Ep 759: OpenAI previews 'Superapp' with Codex update, Anthropic's big Claude updates, Meta created an AI Zuck 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.

Anthropic again shipped like 7 weeks worth of updates over the past 7 days, but all it took was one big release from OpenAI to stifle all of Anthropic's momentum. And while the competition between between Anthropic and OpenAI continues to heat up. OpenAI shareholders are reportedly pushing for leadership changes as they're casting doubt on Sam Altman's ability to lead the company to an IPO. Speaking of AI CEOs and public companies, Mark Zuckerberg is literally turning himself into an AI CEO, and another big CEO is betting in the future that humans won't even be the main consumer of his product. Yeah, there's more happenings in AI this week than there are samples at Costco on a Saturday. So, you know, that means a lot. And if you couldn't keep up with them all, well, you're in luck. I do that for you. Welcome to Everyday AI. If you're new here, my name is Jordan Wilson. This thing's for you. It's a daily livestream podcast and free daily newsletter helping everyday business leaders like you and me keep up with the latest and greatest in AI news. I tell you what matters, what doesn't, and show you how to grow your company and career. So starts here with the unedited, unscripted livestream podcast. But to be the smartest person in AI at your company, our website is your cheat code. That's youreverydayai.com. You can go sign up for the free daily newsletter. We're gonna be recapping not just today's show, but everything else that's happening in the world of AI as we do every single day. But it's also like a free YouTube, but with good stuff on there, right? Because it's, go watch and listen to 750+ episodes of this very show. Interviewed hundreds of the smartest people in AI, so go check that out. All right, also FYI, things to check out. Well, me, if you're in the San Fran area, I will be in San Francisco next week at Sage Future. So if you're around April 20th, is that right? Yeah, April. No, I'm gonna be there like April 26th to 29th. So yeah, if you're in the area, I do always put my LinkedIn information in the show notes, so make sure to reach out to me. All right, let's start with Anthropic because as has been the case of basically the last like 6 weeks, they've had a straight up silly week. So here's what Anthropic released, and yes, this is in one week. All right, uh, so we got updates to their desktop, uh, program. We got updates to, uh, Claude Design, which is a brand new program. We got a brand new model, and those three things might not even be the thing that I'm going to be using the most. All right, so here's what's new and why, why it matters. So, uh, first, uh, Claude Code Desktop updates. So kind of going chronologically here, uh, so Anthropic updated the Claude Code Desktop app and it now supports parallel async jobs with a drag and drop interface for the integrated terminal, file editor, and diff viewer. So completely rebuilt desktop app. It looks different. Some people like it. I personally am not the biggest fan of switching between, uh, Chat, Cowork, and Claude Code. Uh, some people like it. For me, not so much. There's also a new side chat that enables quick questions about code without interrupting your main workflow. But the thing that I'm probably most excited about of all the Claude updates that came this week is the new routines. Uh, so this is within Claude Code. So even if you're not a technical person, go use it. Extremely simple, but routines are automated repeatable tasks in Cloud Desktop that you set up once and then they can trigger on a schedule or even with an API webhook or by an event. So they come in two types. The local version runs on your computer with full access and control, and then the remote version runs on Anthropic's servers and connects to your online apps that you've connected, like Gmail, Slack, Google Drive, and more. Then we got the Claude Opus 4.7 update. So yeah, at least for, I don't know, the next couple of days or couple of weeks, Anthropic has arguably the best model in the world again. So Opus 4.7 is Anthropic's new flagship model, replacing Opus 4.6 and achieving a very impressive 87.6 on SWE-Bench, right? So as we talk about all things, well, agentic, a lot of times it's whatever model is the best at coding, because then that company will use that model to make better versions of all of their models. So even if you're not a developer, the SWE-Bench or the software engineering benchmark is an important one to take a look at. So pretty big leaps there. But probably the things that are most important are the new vision capabilities. So Anthropic says Opus 4.7's vision capabilities have tripled. All right. And the reason why that's important, well, they say it's now better to not just understand images, right? So if you're uploading a PDF with a bunch of charts in it, things like that, but also its ability to output higher quality things as well. So if you're using, you know, Claude to create spreadsheets, presentations, websites, etc., some better capabilities there. Also, the extended thinking mode is now out, and instead it is adaptive thinking, meaning that Opus 4.7 self-regulates its own reasoning depth. So we'll see when this benchmarks and the real reviews come in on that one. For me personally, I'm still using Opus 4.6 over Opus 4.7. I think you can still technically get more, the ceiling is higher with 4.7 obviously, but it takes a lot to get it there, right? I think one thing I do like about Opus 4.6 is, well, it's very fast and very capable. I'm never going to give it my, my most important tasks that require a lot of thinking, that always goes to GPT-5.4 Pro. But for the most part, I loved Opus 4.6. The new 4.7 with adaptive thinking, not my favorite because sometimes it decides it doesn't need to think a ton when I'm like, yo, actually you should think a ton and it doesn't. All right, but don't worry. There was actually maybe the one that got the most pub or the most eyeballs this week. Yeah, another big release. They weren't done with just the brand new desktop routines in Opus 4.7. That's because Anthropic also released Claude Design. So right now this is in research preview for paid subscribers, but it is— well, it looks bonkers. I did play with it a little bit on Friday and over the weekend, but it generates editable visual assets from text prompts so it can create UIs, prototypes, slide decks, one-pagers, and it even creates, yeah, videos. All right, they're not fantastic, but if you're looking for, uh, something simple animation and you go through and you give it a couple iterations, it's actually pretty decent, right? So it's not going to create, uh, you know, your next, you know, service launch or app launch in one shot, but you can actually get some good, yes, actual videos out of this new Claude Design. Uh, so Users can also export assets directly to HTML, CSS, React components, PDFs, PowerPoints, or even Canva. So the tool can ingest your actual codebase and design tokens like your Figma library. Uh, so all the generated designs adhere to your existing brand and component standards. Then the editing is interactive. There's sliders for elements like padding or corner radius, uh, appear next to previews for quick adjustments. So yeah, you do kind of get a mixture of both, right? Kind of get this new vibe design. But then you also get, if you are more familiar with working like a tool like Figma or some different web-based tools, right, where you have more granular control, you can get that as well. So, you know, livestream audience, podcast audience, if I don't know if people want this, right? So if you do, just, I don't know, leave the word design. Again, this will be my, my, my thing for the week. I have my random number. If we get that many people commenting design. Maybe we'll do something on Wednesday on this one. All right, let's go to our next big news story. So this one technically came out like an hour after we did our last Monday AI News That Matters, and I was thinking about not including it because I'm like, this is like 7 days old by now, uh, because we just missed the cutoff last week. But I think it's important to talk about, uh, that's because there was a, uh, leaked internal memo from OpenAI's Chief Revenue Officer Denise that kind of sent some shockwaves through the AI industry. So it wasn't just about the competition between OpenAI and Anthropic, but it was actually a shot at Anthropic. But calling a spade a spade here, because we've seen reports that Anthropic is essentially reporting their revenue in a different way than OpenAI is. And so Anthropic and others are saying OpenAI is in the lead, and OpenAI in this memo is like, no, it's— you're doing your math wrong. So here's what happened. Uh, OpenAI OpenAI's Denise Dresser accused Anthropic of inflating its annual revenue run rate by $8 billion, arguing that Anthropic counts gross revenue from cloud partners like AWS and Google Cloud, while OpenAI reports net of Microsoft's share. Pretty big deal there, right? Basic accounting. We've seen this. I've talked about this one or two times on the show, but essentially now this is the message circulating within OpenAI. The memo also claims claims that Anthropic's annual run rate is closer to $22 billion than not the $30 billion that's been widely reported, potentially putting OpenAI back in the lead with its $24 billion figure. So the dispute is critical as both companies are preparing for IPOs in 2026. Dresser's memo also criticized Anthropic's safety-focused approach, categorizing it as a fear restriction and the idea that a small group of elites should control AI. That's end quote there, signaling a new and more aggressive public stance from OpenAI. The memo also revealed that OpenAI's partnership with Microsoft, while foundational, has limited its ability to reach enterprise customers, especially as demand shifts toward Amazon Web Services' Bedrock platform. So, yeah, interesting. They're not just, you know, shots at Anthropic and their math. Which isn't mathing to most people that are looking at the revenue rates between the two companies, but then also say that Microsoft was kind of limiting, right? Uh, that was probably the first headline that made waves after the, uh, the leaked memo. But then the second one was the $8 billion, uh, kind of boosting of the revenue. All right, well, speaking of this race, well, that's what we're going to be talking about next. So, uh, sorry, a lot over the next couple of months because, well, it's not only going to be dominating the AI headlines and technically I think all of the releases that get pushed, uh, but I think it's also going to dictate the larger tech scene, just the enterprise environment and well, maybe even your retirement, your portfolio. And I think this next, uh, piece is an important one to talk about. Not as much news, uh, as more of an exposé, I guess, um, coming out of the Wall Street Journal, but definitely some news in here. So, uh, according to Wall Street Journal, some OpenAI shareholders are privately questioning whether CEO Sam Altman should lead the company through its upcoming IPO, which could value the company at around $850 billion, or, well, I would say it's probably going to be north of a trillion dollars by the time they IPO, maybe even more than that. Anyways, according to the Wall Street Journal, some of the main alternative, or the main alternative being discussed, is board chair Bret Taylor, known for co-creating Google Maps. Previously served as the co-CEO of Salesforce, the CTO of Facebook, and also chairing Twitter before it was sold to Elon Musk. So yeah, Bret Taylor has essentially led all these big tech companies. So, uh, that's the report that we're hearing. And these discussions come as Altman has reportedly pushed OpenAI toward large investments in outside ventures where he holds personal stakes, including Helion Energy, a nuclear fusion startup, and Stoke Space, a rocket company. Altman proposed that OpenAI invest up to $500 million in Helion, a move that would have boosted the value reportedly of his own significant stake in the company. Ultimately, OpenAI declined the investment but instead signed a long-term purchase agreement instead. So some board members and investors are reportedly concerned that Altman's wide range of outside investments could create conflicts of interest, especially as OpenAI prepares for the increased scrutiny and regulation with going public.

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.

The company has reportedly told employees to focus on core products and cut back on side projects as OpenAI's lead in AI race in the AI race faces growing competition. I don't have a lot of inside knowledge on this one. So as a very, third-party observer, I'd say, I don't know, right? Seemingly, you know, Sam Altman seems to be a polarizing figure, right? People are either, you know, very much in awe of what he's been able to do with OpenAI, or they say he's not the right person and, you know, has all these other things going on. I don't know. To me, he's leading the world's most successful AI company, right? Depends on the metric, right? But when you look at weekly active users, which is probably the most important and the most telling, right? Those people that keep coming back and are actively using the platform, no one's come close to OpenAI. So yes, there's been ups and downs, you know, that's easy for the rest of us to watch. And, you know, all of the drama we saw back when, he got fired a couple of Novembers ago and then rehired a couple of days later. Right. So it makes for, you know, fun, I guess, watching and listening for those of us on the outside, not affected. Um, but I don't know, to me, uh, it seems like Sam Altman's doing a pretty good job, right? No one, including Google, has been able to catch OpenAI when it comes to weekly active users. Google's gaining ground. Uh, I personally am not buying into, uh, Anthropic Anthropic being able to catch OpenAI in the long run. I, I think once OpenAI gets their ads tuned in, they're gonna be blowing past Anthropic. So I've, I've said this all along. I said Anthropic has the better short and medium-term plan when it comes to monetizing, but I think OpenAI is probably smarter playing the long-term user plan because if you can get people using your platform for longer and if you make it stickier and if you make it available, that's the thing, right? I think I saw a meme, you know, somewhere that said Anthropic's uptime graph looks like, you know, a bag of mixed peppers because it's just red, orange, red, orange, and then some green splashed in, right? If you want to use Anthropic, it is down around the clock, right? It's bad. OpenAI is up all the time. So I don't know, say what you want about Sam Altman, but yeah, at least according to the Wall Street Journal, some members, uh, of the board there are looking elsewhere. All right. Uh, speaking of big boards, big AI companies, yeah, Salesforce apparently looking at a different direction that's maybe creating a product that's more for agents than for humans. So Salesforce has announced Headless 360, a major initiative to integrate agentic AI throughout its entire product suite. So the company is releasing over 100 100 new AI-powered tools, uh, and skills for developers aiming to let AI agents handle the reasoning, planning, and execution within Salesforce. So this one's pretty important because you might be like, hey, I follow AI. I didn't see anything about Headless 360. Why is it a big deal? I think it's actually pretty a huge deal, uh, because this move kind of raises a crucial question for the future of business software. Will companies need traditional software, or will the interfaces just be for agents? And it's pretty telling here that Salesforce is coming out with a product and they're saying, well, yeah, AI— or sorry, uh, you know, software is going to be more used by agents versus humans. So, uh, that's kind of at least where they're positioning this with the Headless 360 as a way for users to just kind of automate those customer interactions through agentic, uh, means versus having to go in and log in and do all these things themselves. So Salesforce CEO Marc Benioff previously stated that 30 to 50% of Salesforce tasks were already handled by AI and encouraged workers to focus on higher-value activities such as automation, automating other processes. So the company is responding to recent market pressures, including a dip in value during what everyone is calling the SaaSpocalypse. Earlier this year, right, which was essentially when Anthropic was coming out with all these new plugins and skills. And, you know, everyone's like, well, isn't this now allow you to create new software or to just bypass a lot of software? And, you know, the market was kind of like, yeah, maybe. So now a pretty big move here from Salesforce responding to this. So the headless 360 approach could fundamentally change how businesses use not just CRM tools, but it could just maybe be a preview of what enterprise software looks like in the future. Right. Because maybe these companies are going to be focusing more of their time, more of their resources on creating things not for humans.

Right.

I've kind of opined on that, you know, for the last year or so, like when I'm talking about what the future of the internet looks like. And right now the internet stinks for humans to use. It's an absolute wasteland, right? I spend much more time directing agents to use my programs and to browse the web for me than I do myself personally. And I'm glad I can do that because like I said, the internet is, it's absolutely terrible. It's a distraction. The content is poor, right? So I'd much rather just have the agents have to go deal with that mess. But, uh, regardless, we'll still be following, uh, what happens with Headless 360. Talk about, I don't know, kind of an apocalyptic, uh, name for a, an extension of the product, right? Headless 360. But headline grabbing nonetheless, right? All right, our next piece of AI, uh, AI news. Apparently Anthropic is still in play, uh, at the Department of Defense even though it's not supposed to be. So according to Axios, the NSA is actively deploying Anthropic's Mythos preview model, which is considered the company's most powerful AI yet. Uh, so the, uh, Department of Defense, which oversees the NSA, had previously moved to blacklist Anthropic and ordered its vendors to stop using the company's tool, citing national security concerns. Yeah, well, here we go. The National Security Agency, uh, reportedly using that same company's model. So despite the Pentagon's efforts, the NSA and other parts of the department are broadening their use of Anthropic's technology, according to the Axios reports. So the Mythos model is mainly being used by organizations to scan for exploitable security vulnerabilities within their own environments. Anthropic restricted access to Mythos to about 40 organizations, citing its offensive cyber capabilities as too dangerous for wider release. And the NSA is reportedly one of the unnamed agencies with access. So yeah, the, uh, the drama continues, right? Uh, essentially there was a ruling, uh, just over a week ago that said that, uh, the Department of Defense could keep, uh, kind of their declaration, uh, against Anthropic. So essentially there's been a bunch of legal limbo back and forth. The DOD, uh, initially labeled Anthropic a supply chain risk, the first time that has ever been given to a US company. There's been more rulings back and forth in the courts, but to resolve this, the DC Circuit Court here in the US has fast-tracked the case with final oral arguments scheduled for May of this year. All right, next, Mark Zuckerberg is going full AI. So yes, the company has invested, well, reportedly more than $15 billion on getting their latest AI model off the ground. They ditched the whole Llama, uh, open source, at least for now. So now they have a closed source yet free Muse Spark. But yeah, Zuckerberg's going even harder in the paint when it comes to AI, making himself and actual AI that his employees can use. So Meta, the parent company of Facebook and Instagram, is reportedly developing an AI-powered digital clone of CEO Mark Zuckerberg himself to answer employee questions and boost internal engagement. So the AI version of Zuckerberg is being trained on his voice, mannerisms, public statements, and company strategy with the goal of making employees feel more connected to their CEO. Meta employs nearly 79,000 people, and this initiative aims to give staff direct access to leadership insights, especially as the company flattens its organizational structure, as most companies are because of AI. Zuckerberg has positioned himself as being actively involved in the training process, providing images and voice samples to build the AI characters. Character. So the company has shifted focus from its original metaverse vision to developing AI-generated 3D characters capable of everyday conversations with humans. The experiment could pave the way for other influencers or executives to use AI avatars for communication, potentially transforming how leaders interact with large teams. Meta is also developing a CEO agent, a personalized AI system that helps Zuckerberg prepare for staff meetings and quickly access internal information, as reported by The Wall Street Journal. So this is what companies should be doing, right? That might sound weird, but larger enterprise companies should have some version of their CEO, right? Nothing that's an end-all be-all, right? Like, hey, AI Zuck told me I could get this, you know, this $500 $100 dinner, you know, comped or whatever, right? No, but I think it's, it's, it's something that most companies should be doing is just collecting the insights from their top leadership and allowing more people to access those. Again, not to give people the, the green stamp or the red X on an idea or a project, but at least to help guide their decisions on where they're going. I'd say this is probably be one of the most underutilized, lowest hanging fruit, right? To get your executive suite in a room for literally an entire day, record everything, everyone's viewpoints, right? Assign that out, label it, you know, put it as a little, you know, vector database or, you know, a GPT that people can chat with. It's technically so easy. So to to me, this is kind of like a duh moment, right? And it's actually crazy that it's the year 2026 and this is the first, you know, major story that we're hearing that a CEO is going full AI clone because I think in the future this is going to be very commonplace. All right. Our last big piece of AI news. Yeah, we talked about Anthropic shipped like a month and a half worth of updates in 7 days. OpenAI with one swooping update said, hold my Codex, uh, because it was big. So OpenAI has rolled out a major update to its Codex app in a new release that they said was a preview of how their future super app would operate. Yeah, so, uh, they're being sneaky out in the open and that their future super app that we've been hearing about, well, it's kind of going to be coming from Codex. So we don't know if it'll be Codex, right? Or if Codex might eventually get a different name, but they're essentially trying out all the features. They're bringing together all the pieces in Codex, right? So let me first tell you what's new, and then I'm going to go on my Codex rant. Yes, again. All right. So OpenAI's Codex can now control your Mac. Mac desktop operating apps, clicking and typing with its own independent cursor, allowing users to work alongside the AI without interruption. This marks a big shift from previous AI computer-using agents, such as that from Anthropic, because this version from Codex literally gives Codex its own mouse or multiple pull mouses, mice, right? Multiple cursors. All right, so the new Codex computer-using agent, well, it will work alongside with you. It works in the background, which is really cool to see. So Codex's new computer-use capabilities function at the operating system level, not just the user interface, which is very different from the other most popular one, which is that from Anthropic's Claude. So this enables multiple Codex agents to run in parallel without interfering with the user's own activity. So the update includes a built-in— so yeah, that's just the computer-using agent. There's a lot of other things, right? The other big one is now it has built-in image generation powered by GPT-1.5 Image. It has an integrated web browser. That's really, really cool to see as well. So, you know, the closest comparison, you know, Claude Code, Claude Cowork, it has to use the Chrome extension. The Claude Chrome extension inside Chrome, which works sometimes really great, sometimes it's absolutely terrible, but it's really slow. So this is much faster and much more seamless when it's built inside of the actual Codex app. Also, there's a built-in file viewer as well. And the— yeah, just everything just seamlessly inside of one app. There's also now a new persistent memory feature. So this remembers user preferences, workflows, and text stacks across sessions and can even suggest automations based on previous actions. So users can schedule tasks for Codex to complete in the future, mirroring the proactive automation features seen in Claude Code and others. So who can get this? Well, uh, not all of those features are available for Windows. Many of them are, but the, uh, the updates are all available for Mac users. So you do have to be on a paid plan. but all you need is a paid plan of ChatGPT. So if you've never used Codex before, yeah, bad name. And I've literally been saying this since February. All right, now I'm going to go into my, my little rant here, um, right? So since the Codex app came out in February, I'd been saying how Codex was actually better at coworking than Claude Cowork. Obviously no one believed me. People in the Spotify comments are like, this guy's He clearly doesn't use any of them. And I'm like, yes, I use them all the time. It's better. Trust me, you'll know eventually. And here we are in April with this big app, and I think people are going to know. So go take, right? If you're a heavy Claude Code user, Claude Cowork user, right? I've been saying this whole time that Claude, that Codex is great for non-technical people. It shouldn't be called Codex, right? But go take your your most common, commonly used, you know, prompts or projects that you use inside of Claude Cowork, configure them to run properly in Codex and then compare them side by side. And it's laughable how much better Codex is top to bottom. It's not even close, right? Especially if a lot of your coworking prompts used computer use, right? Which a lot of mine did because it's so much faster and it's so much more consistent. So Yeah, I don't know if, if you guys want to see like a, a side-by-side, maybe just type in, I don't know, showdown or something.

Is that a word?

Showdown? Yeah, it is. All right. And then maybe, maybe I'll do a Codex versus Cowork episode in the future. All right. So that's our main stories, but let's wrap up with some of the smaller updates that didn't make our top AI news because there was a ton this week. So some of these are are things that have already happened, some new releases. Some of these are leaks and rumors, but let's get into it. So first, a trio of leaders left OpenAI, including the former CPO and CTO of B2B apps and the head of Sora. xAI opened up their API to popular open source project OpenClaw. So yeah, if you were— if your OpenClaw was scraping X in any other way, it's probably going to get a lot cheaper and more reliable now. Google released design previews in AI Studio, letting Gemini users select custom app themes. It's actually— this is low-key a banger, banger feature, right? I'm surprised that Google didn't give it a dedicated name and place on their website. It was just like a video on Twitter, and it's really amazing. All right, OpenAI is reportedly testing its Spud model in the arena, which could be called GPT-5.5 and could be released this month. We've also seen users report that maybe it's even being tested live inside of ChatGPT accounts under the Pro model. So we'll see if that pans to be true. But regardless, it looks like we're going to be seeing a new model release from OpenAI maybe as soon as this month. All right, Meta and Broadcom expanded their partnership to co-develop 4 new AI chips. Google, another big piece of news here. Google is reportedly in talks with Marvel to co- develop new custom AI chips, diversifying from just their relationship with Broadcom. Testing Catalog reports that Google is testing out its own version of CoWork, which will just be named Agent. Speaking of Google, they also released their Gemini desktop for, uh, app for the Mac. Uh, US President Donald Trump deleted an AI image, uh, of him, of himself that portrayed himself as Jesus after backlash from evangelicals and Catholic leaders. Uh, Canva released AI 2.0, adding prompt-based tools to create and edit designs across its platform. Uh, that looks pretty cool. I haven't had time to, uh, test it out yet, even though I use Canva all day. Uh, OpenAI rolled out GPT-5.4 Cyber, granting verified defenders advanced tools for cyber defense. So seems to be similar to the Mythos kind of release here, but we didn't get the benchmarks or the hype, right, that we got from Anthropic with Mythos. Here's an interesting one. Allbirds, the shoe company. Yeah, talk about the biggest pivot of all time. Well, now they're setting up to be an AI compute infrastructure company. They raised $50 million. Their stock obviously went through the roof and they renamed the company Newbird AI. AI, right? Have we ever seen a pivot this intense? I don't think so. All right, there's more, uh, reports Meta hired a 5th founding member of Thinking Machines Lab. So yeah, reportedly Thinky, uh, Thinky is losing a lot of their original co-founders. Uh, Anthropic launched Claude for Word beta, uh, enabling tracked AI edits on team and enterprise plans. Google Chrome previewed side-by-side AI mode for search, enabling in-context webpage comparison Khan Academy announced a $10,000 online AI degree in partnership with Microsoft and Google. Mozilla released Thunderbolt, an open-source enterprise AI client focused on self-hosting and data control. That actually looked pretty cool. Wish I had more time in my day to play around with that one. Groq 4.3 launched in beta for X Premium Plus and Super Groq users. The pricing tiers on Grok are just so confusing. X Premium Plus and Super Grok. All right. Perplexity released Perplexity Personal Computer for Max subscribers and those early on the waitlist. We did go over some of these more app-based updates in our Friday Features show. That's episode 758 from— so just go hit the back button. Next, Google released Gemini 3.1 Flash text-to-speech with real-time audio tags for natural voice control. Again, really, really good. Google shipped a ton this week. Snap Inc. cut 16% of its staff, citing AI automation. And look what happened. Their stock rose 6%. Yeah, it's weird. All these things that I said would happen in 2023, they're happening. All right. Adobe and Anthropic announced partnership to integrate Firefly assistant into Cloud. Microsoft released Copilot in Word, including features for legal, finance, and compliance professionals. Google Gemini rolled out personalized image generation using personal intelligence and Google Photos. So yeah, your nano banana creations can get super personal now if you allow them access to do that. And then last but not least, OpenAI launched GPT Rosalind, a life science model for biology in drug discovery research. All right, that is a wrap for the AI news that matters this week. Again, if you're going to be out in San Francisco, hit me up. I will be there. Be there, uh, from Sunday, April 26th through Wednesday the 29th. So, uh, thanks for tuning in. Hope this was helpful, right? Don't spend hours every single day trying to keep up with all the AI news. It's nearly impossible. I try to do it every single day so you don't have to. So, uh, make sure if you haven't already, please take a couple seconds, subscribe to the podcast. I would really appreciate that if you could. Also leave us a rating. That helps more business leaders like you, uh, find our show and hopefully get a no-nonsense, no BS guide to the AI world. It's tough out there. Go help other people out also by sharing this and going to our website at youreverydayai.com. Sign up for the free daily newsletter. Thanks for tuning in. Hope to see you back tomorrow and every day for more Everyday AI. Thanks, y'all.

And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating.

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!