Everyday AI Podcast – An AI and ChatGPT Podcast - EP 503: Google goes AI nuts, OpenAI countersues Elon, Canva drops new AI features and more AI News That Matters

Episode Date: April 14, 2025

Google went AI nuts at Cloud Next, possibly taking the lead. OpenAI is reportedly ready to unveil up to five new models. Canva enters the AI arena with innovative features. This week's AI news is explosive. Catch up now!Newsletter: Sign up for our free daily newsletterMore on this Episode: Episode PageJoin the discussion: Thoughts on the news? Join the convo.Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineupWebsite: YourEverydayAI.comEmail The Show: [e-mail] with Jordan on LinkedInTopics Covered in This Episode:Google Cloud Next AI Announcements RecapOpenAI ChatGPT Memory Feature DetailsOpenAI and Elon Musk Legal BattleCanva AI Suite Expansion OverviewAnthropic Claude Max Subscription PricingMicrosoft's Data Center Strategy PivotShopify's AI Usage Mandatory PolicyUpcoming OpenAI Model Releases OverviewTimestamps:00:00 Technical Glitch with Equipment04:48 Gemini 2.5 Pro Launch07:06 Google's New AI Collaboration Protocols13:28 ChatGPT Enhances Memory for Personalization17:37 AI Surveillance Accusations in U.S. Agencies20:56 Canva Launches AI-Powered Suite23:10 AI Graphic Design Evolution26:02 "Musk's Lawsuit: A Delay Tactic"28:14 Anthropic Launches Claude Max Subscription34:05 Microsoft Pauses Data Center Expansion35:39 Microsoft's $80B AI Investment Strategy42:05 Shopify's Bold AI Integration Shift44:25 Maximize AI Before Hiring48:31 OpenAI Model Tiers Explained51:41 "GPT-5: Concerns on Model Autonomy"54:20 AI Innovations and Controversies OverviewKeywords:Google Cloud Next, AI updates, OpenAI, countersue, Elon Musk, Canva AI features, Gemini 2.5 Pro, Large Language Model, ChatGPT memory feature, AI race, GPT 4.1, Vertex AI, Google AI Studio, agent to agent protocol, Google Workspace, Google Sheets, MCP protocol, Google DeepMind, GDC, Ironwood TPU chip, Lyria AI model, AI video model, AI music generation, Elon Musk's Dodge team, government AI monitoring, Ethical AI use, Canva Sheets, Shopify CEO AI memo, Anthropic, Claude Max, Microsoft Copilot, AI infrastructure, Recall AI feature, data privacy concerns, AI-powered productivity, GPT models, AI operating systems, personalized AI, GMPT 4.5, AI sandbox, Nvidia hardware, test time computing, AI reasoning, ChatGPT preferences.Send Everyday AI and Jordan a text message. (We can't reply back unless you leave contact info) Start Here ▶️Not sure where to start when it comes to AI? Start with our Start Here Series. You can listen to the first drop -- Episode 691 -- or get free access to our Inner Cricle community and all episodes: StartHereSeries.com Also, here's a link to the entire series on a Spotify playlist. 

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 in 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. Google went absolutely crazy with AI updates.

Open AI is countersuing Elon Musk. Canva is trying to be an AI player amongst the big boys. And Open AI could release up to five new AI models this week starting today. Yeah. That's a lot. As always, the world of AI news and developments has been absolutely nonstop. And you could waste hours every single day, try to keep up with the AI news that matters. Or you could just tune in on Mondays and join me. What's going on, y'all?

My name's Jordan Wilson, and I'm the host of Everyday AI. We are your daily live stream podcast and free daily newsletter, helping everyday people like you and me, not just learn about AI. but how we can actually leverage it and use it to grow our companies and our careers. If that sounds like you and what you are doing, you're in the right place. It starts here with our daily podcast and live stream, but it continues on our website. So if you haven't already, please go to your everyday AI.com.

Sign up for the free daily newsletter. So each day, we not only recap the podcast and the live stream with usually some exclusive insights, but we also keep you up to date with everything. else happening in the AI world. But most Mondays, we bring you the AI news that matters. And this week was a doozy and a half, y'all. So like I said, I think that Google may have for the first time definitively jumped into the lead for the AI race.

Yeah, I was at their Google Cloud Next conference in Las Vegas. We're going to be recapping everything. They announced there. And I think from a. strictly model perspective. It is no longer a race. I think Google is ahead, but for how long? Because Open AI is reportedly going to be releasing new models today and multiple of them this week. It's going to be crazy. I'm excited for this one. Let's get into it and go over the AI news that matters for the week of April 14th. And hey, live stream audience, love to see you dropping by.

You know, if it looks or sounds a little bit different, yeah, heading out to the Google Next conference earlier last week, I was on my way out of O'Hare. The TSA line was long. All of a sudden, you know, I'm getting ready to interview someone after I arrived. And I'm like, oh, looks like the TSA held on to some of my equipment. So, yeah, hopefully this won't sound too bad. But yeah, my bike and some of my normal camera. camera equipment gone with the wind gone with the TSA wind so yeah if the live stream at least sounds or looks a little funky today bear with me all right but you came here for the AI news not for the audio visual quality so let's first talk about and hey live stream what's up guys I forgot to say hi to you big bogey face uh Kyle what's going on Ted joining on YouTube as well we have Brian and Dr. Scott Christopher Muhammad Joe Douglas everyone

else. Fred, join it from Chicago. Thank you for tuning in. Let's get into it. So Google went absolutely nuts at their Google Cloud Next conference in Las Vegas. So Google has made some significant announcements at its annual Cloud Next conference focusing on advancements in AI technology and its flagship Gemini models. All right. So I could and I probably will, do just a dedicated episode on everything that Google announced. I think during the course of the multiple keynotes, I wrote down about 27 different updates that I thought would be meaningful

for everyday business leaders, such as yourself. And if I'm being honest, between the newsletter, the podcast, we've covered like five of them. All right, so let's bullet point some of the more important ones. So the big one, potentially Google announced its Gemini 2.5 Pro was available in more places. This is the world's most powerful and highly rated large language model, at least as of this minute.

You know, who knows? Open AI may be announcing something before we even get this live stream out on the podcast. So Gemini 2.5 Pro, which topped the chatbot LM leaderboards, and it's not even close. It's crushing everyone else. It's now available in public preview via the vertex AI. the standalone Gemini app even if you have a free Gemini account you can also use Gemini 2.5 pro inside Google AI studio for free. Additionally, new models from Google, they introduce the smaller,

more lightweight and affordable version of Gemini 2.5 with its Gemini 2.5 Flash, an optimized version of its Gemini 2.5 Pro model. So this new workhorse smaller model is designed to be faster and more efficient for companies building on top of Google's API thanks to its ability to adjust processing power on each task, a technique called test time computes. So yes, Gemini 2.5 is a hybrid model. So it uses kind of reasoning or thinking underneath to plan out its response or plan out its next steps. But it is a hybrid model. So it still has that kind of, you know, old school speed of a transformer. But it has the ability. to just use more compute and think about an issue.

All right. Let's keep going. A lot of new updates. They're not out yet that they were announced. They're rolling out in Alpha to Google Workspace, which I think will be pretty impactful for a lot of our audience. So including audio creation for Google Docs, automated data analysis in Google Sheets,

and Google Workspace Flows. That's one I'm extremely excited about. It's kind of like a slim down. version of Zapier or marketing automation using Google Gemini's AI, but just for all of your Google Gemini workspace products. So yeah, workspace flows. Keep an eye out for that. Also, Google officially announced the adoption of the model context protocol, protocol or MCP. All right. So that's not all. All right. So let me quickly bullet point the rest of these and live stream audience, let me know what you want to hear more of.

Because I think we will do a dedicated Google show because they announced so freaking much. So new updates to their Google agent space platform. They announced a completely new language that called the A2A or the agent to agent protocol, which allows diverse AI agents from different vendors to communicate and collaborate with each other. So yeah, Google really just bringing a genetic AI together both with its adoption of Anthropics MCP protocol, but then also announcing their agent to agent protocol, making it easier for different AI agents to communicate and collaborate with each other. Pretty cool. They had the agent development kit, Gemini on GDC, which I think is

pretty big. That allows enterprise companies to run Google's powerful Gemini model locally on-prem in your data center by using, so this is with by using Google's distributed cloud with Invidia hardware. So that one's pretty big. It should be rolling out in quarter three. So just a couple of months. So yeah, that one, a move we haven't really seen,

you know, from Google so far. These are similar, you know, Microsoft does have similar offerings. So this is pretty big, allowing companies to run Gemini locally in their data centers. Huge. All right. Also, Ironwood, which is Google's new TPU chip, so their newest AI chip that delivers a reported 10x performance boost for efficient AI inference powering their next gen Gemini models.

We briefly talked about the new Help Me Analyze in sheets. The very popular audio. Overviews, right? So if you use Notebook LM, if you listen to this show, you know, I love Notebook LM and you should be using it as well. So the audio overviews feature where you can generate a podcast based on your information or documents, that's rolling out to Google Docs as well. So that should be a pretty big one. They also announced a text to music model, a lot going on in the creative space,

which I was actually pretty impressed with. So right now, it's kind of like a Suno competitor. It's not as good, at least from the demos that we saw. But you can create custom royalty-free 30-second music tracks from just text prompts using Google's Leria AI model. And that's in Vertex AI. And that is available now inside Vertex AI. So obviously inside Vertex, you are paying by usage, right? So you're not paying a subscription fee.

And then another one, V-O-2 is now available for almost everyone. So, yeah, if you have a paid Gemini account, if you log into AI Studio. So yeah, this is where it gets a little confusing, but, you know, I get the kind of separation. So when we talk about Google Gemini, you have your kind of front-end Google chatbot, right? So Gemini.com, then you have Vertex, which is more for enterprise and developments. And then you have Google's AI Studio, which is kind of a sandwich. box, so to speak. So Google AI Studio is free. However, there's no data protection. So anything you put in there, Google does use to train its models. So keep that in mind. But they do have the new V-O-2

AI video model, which is by far the best AI video model in the world. It's available. So you might, if you have a workspace, a paid workspace account, it might not be in there. I logged into my personal Gmail account in AI studio. And I do have access to V-O-2. Some of the, and there's not just the access and availability on VOT, but there's also some new things you can do. You can control, you know, camera angles, some things like that. Really cool. And then not laugh, but not least, Google's remaking the Wizard of Oz using its generative AI tools,

including VO2, and they're debuting it in the sphere in Las Vegas. So I was there for that announcement. pretty impressive, right? I had never been to the sphere before, but, you know, really how Google is using their different AI tools to remake the Wizard of Oz. It looks really good. All right.

Yeah, I told you all. That was a lot from Google. So, yeah, live stream audience, what more out of all those Google announcements do you want to see or hear more of? We'll maybe, you know, make sure we cover that in whatever we do get time to cover all of those Google announcements.

All right. Next. Open AI has expanded chat GPT's memory feature to include all past conversations. So OpenAI has rolled out a significant update to chat GPT's memory feature, allowing it more, allowing it to be more comprehensive and personalized. So this new feature allows the chat bot to reference all previous conversations to deliver tailored responses marking a step toward AI systems that evolve alongside overtime with users. So OpenAI's updated memory feature now enables chat TVT to not only, you know, remember saved preferences, but also insights from all of your past chats aiming to provide a more relevant and

personalized interaction. So according to OpenAI CEO Sam Altman, this development reflects the company's vision of an AI system that grows with users over their lifetime offering increasingly useful and customized experiences. So the new updated memory feature works in two different ways. One way is not that impressive or very useful, in my opinion, which is the saved memories. So that's, uh, you know, that's been out for a very long time. That's just where, uh, you know, chat, UVT just kind of automatically add certain things that you talk about to its memory, essentially like a memory bank and then it recalls on those things later.

But now it just has chat history. So it can remember and gather all information from all of your previous chats. So you can opt in or opt out. So for our live stream audience, I have this on my screen. So if you have a paid chat gvt account and you want to give this a try, you can go into settings, personalization, and then under memory, you can toggle it on or off. It's a new, a new option that says reference chat history. So I think this is going to be one of those divisive features, I think.

I think casual or more casual chat GPT users who really use chat GPT for a specific focus are going to absolutely love this. Myself, probably not. Not going to like this probably at all. So, you know, even though I do have, I don't know, I lost count at least seven, seven different chat GPT accounts, right? So, you know, companies hire us to teach them chat. So, you know, I've access to a couple enterprise accounts.

At everyday AI, we have a team account. I have my own chat chatt chbt plus and chat chpt pro accounts, a couple of free accounts. Because I'm always needing to try things out for you all, right? But even in those individual accounts, I use chat chvety for absolutely everything, multiple lines of businesses, multiple clients, multiple, you know, personal projects, things happening in my personal life as well. So, you know, this ability for the, you know, chat chvety to automatically remember things from past chats and then reflect that in its responses, at least if you use chat chvety for a lot of different reasons, you might not like this, right? As an example, sometimes I want chat chad t to be very short, speaking bullet points, you know, for maybe things about, you know, my own personal health or a vacation I'm wanting to take, right?

And then when I'm researching something for everyday AI or for a client, I don't necessarily want those preferences to be reflected in the response. So, eh, 50-50. We'll see how it goes. Oh, gosh, big bogey face. Thank you. I forgot to mention Google DeepMinds Firebase IDE. Yeah, might have to throw that in.

Our next piece of AI news, this one's not good if it's true. But Elon Musk's Dodge or Doge, I don't know what's called Doge team, has been accused of using AI to monitor federal agencies for anti-Trump sentiment. It's not good if true for, especially for our viewers in the U.S. So Elon Musk's Department of Government Efficiency or Doge is reportedly leveraging artificial intelligence to monitor federal agency communications for perceived hostility toward President Donald Trump raising concerns about transparency, legality and ethics. So this is according to a report from Reuters. So according to this report,

Trump administration officials claim Doge is surveilling communications with at least one federal agency to identify anti-Trump sentiment using AI, marking a controversial use of the technology in government operations. So the Environmental Protection Agency is allegedly one of the agencies targeted with AI monitoring employee communications for perceived disloyalty to Trump or Musk's agenda. That's gross. I don't know. I don't want to go on an accidental hot take Tuesday. This is the AI news that matters.

So we'll try to keep it middle of the road here. But Doge staff are accused of bypassing traditional betting processes and operating in secret. secrecy, including working collaboratively on Google Docs instead of circulating official drafts, according to sources. So the Trump administration argues Doge is exempt from federal freedom of information laws, you know that thing protecting our democracy, prompting legal challenges from watchdog groups demanding transparency. A federal judge recently ordered Doge to hand over records to citizens for responsibilities and ethics. and Washington, but at least so far, the organization has not done that.

So critics, including Democrats and Republican, argue that Doge's actions are part of a broader efforts to purge nonpartisan public servants and replace them with loyalists aligned with Trump and Musk's agenda. So Musk has previously suggested AI could replace government workers entirely, raising concerns about the ethical implications of automating federal operations. Yeah, Joe says this is disgusting. Douglas says, no mellow Monday, bring the heat. Fred is saying kind of big brother. Yeah, this is not good. The fact that this is reportedly happening in the federal government is absolutely bonkers. Put your politics aside, right? You know, throw out, whether you're a Republican,

a Democrat, moderate, doesn't matter. The fact that we have reputable reports that the government is using AI to spy on federal workers and to make sure their communications ally with a certain organization's agenda. Again, throw out, throw out your political bias.

That is absolutely boggers. And if true, an extremely sad day for democracy here in the U.S. So I don't know. I'll just say gross. Hopefully it's not true, but it probably is, which is extremely troubling.

All right. On to news that isn't indicative of democracy dying. Canva has expanded beyond its normal visual design suite and is going all in on everything AI. So Canva has unveiled its new kind of 2.0.6. suite, including a lot of new AI updates like Canva Sheets, a visually dynamic spreadsheet tool that integrates AI powered features like Magic Insights and Magic formulas. The platform right now aims to rival traditional spreadsheet tools like Microsoft Excel by focusing on visual storytelling, offering bright and engaging designs that Excel does it prioritize. So not just Canva Sheet.

There's a new AI-powered charts app that complements Canva sheets as well. Also, there is just this new Canva AI kind of interface. So you can do with a simple text prompt like you could in chat, GPT. There's a couple of different new features in this dedicated Canva AI section. There's a design assistant. It can generate and render code, right? So it's Canvah's trying to get into the IDE game as well. right, like a slim down version of cursor or something like that.

And also, you know, with your normal kind of design. So, you know, creating social media design and then being able to iterate on design based on both reference images and follow-up text prompts in natural language. Also, the Design for Me feature got a big upgrade and allows users to upload images and received AI-generated design suggestions. while the canvas AI powered photo editor can inject objects into AI generated backgrounds with realistic lighting adjustments. Yeah, that one I did watch highlights of the keynote, didn't get to watch Canvas whole keynote because I was at the Google Cloud Next conference. So again, thanks to Google for partnering with us on that one.

But it looks pretty impressive, right? But let me just say, a lot of these features right now aren't out. They're quote unquote coming soon. So who knows when we'll get all of them. When we do, though, this is pretty impressive, right? Who I don't want to be right now, I don't want to be Adobe, right? Canva, some power plays, right? And then, I mean, everything that we talked about last week in our AI News That Matter segment with chat, GVT, their new 4-0 ImageGen, extremely impressive, right?

for graphic design, iterating on photos, editing photos at a very high level with prompts. The same thing inside, you know, Google, you can do the same thing with their new Gemini 2.0 Flash, I believe it is. You can, you know, it's multimodal by default. You can upload your own photos and edit them with text prompts. So between what Google and OpenAI and now Canva have released a lot of pressure on Adobe, not just to see how they respond, but to really see their position going forward.

And oh, yeah, apparently, I put this out in the newsletter, I think on Friday. Apparently, y'all want to see a lot more coverage on Canva. I had no clue. Apparently our audience, you really want to know and see everything that Canva has to offer with its new AI suite, live stream audience. Let me know if that's you too.

Maybe that's just our, you know, newsletter audience. But I'd like to know, Douglas says it's time for a Canva show. All right. So we'll see. All right. More drama. There's always drama in the AI world.

So Open AI has filed a countersuit against Elon Musk, accusing him of attempting to hinder its business and seize control of AI innovation for his personal gain. So Open AI in their countersuit just filed alleges that Elon Musk, obviously a former co-founder of Oversex. open AI going back to the early days. So they've alleged that Elon Musk has engaged in quote unquote bad faith tactics to slow down open AI's progress and gain control of leading AI technologies. The company claims Musk's actions are driven by personal motives rather than a commitment

to advance AI for humanity. So Elon Musk in a series of, I don't know, I've lost track how many different lawsuits, Elon Musk has filed over the last year and a half. against Open AI, but he's previously sued Open AI multiple times in at least multiple updates. He's sued Open AI CEO Sam Altman to block changes to the company's corporate structure. So Musk was an original co-founder in Open AI but left the organization in 2018, citing disagreements over its direction. A federal judge has set a trial date for March 2026 in Musk's lawsuit against Open AI, fast-tracking the legal fight. So in Musk's original lawsuit, he accused opening eye of straying from its original mission,

which is not illegal, by the way. So yeah, I talked about this. You know, let's let's just be honest. You know, it's Musk's original lawsuit is a joke. I don't think anyone believes it has legal ground to stand on. It's more or less a delay tactic, right? You know, and I have to say that because there's actually been organizations that I've talked to over the last few months that have been like, oh, I'm not sure if we're going to go, if we're going to start using chat TVT because, you know, Elon Musk is suing them and it's not a real lawsuit. It is a delay tactic here. Not that I'm trying to choose sides, but, you know, I would say anyone, you know, that can read,

can go in and see this is just a delay tactic. And obviously, you know, Elon of Musk has a competing company in XAI and it's Grock chatbot. So, yeah, and now a position of power in the government. So it could get a little messy. So in short, Musk made a kind of tongue in cheek unsolicited $97 billion bid to acquire the company earlier this year. And in response, Altman posted, Open AI CEO Sam Altman posted a sarcastic offer to buy Twitter instead for $9.74 billion. So Musk's competing AI company, XAI, has struggled to match Open AI's success. So now, you know, Elon must instead has resorted to filing some what a lot of people are calling frivolous lawsuits that are really just delay tactics and are actually just making it harder for Open AI to transition from a nonprofit to a for-profits company.

So in the countersuit, OpenAI insists that Elon's actions are self-serving and not aligned with its mission to benefit humanity. So in a statement, the company said, Elon's quote unquote, Elon's never. been about the mission. He's always been about his own agenda. All right. So we will continue to have more juicy AI developments. It's like a soap opera. Grab your popcorn. All right. Our next piece of AI news, Anthropic, has unveiled a new premium subscription tier called Claude Max. So Anthropics, new subscription plan, Claude Max comes in two tiers, a $100 per month. option, offering 5x higher rate limits than the normal Claude paid plan in a $200 a month plan with 20 times higher limits.

So the higher tiers provide priority access to Anthropics' latest AI models and features, catering to power users who require extensive usage. Adobe just introduced an entirely new way to create, bringing the power and precision of its creative suite into one conversational experience. Meet Firefly AI Assistant, now live in the Adobe Firefly app, the all-in-one creative AI studio. Powered by Adobe's creative agent, Firefly AI Assistant lets you start with your vision, just describe what you want, and shape the outcome as it takes form with the Assistant.

The Assistant orchestrates multi-step workflows, drawing on 60-plus pro-grade tools across Adobe Creative Cloud apps, including Photoshop, Illustrator, Premiere, Lightroom, except. and more to help bring your ideas to life. You can also get started with creative skills, a growing library of pre-built workflows for common creative tasks like batch editing photos, creating mood boards, portrait retouching, and creating social variations. Every step the assistant takes is visible so you can refine, redirect, or take over at any time. You stay in the driver's seat as the creative director.

Adobe Firefly AI assistant now in public beta. See it today at. firefly.adobie.com. So the launch follows open AI success of chat GPT Pro, similarly priced at $200 a month, which reportedly contributed $300 million in annualized revenues to open AI just two months after its release. So Anthropic is reportedly hoping, hoping for similar financial gains. So Anthropics product lead hinted at the possibility of even pricier subscription plans in the future, although Anthropic currently does not offer unlimited usage like Open AI does.

So the company has reportedly seen significant demand for its new Claude 3.7 Sonnet AI model, which uses advanced reasoning capabilities to deliver more reliable answers to complex questions. Live three, audience, what do you think about this? I have thoughts. I have thought. Let me just say this. When Open AI announced their $200 a month pro plan with it came exclusive access at the time to certain features and modes that were not available elsewhere. So as an example, O1 Pro mode, which is still, even though we have Gemini 2.5 Pro as the world's most capable model, I still use the O1 Pro model a ton, especially when dealing with a lot of data.

or working with extremely complex business tasks that require a lot of reasoning. Also, at the time, OpenAI said, hey, you get unlimited, you know, unlimited usage with the $200 a month plan, as well as at the time, SORA access. So if you wanted at the time to access SORA, Open AIs, AI's AI video model, that was the only way. So with Claude here, Anthropics $200 a month subscription, I don't see any of this.

I don't see any of this, right? I had an episode. You know, let me just go ahead and let's see if I can find what episode this was in case you want to go listen to it. It was a while ago. It was, let's see. Where was it? Oh, all right. I'll put it in the newsletter.

But I did an episode a couple of months ago, essentially on the three reasons why that I thought entropic clod was failing, right? It was just losing its footing as a top AI lap. And one of the reasons was the rate limits, right? So on the paid Claude plan, I kid you not, y'all. I'm always testing the boundaries of all the different AI models. I subscribe to almost each and every one if it has a paid plan, especially, you know, a, you know, quote unquote AI chatbots. I have the paid plan. If you looked at Claude the wrong way, it rate limited, right? Where even with, you know, Chad GPT on the $20 a month plan, the limits are pretty generous, right? You could have dozens and dozens of messages in an hour and you're not going to hit your limit.

With Claude, on a paid plan, the $25 a month pay. plan. It's very easy, especially if you're working in a large context window, right, which is one of Anthropics reported benefits, right? I would hit my rate limit in less than 10 minutes consistently. Yes, I'm a power user. Yes, I'm usually having multiple tabs open working in long context, but I don't hit those when using Google Gemini, when using Open AIs chat GPT, when using even the $20 version of Microsoft co-pilot, the web, base version or any other AI model. So Claude has been absolutely terrible from the get-go in terms of letting people actually use their product, right? So, you know, there's a lot of

reasons why I am extremely bearish on Claude, but this, in, Anthropic in general, but this does not give me any confidence in the long-term stability of the company, right? To me, this shows, well, you probably should have had your rate limits fixed to begin with. Yes, there's going to be probably thousands of people, maybe more, that are using this and enjoying it, because those people that are using Claude on the front end and running into rate limits and who would gladly pay more, this is good for them.

I don't see this ultimately being something that is going to help, that is going to help Anthropic in the long run. All right. So Fred said maybe it's episode 301. Goddard love it when our live stream audience. Thanks, Fred. Hold on it down for Chicago as well. Got a love on our live on our live stream audience is better at finding information on everyday AI than I am.

All right. So, I, y'all, if I'm being honest, I don't buy. You know, if I was buying stock at a company right now, I know Anthropic is not a publicly trade company. I would avoid Anthropic. I'm just saying, yes, I use it. Yes, there's still some specific use cases that I think it is. best in. But at least since Gemini 2.5 Pro from Gemini from Google came out, those advantages to Empropic are very, very slim. All right. Our next piece of AI news, Microsoft is slowing down

or pausing some of its ambitious data center construction projects. According to reports, that includes a $1 billion projected facility in Ohio, signaling a recalibration in the tech giants approach to AI infrastructure expansion. So Microsoft confirmed it is halting early stage construction on rural land in Ohio and will reserve two of three sites for farmland. This comes as part of a broader effort to refine its infrastructure scaling strategy. So the company has been executing its largest infrastructure expansion in history to meet surging demand for cloud and AI services, but is now re-evaluating some projects due to evolving customer needs, and that is according to Microsoft's president of cloud computing operations.

So other paused projects include the later phases of a large data center in Wisconsin, and analysts report that Microsoft is scaling back international expansions as well, in canceling U.S. leases for third-party operated data centers. So should be interesting because Microsoft did announce plans to spend over $80 billion globally in AI infrastructure this fiscal year, doubling its data center capacity in the past three years. However, it is strategically pacing its investments to align with some of its bigger business priorities and evolving customer demand. So this is definitely a story to keep an eye on.

And a lot of people are going to wonder, okay, is this because you have companies like Google, at least right now coming in with cheaper prices? Is this because of open source models? Are people using as an example Chinese models like Deepseek and Quinn? So this is something to definitely keep an eye on from Microsoft. So another thing to keep an eye on from Microsoft,

our next piece of AI news, Microsoft has released a handful of new AI updates in a new operating system update, including the controversial Recall AI feature. And that feature has faced both backlash and multiple delays over privacy concerns. So Recall AI allows users to capture encrypted screenshots locally, automatically, to search through past activities, such as apps,

websites, images, or documents. And you can go and find that content by just describing it in natural language. So, yeah, recall AI when it was announced, I'm like, this is crazy. For me personally, I don't care about my data privacy. I don't. Right. I don't care. Right.

It's like, okay, I'm on all these websites. I'm working in all these documents. I'm creating, you know, these presentations. I don't care if my local computer, because a lot of this runs on device, it runs edge, right, on an encrypted server. Yes, take it. I don't care. I know not everyone has the same lax or, you know, lax-ask approach to data, especially if you are an enterprise company. So I get that, right?

I'm a smaller company. Every day, I is a smaller organization. So I'm not as worried. But, I mean, just the technology is extremely. impressive and very useful, right? So imagine you're like, oh, man, what was that document two weeks ago? I was emailing someone about it. You know, it had, you know, that green cover.

And it was something about, you know, marketing our, you know, new product. You can literally just say that. You can go into Microsoft Recall, say, what was that, you know, email and that presentation from a couple of weeks ago. It was green and it was about marketing our new product. And because it takes these. consistent, ongoing, encrypted screenshots, and it can use co-pilot and co-pilots underlying GPT-40 computer vision technology. It can instantly tell you that. So I think it's extremely impressive, but there's been a lot of issues early on because earlier versions of recall

faced criticism for storing sensitive information, including passwords, credit card numbers, and private keys as unsecured text files. So this did lead to widespread backlash from initial users testing it, privacy advocates, and cybersecurity experts. So in response, Microsoft has added new safeguards to the feature, including requiring users to explicitly opt in. So before it was an opt out feature. So now it is an explicit opt in to saving those snapshots and encrypting them on local storage. So access to the snapshots would be protected by Windows Hello authentication to ensure only the user can view them. So right now, this new rollout, the tool will be initially available to Windows insider members and optimized for select languages,

including English, Chinese, French, German, Japanese, and Spanish. And it will reportedly roll out to the European economic area later this year. Yeah, it doesn't just work on any PC. You do need a newer PC, such as a co-pilot plus PC. Minimal requirements are 16 gigabytes of RAM, 256 gigabytes of storage, and on-device edge AI. Microsoft also introduced some additional features in the same build, including ClickToDo, which simplifies actions like text copying, object erasing, and background removal through intuitive shortcuts. Users can activate it with the Windows key or search for it in the task bar.

So other improvements just rolled out in this newest update include enhanced Windows search capabilities, a speech recap feature for live transcription, and quick access to spoken content and updates to productivity tools like paint and photos. All right. So if that's something you're interested, be on the lookout for that. Yeah, Big Bogey Face says it's a security nightmare. I get that. It is, it can be, right? So we'll have to see if this newest release where Microsoft addressed some of these security features and added the new safeguards.

We'll see how the public responds to this as it is finally beginning its rollout to all Windows insider members in those countries. All right. our next piece of AI news. A leaked internal memo from Shopify CEO Tobias Lutki has gone viral. Because essentially he's saying all of his employees need to absolutely use AI always, right? I'm generalizing here. We covered this in our newsletter. But more or less, his memo was like, hey, if you have a problem, you better be using AI immediately to solve it before even going to other humans.

So he outlined in this post, which was originally leaked, but then he just said, hey, it seems like people are talking about this. So here it is. So he did eventually release this full internal employee memo after it was leaked. But he essentially outlined his bold vision for integrating AI into every aspect of Shopify's operations. So the memo signals a major cultural shift in how businesses approach AI adoption. So some of the key takeaways, he said that using AI effectively is no longer optional, regardless of your job role, marking a significant operational and cultural shift for the company. Employees must justify love.

Well, I personally love this. A lot of people hate this. And this is why, you know, kind of went quote unquote viral online. Because in the memo, he said employees must justify non use of AI before requesting additional. resources. So that part, I think is one of those things that, you know, people were like, wait, you have to justify not using AI if you're just, if you're requesting additional resources. So whether that's more people on a project, more resources, more budget, et cetera.

First, you have in those cases, if you did not use AI, you have to justify, right? So maybe you need a researcher on a project. instead you're like okay you have to then justify why all the AI research tools could not do the job so definitely a bold shift but I think it's like let me be honest this is going to be very common in the coming months right as more and more enterprise companies go all in on generative AI as the technology right whether that's from Google right so if you are a Google team using Google Workspace, whether you are a Microsoft team, right? I think Apple, unfortunately, we might be waiting until like 2032 for usable AI in their

operating system. If recent trends are any indication, but I think for everyone else, right, and even if you are an Apple team, you're probably using, you know, Google's suite of products anyways, there's a good chance you are. This is common. I think this is actually a no-brainer, right? If I had a large organization and I had people requesting additional people to work on a project, to hire someone new for a project, for additional ad spend, for additional dollars,

I'd be like, have you used every single AI resource available to us yet? That's what I would ask, right? So I know a lot of people are losing their noodles over this and it's like, oh, you know, employees are having to justify not using AI when requesting additional resources. For me, it's like, duh, right? It's like, have you done your job? Yeah, do you know there's these deep research tools that I think perform at the level probably higher than junior consultants?

Yeah, so consultants that are just getting started out at big four consulting companies. If you actually know how to use these deep research tools, they're probably just as good, if not better, right? Because they're faster. And when you use them correctly, they can be just as accurate. So, um, I know,

I know people are, uh, you know, saying this is controversial. I don't think so. All right. Last but not least, as we wrap up, y'all, Google at Google Cloudnex,

I think, undoubtedly took the lead from open AI. Yes, open AI. I think is better as a single application. Like we talked about with Google and its Gemini suite of products, it's a little disjointed,

but that's also because of Google's just vast offerings across the ecosystem, right? So yes, you know, there's essentially four different ways, you know, to use Gemini. You can use it at, you know, gemini.com. So the chat bot would be number one. Number two would be AI Studio,

which is the sandbox. Number three would be Vertex, which is more for developers and enterprise. users and number four would be just across the different workspace platform so if you're in google docs as an example you can use google gemini so it's a little uh difficult so that's why i think even though google uh is probably in the lead when it comes to a i models i still think open a i is in the lead when it comes to picking an a i operating system which is something i've talked about in this show as well because from the application layer it's a little simpler right and it can be

pretty difficult, if I'm being honest, for organizations to take advantage of everything that Google just announced. There are so many things that they announced. Even executives that I talked to at the conference were mind-boggled at how much AI they had to offer. So I think it's a little easier to take advantage of it in Open AI, but yet we may be getting five new AI models from Open AI starting today. Yeah, so even by the time you hear this on the podcast, You know, since I have to improve this audio a little bit, since my mic is, you know, I don't know, in some TSA guy's locker in Chicago O'Hare, we could be getting five new models from Open AI this week. Here's what we could be expecting.

So Open AI is going to be releasing, according to some leaks and some drips, five new models this week. And those might include GBT 4.1, 4.1, nano. 4.1 mini 04 mini and full 03. Okay. Confusing a little bit. I know one of the, uh, I'd say it's probably three big, uh, three big series here. So one would be the GPT 40 would be getting updated to 4.1. So we're not sure if it's going to be called 4.1. Oh, that would be confusing if they adopted this O naming for the, uh, kind of Omni model. But I get it, uh, because they're reasoning models have the O in front. So it looks like Open AI may just be dropping the 40 and just going with 4-1,

even though you might be saying, wait, there's GPG-4-5. Yes. But it does look that like OpenAI is even trying to differentiate and have three different tiers of models. So it looks like the GPT-4-1 will be kind of the common everyday workhorse, which is your IQ model. Then 4-5, which is not getting updated in this. latest round according to reports. It's kind of the EQ model, the emotionally intelligent model. And then you have your thinking models or your O series models. And even there, it gets so

confusing because you have O1. You have O1 Pro. Then you have the O3 models such as O3 Mini. 03 mini. Now, according to reports and leaks, we're going to be getting the O3 full as well as 04 Mini. So not only are there technically now three different types of open AI models. Even on the reasoning side, the O series, there might be three different tiers of thinking models in that third tier. Confusing, I know. So like I said, GPT4-1 is probably going to be the headliner of this series.

I don't know. Maybe people will want 03 full or 04 Mini. We'll see how the benchmarks and the ELO scores come out in terms of human preferences. but GPD 4-1 is expected to be a successor to the multimodal GPT4O, which is designed for enhanced versatility and performance. Meanwhile, O4 Mini would be its latest, in theory, reasoning-focused model, highlighting OpenAI's push into specialized AI applications. So the updated model art and icons for these new releases were spotted on Open.

AI's website. Also, Open AI CEO, Sam Altman said on Sunday, he said this on Twitter. He said, we've got a lot of good stuff for you this coming week, kicking it off tomorrow. And then, you know, we had essentially on Open AI's actual website, you know, kind of leaks, kind of not, you know, some online sleuths found them. But essentially it was the image art for these new models. So anytime that Open AI releases a new model. They usually have this kind of, you know, nice little gradient, you know, image that goes with the new model. And it was confirmed.

At least there are those images for GPT41, 41, Nano, 41, Mini, 04, mini and 03. My gosh, Alphabet Soup. I will say this. We'll obviously be covering all these releases as they come out. It is a little confusing. I get it. But here's the reality.

GPT5, it looks like, it's just getting kicked even further and further out, which me personally, I'm fine with, right? So previously, CEO Sam Altman said that they would not be releasing, you know, more non-reasoning models before GPT5 came out. And they also said GPT5 was going to be more of a system, right?

So it wouldn't necessarily be a model you'd choose, but instead, GPT5 would be a system that you would talk to, and then essentially it would have kind of hybrid models or it would have an intelligent system that depending on your prompt, it would choose for you which model it should use to complete your query or to produce an output for whatever input that you entered. Personally, I'm not looking forward to that GPT5 ecosystem. I would much rather, right, hopefully whenever we get GPT5, hopefully we still get the option to

override whatever the system, the intelligent system decided. Because let me be honest. Power users, I think when and if GBT5 is released and if it is released as it has been kind of teased or promoted, I think power users aren't going to like it. I think power users like the option to be able to pull down and choose, right? Last night, I used five different open AI models all for a specific reason, right? I needed that specific model, right? Whether it was 03 Mini or, you know, O1 Pro or GPD 4O or GPD 4.5,

I needed that specific model for the specific query that I was putting in there. Would a GPT5, you know, system be able to get that right? Maybe, but I'm guessing like most of the time, it would either require way more prompting from power users or having to rerun it because you're like, oh, this is not what I was wanting. wanting the model to do. Whereas if you choose the model from a drop-down menu, I think it's a little easier.

Oh, gosh. I am exhausted from all that AI news. All right, y'all. Let me know also which one of these should we cover more this week. I think we have an open slot or two this week. Apparently a lot of you all want us to hear more about Canva. I think we are going to do a dedicated Google recap, from their Google Cloud Next conference because it was insane.

There's been, like I said, I wrote down more than 27 announcements that I thought would be extremely helpful to the everyday user and we haven't even covered them all. But let me know, live stream audience. What do you want to see more of? If this was helpful, please let me know. But let me do the very fast recap of the AI news that matters for this week, April 14th. So Google went absolutely nutty and unveiled a dozen of major AI updates. to both Gemini and its other platforms at Google Cloud.

Next, Open AI released a new chat GPT memory feature that allows it to pull from all past chats. Next, according to report, Elon Musk's Dodge team is accused of using AI to monitor federal agencies for anti-Trump sentiment. Canva announced at its conference that they're expanding beyond visual design and going all in on AI with a lot of different features, including an AI coder. Open AI is now counter-suming Elon Musk over alleged tactics to slow the company's AI progress. Anthropic has unveiled an expensive new $200 per month and a $100 per month option that doesn't really do anything too new, aside from give you higher limits. Microsoft has paused a $1 billion dollar AI infrastructure plan. in Ohio. At the same time, Microsoft is rolling out and ramping up its controversial recall AI feature

in a new Windows 11 preview build. Shopify CEO has declared that AI proficiency is mandatory for all employees in a viral company-wide memo. And then last but not least, Open AI is set to launch potentially up to five new AI models with probably at least that kickoff being started today. So we may be getting GPT41, GBT 41 Nano, 41 Mini, 04 Mini, and 03. My gosh, a lot going on. I hope this was helpful. You don't have to waste hours every single day trying to keep up with the AI news, with AI developments. You can come and stop by on Mondays for are AI News That Matters series. So I hope this was helpful.

If so, please share this with someone. You know, click that little repost button. It takes you like, you know, if you're listening on Twitter or LinkedIn, takes you like five seconds to do that and to keep your network, your family, your friends, your coworkers more up to date with AI developments because it's changing everything about how we work. Our team puts in countless hours, giving you up to date, free, unbiased information. So please, if you find this helpful, share this with others.

If you're listening on the podcast, I'd appreciate if you click that follow button on Apple or Spotify, if you could leave us a rating as well. Then go to Your EverydayAI.com. Sign up for the free daily newsletter. I hope this was helpful. Absolutely crazy week in AI news. I will 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. 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 everyday AI.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.