Ep 694: ChatGPT ads incoming, Gemini gets personal with you and official with Apple and more AI News That Matters

This is the Everyday AI show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business and everyday life.

The AI models themselves didn't change. However, just about every way that we use AI could change from what happened this week in AI news. I mean, OpenAI is going to be rolling out ads on their platform. Google partnered with Apple. Now officially they changed the way that Gemini works. And Claude Cowork could be the biggest change in all of this in terms of how we use large language models. So if you missed anything this week, don't worry. We're going to be catching you up on this week's edition of the AI News that matters on Everyday AI. What's going on, y'? All? Welcome to Everyday AI. My name is Jordan Wilson, and if you're new here, well, Everyday AI it's Every day. It's a daily livestream, podcast and free daily newsletter helping everyday business leaders like you and me not just keep up with what's happening in the world of AI, but how we can make sense of it to grow our companies and our careers. So if you're like me trying to keep up and try to get ahead, well, it starts here. But to take it to the next level, make sure you go to our website at your everyday AI dot com. Sign up for that free daily newsletter where we're going to be recapping not just these stories, but everything else that's happening in AI today. And that's how you get ahead to grow your company and your career. All right? And FYI, our inner circle is live. As well as our free prime prompt polish course. We're rolling out slow access to everyone, FYI. So if you voted in our poll last week, we're going to be giving you access here in the next day or two, making sure everyone has a smooth transition on. But eventually we will be opening it up to everyone else. All right, let's get into the AI news that matters for the week of January 19th. All right, what do large language models need? Well, they need training data, probably just as much as anything else. And that now has a big change after the Wikimedia foundation has signed a major AI licensing deal. So let's talk about it. The Wikimedia foundation has announced landmark licensing agreements with top AI companies. So they've signed deals with Amazon, Meta, Microsoft, Mistral and Perplexity, So those companies can access Wikipedia's data through their new paid Wikimedia enterprise platform, joining earlier partners like Google and Ecosia. So these agreements replace what's been going on, which is just web scraping, right? All these companies scraping the open web with now structured machine readable data delivered via specialized APIs covering over 65 million articles in more than 300 languages. So the Enterprise Platforms offers AI companies real time updates and snapshots of Wikipedia content, helping models stay current with live edits while reducing the need to clean the process of that raw and sometimes dirty and unorganized data. So by formalizing partnerships, Wikimedia aims to offset rising infrastructure costs caused by aggressive AI scrapers, including server maintenance and bandwidth expenses. The deal provides a new revenue stream for the non profit, which has seen traditional Wikipedia traffic and donations decline as AI chatbots increasingly summarize its content. This also avoids litigation by treating Wikipedia's factual content as not subject to traditional copyright, provided AI models credit human contributors and do not reproduce text verbatim. So the whole Wikipedia thing is very interesting, right, because Wikipedia is not creating technically its own new sources of information. Like I said, they have human editors essentially, you know, write what has happened and then attribute that in the, you know, in the footer, more or less. So in a way, Wikipedia was almost kind of like the original large language model, right? You wouldn't have to go read 30 different pieces of information. You hopefully got an unbiased and, you know, all inclusive kind of answer to a question. So, but for the most part, every single large language model has used Wikipedia as a main source of information. So this is a pretty big step forward in kind of figuring out what the next step is going to be. But like I've said for a long time, it's lawsuits, partnerships, or the media company goes out of business. There's really no third option. All right, our next piece of AI news. We've technically reported on this quite a bit, but now it's officially official. Apple has announced a significant multi year partnership with Google to power some of Apple's most important AI services, including a more personalized Siri. So they're going to be using Google Gemini's models. And this move marks a pretty big shift in Apple's strategy because before they were trying to do it themselves and they've been trying to do it themselves for, you know, four plus years. And they've been promoting this Apple intelligence that hasn't worked and they faced many lawsuits. So now they essentially said, hey Google, we need your help. You're really good. And that's what's going to happen now. So here's some more details on this pretty big Partnership, Apple will rely on Google Gemini's AI models to enhance services like Siri making the first marking the first time that Apple has used Google's AI or foundational technology in its ecosystem. So the partnership was officially announced in a joint statement from both companies promising innovative new experiences for Apple users. Although we've known about this for at least two months or so, right, there's a lot of reporting earlier on, back and forth. Is it going to be OpenAI, Is it going to be anthropic, Is it going to be Google? And then we heard some official reporting two months ago that it was going to be Google, but now the companies have confirmed this in the joint statement. So this is not Apple's first external AI deal. In June of 2024, Apple also partnered with OpenAI to bring ChatGPT to its Apple devices as part of the Apple Intelligence suite. So yeah, they couldn't figure it out in time. So essentially if you ask Apple's Apple Intelligence anything really, it either gave you a normal answer that wasn't really good and didn't really use your data or just said, hey, you need chat GPT for that and it just kicked you over to ChatGPT. So the financial terms of the Apple Google deal have not been disclosed, but past agreements between the two firms have been valued in the billions of dollars. Regulatory scrunchy is obviously expected here as previous deals between Apple and Google, such as the deal where Google reportedly pays tens of billions of dollars of deal a year to be the default search engine on Apple devices. Well, this one is going to also invite some regulatory scrutiny in the us, UK and EU over potential anti competitive practices and market dominance. I don't think it has any teeth though. Right. Especially when you look at, you could make the argument, right, that OpenAI is the number one player. Obviously OpenAI and Google are going 1A and 1B. So this is. Apple is not even a contender in the AI space. So I don't think this is going to be viewed as, you know, the two AI giants teaming up for a monopoly. Not at all. Because Apple is not in the conversation. All right, our next piece of AI news. Yeah, another big deal worth a lot of billions. That's because OpenAI has struck a computing partnership with with Sarabra's systems valued at over $10 billion. So the multi year deal which will run through 2028, gives OpenAI access to Sarabra's specialized wafer scale engine chips, aiming to reduce the company's reliance on Nvidia's GPU for large scale AI training and real world applications. So Sabra's hardware will be delivered in phases, supporting OpenAI's push to build a more resilient and diversified infrastructure for massive AI worklo. Industry analysts see this as a direct challenge to Nvidia's dominance in the AI chip market, marking the first time a major AI lab is betting heavily on non Nvidia alternatives at this scale. The deal also draws attention due to OpenAI CEO Sam Altman's personal investment in Sarabras, though the agreement was formalized at the corporate level. All right, so this, I mean, this isn't new for OpenAI. Right? They're striking a lot of different partnerships with different infrastructure and chip makers just to try to be able to scale things a little more quickly. On the compute side, OpenAI has been saying for more than a year that they have different features and models usually ready well ahead of time. But they said that they're not able to serve those features due to compute purposes. But maybe that'll change because they made a big switch, obviously, with the ads that we're going to be talking about here in a couple of minutes. All right, speaking of a big change, Google is finally doing something that I personally wanted, but I don't think most people are going to use. But I think you should. That's because Gemini's new personal intelligence feature is rolling out as a beta version to Google AI Pro and Google AI Ultra subscribers in the US Promising a more personalized AI experience by securely connecting with users Google's apps. So now Gemini lets users be able to link up their gmail, Google Photos, YouTube and their search history for tailored AI responses, making it easier to get personal answers about travel plans, purchases, recommendations, or anything going on in your work. So this beta feature is available for personal Google accounts, but not for business, enterprise, or education users. That's the big downside right away, right? Like, unfortunately, Google does roll things out sometimes first or sometimes just only to personal Gmail addresses. Sometimes, you know, these things might get released to businesses later. So that's the thing I'm a little on the downer about, right. So I have this on my personal, you know, Gmail Ultra plan that I have. I really wish I could get this in my work account. That would be amazing. So I might have to do one of those workarounds where I, you know, forward some of my, you know, work stuff to my personal email. I don't know, it gets a little messy. So some more details. Google says that privacy is emphasized, connecting your apps is optional. And Gemini does Not train. Train directly on your personal data like your Gmail inbox or photo library. So Gemini can reference details from your emails or photos, such as pulling up your car's license plate from a photo or suggesting tire options based on past road trips. So the feature works across the Web, Android and iOS, and users can control which apps are connected or disconnected to them at any time. So the rollout starts with US subscribers, but Google plans to expand access to more countries and eventually free tiers. So I have been trying this out since it was announced. I did get access to it fairly early. I really like it. And here's the thing, at first I was a little frustrated because I'm like, you know, it's pulling things from, you know, my email or, you know, certain history from my Google account that are maybe too far out and no longer relevant. But there is kind of a setting that's a little hidden where it's kind of like a custom instructions, but you can essentially just chat with Google Gemini to update it from anything in your personal intelligence that maybe isn't right or is outdated. So at first I was like, ah, there's some things I'm not really feeling. But once I found that not really hidden setting, but kind of hidden setting, I think that it improved a lot. All right, speaking of Google. Yeah, they're trying to completely change the way that people do shopping. And they're starting this new venture by partnering with the world's largest retailer. So here's what's happening. Google is expanding its Gemini AI chatbot with new shopping features, partnering with major retailers like Walmart, Shopify and Wayfair to turn the app into a virtual merchant and assistant. So here's what that means. Essentially, you will be able to check out in AI mode in the very near future without even having to visit these websites like Walmart, Wayfair, Shopify, et cetera. So the entire shopping experience is moving on to Google AI mode, my hot take. Not that you need it here on the Monday AI News Show. I've been saying for a long time the traditional Internet is going to die. And this is another indication of that, right? You're not even think if, if you're an avid shopper at Walmart or Shopify or Wayfair, think of how much time you probably spend on those websites. I remember, I don't know, looking for some piece of furniture on Wayfair before, and I spent so much time on that website. Right. And these companies obviously make a lot of money because they serve you different ads while you're on their website, they obviously, you know, sell you products, but pretty big change in how AI and shopping can get done. So the new instant checkout function will let customers buy products from participating retailers directly inside the Gemini Chat without needing to visit a separate website. So Walmart customers who link their Walmart and Google accounts will receive personalized recommendations based on their previous purchases and can combine new purchases with their existing Walmart or Sam's Club online carts. Payments will be processed using cards linked to Google accounts at first, but Google plans to add PayPal and other payment providers soon. So this move is part of a broader race among Google, OpenAI and Amazon to create seamless AI powered shopping experiences that take users from browsing to buying in a single conversation. All right, well that might change how shopping is done and our next piece of AI news might change how drug discovery is done. That's because Nvidia and Eli Lilly have announced a major one billion dollar partnership to open up a cutting edge AI drug discovery lab in the Bay Area aiming to speed up the development of new medicines with a five year $1 billion commitment. So the new lab is reportedly going to be opening by the end of March and will house teams from both companies working side by side to accelerate drug discovery using AI and high performance computing. So Lilly's newly built AI factory, powered by Nvidia's DGX Superpod and future Vera Rubin Systems, is designed to scale up drug research in a hybrid cloud environment. So the initiative centers on generating high quality ground truth data through large scale lab experiments which will train next generation biology foundation models with multimodal data. So the goal is to create a closed loop feedback system. Better experiments produce better data, better data improves models and improved models make future experiments more targeted and efficient. So the scope extends beyond research and development with plans to apply AI across Lilly's manufacturing and commercial operations, potentially transforming the entire business. So the partnership builds on Lilly's October announcement of creating the most powerful pharma owned AI supercomputer aiming to industrialize drug discovery in development. All right, and this one, even though the chat GPT ads is the story that grabbed all the headlines and the chatter, this one from Anthropic might be one of the bigger, I don't know, new AI announcements we've seen in the last year or two. So Anthropic is making headlines with the launch of Claude Cowork, a new AI agent designed to handle everyday computer tasks for non technical users. So the move is seen as a major step to just bring AI that works the way we do. So Anthropic has released Claude Cowork as a research preview first for MAX subscribers, but they actually just rolled it out to the normal 20amonth paid users as well. So here's what Claude Cowork is. It can autonomously manipulate, read, analyze and create files on a user's computer, making it a hands off assistant for tasks like organizing downloads, converting receipts to spreadsheets, and drafting documents from desktop notes. So unlike the developer focused Claude code Cowork, Codework is built for non technical users and aims to simplify complex tasks without requiring coding knowledge. Here's the thing that's actually kind of crazy. Anthropic's team said that they built Cowork in just a week and a half and it was built using Claude code. So Claude Code built Claude Cowork in a week and a half, which if you go use Claude Cowork, you're like, okay, this is crazy that this was built by AI in a week and a half. So here's what it is. Strip away the bullet point features, all the technical terms. If you've ever used Claude Code, it's very good. It's one of the reasons, I think, that Anthropic models have continued to just soar in popularity, because their direct connection inside cloud code. Right? It's a coding tool. But what Anthropic found, and what myself and a lot of other people were realizing was people were using Claude code for non coding reasons. They're like, wait, this thing can control your computer? Yes, that's great if you're trying to code something and you know, access your local files. But what about if you just want this thing to do work for you? Are you still running in circles trying to figure out how to actually grow your business with AI? Maybe your company has been tinkering with large language models for a year or more, but can't really get traction to find ROI on gen AI. Hey, this is Jordan Wilson, host of this very podcast. Companies like Adobe, Microsoft and Nvidia have partnered with us because they trust our expertise in educating the masses around generative AI to get ahead. And some of the most innovative companies in the country hire us to help with their AI strategy and to train hundreds of their employees on how to use Gen AI. So whether you're looking for ChatGPT training for thousands or just need help building your front end AI strategy, you can partner with us too. Just like some of the biggest companies in the world do. Go to your everydayai.com partner to get in contact with our team, or you can just click on the partner section of our website. We'll help you stop running in those AI circles and help get your team ahead and build a straight path to ROI on Gen AI. And that's essentially what Claude Cowork is. It connects your. It can control your local computer, your local files, your terminal, etc. It can control your local browser as well as controlling anything in the claw, in the, In. In the. This is hard to say in the cloud from Claude. That one was hard, I don't know why. In the cloud from anthropic. Right. So any of your skills, your information, right. That you have in Claude, combine that with its ability to control your desktop, which is huge. Control your browser via a Chrome extension, right, and you have what I think is the first kind of official foray into the future of work. And this is right now available to only Mac users. And we are doing a show on this on Wednesday for our working Wednesday, FYI and full transparency. I had to pre record the show because I'm actually on a little trip right now and feeling a little sick too. Great. You know, you always want to rest and recharge and then you, you know, don't feel the best. However, I did have to record it right before they announced that this was coming out to all pro plans as well. So I think in the show I probably mentioned that you have to be on the 100amonth max plan, which you did. And then an hour after I was done, they rolled it out to everyone. So if you're on a paid plan and if you have a Mac because you need Mac because it is a program that you download and it run, it runs on your computer. So you have to have Mac and at least a pro plan. And you can use Claude Cowork. So make sure to join us Wednesday as we go over that. All right, and our last big story, at least OpenAI is bringing ads to chat GPT, which might sound surprising, but if you listen to this show, I've been saying for many, many months that ChatGPT was rolling out ads. We actually did a pretty, which I thought was a really great show detailing it all the nine reasons why ads are coming and you can't do anything about it. And people at the time are like, no, this is never happening. And I'm like, yes, it's absolutely happening. And now we get the official word from OpenAI that yes, it is happening. So OpenAI will begin showing ads to adult users on the free version of ChatGPT starting in the US in the coming weeks making its first official move into digital advertising. So pay subscribers, including Plus Pro and Enterprise users, will not see ads, keeping their experience ad free. However, there is a new tier of plan that was first more of a global plan offered in other countries called Chat GPT Go. So this lower price plan, I believe it's about $8 a month and they opened up the plan to anyone in the world. So this new paid plan will get ads, which is pretty, pretty significant if you ask me. So that newly launched low cost Go subscription will also include ads for users who are on that plan. So ads will appear at the bottom of ChatGPT's answers and will clearly be labeled to avoid any confusion. So it's not like ChatGPT is going to be trying to tell you something and it's not going to be marked as an ad. So it will, at least according to OpenAI and what they release, it will look traditionally like older Google Ads that were very easy to see. Oh yes, this is an ad, but will appear actually at the bottom of the screen and not at the top like traditional ads may appear in search results. So OpenAI says ads will not influence ChatGPT's responses and promises never to sell users data to advertisers. So users under 18 will be shielded from ads. And sensitive topics like politics, health and mental health are excluded from ad placements. CEO Sam Altman previously expressed concerns about ads eroding user trust, but now says the company hopes this business model can work for those that want free AI access. So the move comes as OpenAI faces major spending commitments, including a reported over $1 trillion in infrastructure deals signed in the coming years. So the company is on track to reach a 20 billion annualized revenue rate. And ads could help meet those ambitious financial goals. So users will be able to learn why they see specific ads. They can dismiss the ads and submit feedback to improve the experience as OpenAI refines the rollout. Here's the thing. I know a lot of people are going to lose their noodles over this, right? Because OpenAI has nearly 800 or, sorry, almost 900 million weekly active users right now. But their last reported official number is 800 million weekly active users. And they've said that only 5% of those users are paid. So that means that there is probably more than 700 million people who are using ChatGPT every week on the free plan. And here's the thing, I don't care what you're talking about. If you're not paying for something, usually that means you are the service, right? Like eventually you're gonna have to pay for something in some way, shape or form. You can even say, oh, well, email's free. Yeah, it is, right. But a lot of free email providers over the years have had to monetize. And what that means is, you know, they change their storage and you used to get generous storage and now it's like, oh, if you want to receive an email, essentially you gotta go delete everything or start paying us every single month. So this to me is not a surprise and I think it's obviously the right move from OpenAI. They've given away the world's most powerful intelligence in what I think is one of the best form factors. I think Chat GPT has the best user interface, user experience, and it's not really close. And I think most people will agree with that. That's not like a hot take of any sorts, Right? So they have the most users, they have one of the best products. Right. They don't have the best models per se anymore, I don't think. Right. Yes, I think GVT 5 II Pro is the absolute best single model, but it is very slow and time consuming, so that's not what most people are going to use. So I think when you talk about day to day, like daily driver models, I think it's still Gemini 3 Pro and maybe even Claude Opus 4 5, depending on what you're using it for. So OpenAI still has. Traditionally they kicked off this generative AI phase. They've offered an amazing product for free to hundreds of millions of users. And if these free users want to keep using the best product, it costs a lot of money. Right. I've been saying this for a long time. OpenAI CEO Sam Altman has said on a lot of their paid plans, OpenAI loses money. So obviously on the free plans, they lose a ton of money. And just the fact that you could pay $20 a month or nothing and get this level of intelligence is absolutely silly. So I do think that this is the right move and I'm all on board. Right. I'm a little different. I've talked about it a lot over the years on the show. I'm someone that's like, yes, company, take all my data, I don't care. Take my health data, take my financial data, don't really care. Right. As long as it helps me, you know, make better decisions in my life. And as long as at any time I can, you know, kind of click the disconnect or delete button if I have that option, right. So I do think there's going to be a, an uproar, right, because as they get this right and as they refine it, they're going to make some mistakes, right? There's people are going to screenshot things that are going to go viral and look really bad, like, oh, I can't believe, you know, an ad was showed like this kind of ad was shown in this kind of way. But it is so much different than Google search. Right? And as someone that, right. I've had a marketing agency for eight years, so I've spent millions of dollars managing ads or, sorry, I've managed millions of dollars in advertising. And it's a much more straightforward to, you know, manage advertising that's keyword based, right? It's like, oh, when someone is searching for a certain service or product, that's what you're bidding against. This is completely different because think of the level of insights that users give to large language models like Chat GPT and the amount that they know about users is huge and it goes so much further than actual keywords. All right, so like I said, go listen to the, the Chat GPT ads episode if you want more on that. That was episode 641 if you wanted to go check that out. All right, that's it for our main news stories. But there was a ton, my gosh, a ton of other news stories that didn't make our top, top stories cut. So, so let's go over a quick bullet point rundown of what's new and what's next. These are kind of, some of these are rumors, some of these are just smaller news stories that didn't make the cut. So let's get straight into it. So ChatGPT workspace apps in custom GPTs are rolling out for beta for business users. So yeah, you'll be able to use apps in your custom GPTs, which is pretty cool. BlackRock raised 12 and a half billion for AI infrastructure fund targeting US data centers. Some hiring drama. So OpenAI rehired Barrett Zoff and Luke Metz who both left OpenAI to be co founders at Thinking Machines Lab. But now they are back with OpenAI. Anthropic is testing knowledge bases for Claude Cowork with topic specific persistent memories. Manus AI partnered with similar web data for paid accounts which is actually huge, y'. All. Like similar web is a treasure trove of data and usually very expensive and somehow you now are getting certain level of that for free if you are on a Manus paid account. Google launched VO31 updates so now you have 9 by 16 vertical video and 4K options for paid plans. OpenAI will be rolling out Chat GPT jobs according to leaks. A career assistant for job search, Resume and Advice. OpenAI invested in Merge Labs brain computer interface company which their CEO Sam Altman is a co founder in. OpenAI launched ChatGPT Translate to compete directly with Google Translate. Meta launched their Meta Compute initiative to build the tens of gigawatts of AI infrastructure. Claude added AI health integrations. Google's Anti Gravity is rolling out agent skills. Some new leaks show that Anthropic is testing new widgets for Claude Web app including weather stocks, sports scores in places with maps. I like that, you know. OpenAI added that in ChatGPT a couple months ago. Perplexity has had that for a while. Google's been rolling it out. So Anthropic getting on the board there as well. Google released Translate Gemma, their medical model in an open source way. That or sorry, not their medical model, their Translate model that delivers translation in 55 languages. ChatGPT lost access to WhatsApp after Meta enforced their broad AI usage restrictions. So yeah, Meta said, yeah, you can't use WhatsApp anymore. Chat GPT Some new reports show that Anthropic and OpenAI are getting closer to an IPO. Cursor's team, this Was Nuts detailed how they used GPT 5.2 to build an actual working browser. So if you don't know anything about coding or browsers, that's absolutely insane. This thing was 3 million lines of code and it was actually a working browser that GPT5.2 built, right where normal browsers usually take years and millions of dollars in enormous engineering teams. So yeah, apparently cursor just used GPT5.2 and got the job done. Microsoft launched an elevate for educators to help help teachers better use AI. A couple OpenAI things. OpenAI improved Chad GPT's memory and rolled out a better way to remember past chats. They also released GPT5.2 codecs, enhancing coding abilities for people using those models. They also acquired Torch Health to accelerate Chat GPT Health, which they announced last week. And Healthcare AI adoption at large. Salesforce launched AI Slack Bot for businesses enterprise customers as well. I think that'll be huge, right? So many people, often with all the Slack messages coming in. So pretty big I'd say for this new AI Slack Bots project. Sorry, losing my voice here. And then last but not least, Google Open Source, their Med Gemma 1.5 Imaging Text Model and Med ASR Medical Transcription AI all right, we covered a ton in this AI News that matters. Remember, we do this almost every single Monday, whether I'm working or not, right? Like, like here I am on vacation bringing it to you, right? So every single Monday we do our AI News that matters. Make sure to join us Wednesday for our AI at Work. We're going to be going over Claude co work and we're dropping one or two episodes as well of our new Start Here series, helping beginners and advanced users double down on AI at the beginning of 2026. I hope this was helpful. If so, tell someone about this. Repost this if you're listening on LinkedIn. If you're on the podcast, please take like 30 seconds. Go ahead, follow the show and leave us a rating. I'd super appreciate that. That's the only way we can keep this thing going. And then go to your everydayai.com Sign up for that free daily newsletter and thanks for tuning in. See you back tomorrow and every day for more Everyday AI. Thanks, y'. All.

And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going for a little more AI magic. Visit youreverydayai.com and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.