of AI news for the week. Apparently, Microsoft CEO warned that AI could make them irrelevant. Yeah, one of the biggest providers of AI in the world. So, according to reports, Microsoft CEO, Sadia Nadella, bluntly told employees at a recent internal town hall that AI could make Windows office and even Azure less relevant

and that Microsoft itself could become irrelevant if it clings to pass success. So he said that he is haunted by the former cautionary tale of once giant company, digital equipment corporations, collapse and use DEC as a warning that great companies can all but disappear if they miss a platform shift. So Mandela said AI already writes content for Excel and PowerPoint and could undercut the value of Microsoft's most profitable products if the company does not move faster. So he told employees to stop protecting legacy franchises and instead build for what the market

will demand next, calling for a true growth mindset, not nostalgia. So Nadella also reportedly acknowledged internal cultural strain, including complaints of reduced empathy and increasing rigidious and warned that without trust, the organization will fail under the AI era pressure. The message to teams was clear, ship AI native products, and be willing to disrupt Microsoft's own businesses before competitors do it. This is one that also kind of shot me. There was a lot of AI news this week that I was like, wow.

But I think that there's an important message here from Microsoft's CEO. So again, this is coming from. one of the largest companies in the world, right? Technically the second largest company, Microsoft's the second largest company right behind Nvidia. If the CEO of the second largest company in the world is saying AI could make us irrelevant,

I mean, it's a sobering message. But I think we should also be thinking about this as business leaders. You have to take AI really seriously. And I like what Satya Nadella said about not hanging on to legacy, right? And I think this is important when I tell people, you need to stop, like, throw out this concept of upskilling and reskilling. That's building off of your legacy or that's building off of, you know, your company's past.

You really have to unlearn and start from scratch. You know, if the CEO of Microsoft is feeling, this way about AI. Like, hey, we've got to be faster. We have to be more innovative. Then I think every one of us probably should take a lesson from him and start thinking about that as well. All right. There's a lot that didn't make the roundup this week. So I'm going to go through this fast bullet point style of what's new and what's next because there's a ton. And a lot of this is important. Maybe all of this doesn't hit you, but a lot of it will. So let me sit on my coffee here as we get ready to run down the bullet points of what's new and what's the next.

All right. Open AI release a new model. GPT5 codex specifically for its codex platform and people are loving it. Notion release agents with their notion 3.0 agent that has deep research connectors and memory. ChatGPT rolled out a toggle to choose custom thinking efforts in GPT5 thinking. So yeah, confusing. Now there's like 18 different models. if you count all of those.

Sam Oldman said that OpenAI is releasing some compute-intensive offers over the next few weeks. So a lot of people are thinking that might be a new version of SORA, a new version of their GPT image gen. It could be their new agentic browser. We'll see. But he did say that it might only be rolling out to pro users at first. They're users on a higher plan. Winsurf released a new model co-named Supernova, which a lot of people, are saying is an updated Claude model.

Google released a Gemini desktop app for Windows. It looks really cool. I kind of wish they would release that for Mac, but we do have the new, you know, Gemini in the browser there. So Luma AI released Ray 3, the first AI video model that reasons. Microsoft is releasing a career coach inside of co-pilot. Google made Google gems shareable, small thing,

but I'm looking forward to that one too. XAI released Grock for FAST, which has received some impressive benchmarks so far. Open AI may be making projects shareable soon as screenshots of that feature have leaked. So that'll be, I think, helpful for teams, not only being able to share GPTs, but also projects. Oxford became the first UK university to offer free ChadGPT EDU access to all students. Perplexity announced they're working on a browser or sorry, on a server, face AI agent, not just in the browser. Suno teased V5 of their popular AI music platform.

At the International Collegiate Programming Contest or the ICPC, Google Gemini 2.5, one gold with 10 out of 12 tests passed in OpenAI, said it won gold with 12 out of 12, including using a model for the 12th question that no one else got right. That is not publicly available yet. Meta rolled out. their AI powered ray band display glasses. A little shaky demo, but the tech looks good. Chad GPT rolled out an updated personalization in memory page with new settings.

Microsoft released new co-pilot agents across teams and SharePoint. 11 labs launched Studio 3.0 that blends voiceovers, music, sound effects, and more. And Microsoft is reportedly testing its portrait features in co-pilot labs, powered by the VASA 1 model. I talked to you guys about VASA 1 18 months ago. I'm like, this is so good. It's dangerous. Even 18 months ago, it still holds up to today's technology.

So I will keep an eye on that in the coming weeks. So that was a ton of AI news. Let's quickly go over a recap. And hey, let me know what should we do tomorrow for Hot Take Tuesday? We're going to do a poll in the newsletter probably. So let me know what you want to see. here. All right, quick recap, the biggest AI news stories of the week. Number one, open AI rating Apple talents to apparently build multiple AI hardware devices. The U.S. government has given the green light to Mata's Lama models for federal use. Google has rolled out Gemini in Chrome to millions of U.S. users. Chad GPT and Claude released new studies showing how millions of users are actually using their platforms. Google and Part. partners unveiled AP2 for AI-driven payments.

I think it's actually going to be pretty big. US tech giants have pledged more than $30 billion for AI expansion in the UK. A college student is now reportedly leading the XAI data annotation team after 600 plus layoffs. Invidia invested $5 billion into one of its biggest rivals in Intel and are working on chips together. and Microsoft CEO, Sadie Edna Della, said that AI could make Microsoft irrelevant if the company doesn't turn it around. All right, thank you for tuning in.

If you haven't already, please go to your everyday AI.com. Sign up for the free daily newsletter. We'll see you back tomorrow and every day. For more Everyday AI. Thanks y'all. Meet Firefly AI Assistant. Now live in Adobe Firefly, the Allman One Creative AI Studio. Just describe what you want to create in your own words and the assistant handles the rest,

orchestrating multi-step workflows across Adobe Creative Cloud apps, including Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome while the assistant accelerates execution. Stand control with the ability to step in and refine at any time. See it today at firefly.adobie.com. And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating.

It helps keep us going. For a little more AI magic, visit Your EverydayAI.com and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.