Everyday AI Podcast – An AI and ChatGPT Podcast - EP 650: Apple and Google’s deal for AI-powered Siri, OpenAI’s dragged for fed financing talk, a new open source LLM leader & more AI News That Matters

Episode Date: November 10, 2025

Could a model named…. Kimi K2 Thinking wreak havoc on the U.S. AI scene? 🤔And why is Apple gonna pay Google a billion dollars a year? 🤝And is OpenAI suggesting a government insurance if AI fails? 🛟So many AI questions. We’ve got your AI answers. On Mondays, Everyday AI brings you the AI News That Matters. Newsletter: Sign up for our free daily newsletterMore on this Episode: Episode PageJoin the discussion: Thoughts on this? Join the convo and connect with other AI leaders on LinkedIn.Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineupWebsite: YourEverydayAI.comEmail The Show: [e-mail] with Jordan on LinkedInTopics Covered in This Episode:Apple-Google Gemini AI Siri PartnershipOpenAI Federal Financing Backstop ControversyMoonshot Kimmy K2 Open Source LLM BenchmarksAnthropic's $350 Billion Valuation SurgeMicrosoft Launches Humanist Superintelligence TeamIBM Announces AI-Driven Job CutsMeta's $600B AI Infrastructure InvestmentGoogle's Project Sun Catcher: Space Data CentersOpenAI $38B AWS Cloud PartnershipGoogle Gemini Deep Research Across Workspace AppsTimestamps:00:00 Apple Chooses Google Gemini AI06:12 "AI Shorting Sparks Market Stir"08:44 Google Deepens AI Race Investments12:33 "AI Model Performance Concerns"16:38 "Microsoft’s AGI Push Begins"18:25 "Humanist Superintelligence vs AI Autonomy"21:08 IBM Cuts Jobs for AI Focus25:57 Meta's AI Rumors: Open vs Closed29:04 Google Stress-Tests Space TPU Chips31:52 "OpenAI Clarifies: No Federal Backstop"35:10 "OpenAI's Mixed Messaging"38:17 AI Updates: GPT-5, Gemini, Research41:12 Everyday AI: Subscribe & ExploreKeywords:Apple and Google AI partnership, Gemini AI, Siri AI update, billion dollar AI deal, Anthropic AI, OpenAI federal backing, AI government bailout, Michael Burry AI short, Nvidia stock drop, Palantir stock, US Department of Justice antitrust, Tim Cook AI acquisition, AI production risks, Aria AI sandbox, Microsoft superintelligence, Mustafa Suleiman, humanist superintelligence, frontier AI models, recursive self-improvement, Send Everyday AI and Jordan a text message. (We can't reply back unless you leave contact info)

This is the Everyday AI Show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business, and everyday life. Meet Firefly AI Assistant, now live 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. Two tech titans have formed an unlikely partnership on AI that's worth a billion dollars a year.

The leading AI startup is in hot water after kind of not really, but kind of suggesting federal backing for AI projects. And there may be a new open source AI model that could shake up the AI space way more than Deepseek ever did. Yeah, a lot of drama, a lot of new partnerships, a lot of questions that hopefully will get answered today on everyday AI. What's going on, y'all? Welcome to Everyday AI. My name's Jordan Moulson, and this thing's for you. It's a daily live stream podcast and free daily newsletter that helps everyday business leaders like you and me make sense of the who's who, the what's what, and the who did what? Yeah, every single Monday, we do this thing called the AI News That Matter.

So if you're looking to grow your company and career by understanding. standing and leveraging AI. This is where it happens. And Mondays, especially, if you can only join one day and you have to know who's moving, who's shaking, this is where you do it. So if you haven't already, please make sure you go to our website at your everyday AI.com. Sign up for the free daily newsletter. We do it each and every day, Monday through Friday. We're going to be recapping the highlights from today's show as well as giving you all of the other AI news that's happening today. But for the news that matters for the week, let's get into it. And And let's start with the two tech titans shaking hands once again.

That is Apple paying Google a billion dollars a year to use Gemini. So according to reports, Apple is making headlines for choosing Google's Gemini AI over others to power the next generation of its smarter, well, maybe smarter, Siri assistant. So Apple will reportedly pay a billion dollars a year to Google for a specialized version. of Gemini AI, a version with 1.2 trillion parameters, according to Bloomberg's Mark German. So the decision reportedly came down to cost with Google's offer beating Anthropics $1.5 billion annual price tag. So the new Gemini model will reportedly launch on Apple's private cloud servers in spring 2026 handling complex Siri tests, while Apple's own models continue

to run on devices for personal data. So who knows that that'll work, but at least the ones with Gemini will probably work. So Apple does not plan to really highlight or publicize Google's involvement in Siri and its future marketing, keeping the partnership mostly behind the scenes. So despite lagging behind rivals in AI innovation, Apple is leveraging partnerships with leaders like Google, and they have a current partnership with OpenAI to add generative features to its devices. Well, these two tech titans, even though they're competitors, especially on the hardware side, they're actually frenemies because Google has been paying Apple between $15 to $20 billion a year

to remain the default search engine on Apple devices, a deal that has drawn antitrust scrutiny from the U.S. Department of Justice. And Apple's reliance on Google for both search and AI highlights its strategic choice to just buy or kind of rent versus build the core technology in some areas because they've failed to do so. So Apple CEO, Tim Cook, has not ruled out acquiring an AI company, though they can't do that to Google, obviously, signaling that Apple may still be making bigger moves in the space. So yeah, if you want, I'm not going to give you my opinion on this because we actually had a poll on the podcast. So, yeah, if you're listening on the podcast on Spotify or Apple, Actually just Spotify.

Make sure to scroll down. And most days now, we're going to have a poll. And you all said you wanted this as the hot take Tuesday. So if you want to know more on this story, my opinion, make sure to tune in tomorrow. That's going to be our hot take Tuesday episode. All right. Next, no federal bailout for AI. So the White House is taking a decisive stance on artificial intelligence policy as the U.S.

seeks to strengthen its position in the fast. evolving global tech race. So here's kind of what happens. And we're going to unpack this a little bit more later in one of our other news stories. But David Sacks, the White House AI in CryptoSar, announced there would be no federal bailout for AI companies. And here's why.

The statement came after Open AI made some government backing comments, which we'll get to later. And as well, a bold bet against AI, from Michael Burry. So Michael Berry is the famed investor who rose to being a household name after predicting the 2008 housing crash. So yeah, the big short, right, the housing crash was Michael Burry. And he made headlines by acquiring over a billion dollars in put options against AI leaders in Vidia and Palantir. So yeah, essentially the big short guy is shorting against a AI, and that's caused a lot of stir in the market, and it sets global stocks down.

But specifically, the two that he took positions against, Nvidia dropped about 7% in the days after that, and Palantir dropped 11%. So between the Open AI comment, kind of hinting at government backing for AI projects, and the big short kind of play there, it kind of has. had a lot of people wondering, well, is the government going to get behind, right? Like how they bailed out the banks. And David Sachs essentially said, no, not a chance. And Sachs also noted that the U.S. currently has at least five major frontier AI model

companies assuring that if one fail, others are ready to step in and fill the gap. So yeah, a lot more on the open AI side of that later. And if I'm being honest, I thought, right, because I put it out to a poll to our audience. I thought people would want to hear about this as the hot take Tuesday, but you all wanted the Apple and Gemini. So let me just give you my quick two cents on this. I think this is going to be an evolving conversation. Right.

So without getting too much into the geopolitical side, a lot of people, when they're looking at AI, they're looking at it face value. They're looking at it as a technology, right? Like, I don't know, a CRM or the, the cloud or social media. It's not that, right? Without getting too deep into it, let's just say this.

I think AI is ultimately going to become much more important than even the military. Because if you really can control AI or achieve artificial general intelligence before any other country, what that means from a military standpoint translates to the geopolitical and obviously on the commerce side. So AI is extremely important. So I do think there's actually cases to be made for the government to have some sort of backing or skin of the game when it comes to the AI infrastructure. Obviously, they've been supporting it through Stargate and other initiatives,

but I actually don't think it's a crazy kind of road to go down. But more on that here in a couple minutes. All right. Next piece of AI news. Well, more Google news. that's because Google had a big deal with Anthropic, which saw their valuation soar. So Anthropic may soon be valued at over $350 billion as Google just reportedly entered talks into deepening its current investment in Anthropic.

So Google's move reflects its aggressive commitment to AI and signals a new wave of competition in the sector. So yeah, it's not Open AI. running away with this with a reported one trillion dollar IPO that may or may not be happening. Now all of a sudden you see Anthropics valuation soaring. So that $350 billion figure is a significant leap from previous rounds, like twice as much as from earlier this year, highlighting the just straight up silly growth in the sector in the sector as investor confidence around AI leaders continues to soar. Google already holds an equity stake in Anthropic, but this new investment round could cement its influence as the generative AI kind of competitor to themselves and Open AI, right?

So I've always said it is kind of a two, kind of a two horse race between Open AI in Google. So some people might be confused. Okay, why is Google sticking billions of more dollars into someone that is technically a competitor? Well, hey, if Anthropic ends up eating some of the market from, let's say, Microsoft or Google or Open AI, well, Google will still profit. So it makes sense to even still be funding one of their bigger competitors, especially on the coding side, right? Gemini 2.5 Flash, Gemini 2.5 Flashlight. He's gotten extremely popular. We know there's going to be a Gemini 3 coming any day, week or month now.

So yeah, pretty, pretty interesting news. All right. Speaking of interesting news on the AI model front, a new open source model benchmarking above frontier proprietary models. Yeah. So if moonshot AI's benchmarks are to be believed and we will see just how well they're verified by third parties, we could have that. So moonshot AI released a new open weight model. called Kimmy K2 Thinking, and its posted benchmark scores surpassed Open AIs, obviously proprietary GPT5, on several key tests. So the most, probably one of the more attention-grabbing stats,

is K2 Thinkings 60.2 on browser comp, a demanding web search and agentic reasoning benchmark, which beat GPT5's 54.9% and Anthropic 4.5 thinking, add to 24%. So K2 is released under a modified MIT license on hugging face that allows full commercial use, but requires prominent Kimmy K2 attribution only if a product serves over 100 million monthly active users. All right. So like, what does this mean? Well, an open source model, if you don't follow it, that means you can actually download it and you can run it and you don't have to pay anything. Obviously, there's options to pay this via the API. to use it on their website.

But having an open source model that is now entering frontier model territory, this is pretty big. Right. We'll see how well it sticks because we have seen some of the open source models specifically from China that get released with very high internal benchmarks. But then when they go through other benchmarks and other third parties, they seem to waver a little bit. We saw that with Quinn 3 coming out internally, what was reported,

with some very lofty benchmarks, but through third parties, they maybe were a little overfitted for specific tests. So we'll see how, you know, as everyone gets their hands on this and puts it through the paces, if Kimi K2 thinking is still at that level. Is it going to be a GPT5, uh, Claude Sonet 4.5 level model? If so, that's silly. It is silly that we could have an open source model, right? I think this is what the premise of, uh,

Meta's Lama for Behemoth was supposed to be, and they just didn't release it, I think, because they weren't happy with the result. So we kind of got a very good version of what we never actually received from Meta with the Lama for behemoth, but y'all, this could be pretty big. And for whatever reason, this isn't kind of carrying with it the same kind of cloud of uncertainty as Deep Seek was, although I think that might unravel, because there was. was some original reporting with an unnamed source that said the Kimmy K2 thinking model only cost $4.5 million to train. So I don't believe that. I don't believe that's true, right? We heard the

same thing, similar Pinocchio tales that were told by Deep Seek back in January and the media ran with it until a semi-analysis report debunked those training costs claim. So I wouldn't pay attention, right? Like, if this has caught your attention and you're like, okay, should my company be using Kimmy K2 thinking, well, maybe, but don't pay attention to the $4.5 million. Pay attention to your own internal benchmarks. But this could be a pretty big shift. I think what this might ultimately do is it might force Google to release even more versions of their Gemma, which is an extremely capable open source model. It might force open AI to push more on the GPTOSS, right?

Very good open source model, but that's more GPT40 territory, not GBT5. So this could really shake the proprietary market and not just the open source market. So we're going to be continuing to keep our eye on Kimmy K2 thinking. We'll probably put it through the paces ourselves, right? But yeah, you do have to have a pretty powerful computer. I mean, Apple's probably happy about this. They're probably going to sell some more of their Mac studios. Probably Nvidia is very happy about this because a lot of people think, okay,

what does this mean in the long term for, you know, running AI in the cloud like everyone does, right? Some people are thinking, oh, well, no one's going to go use or pay $20 a month for open AI. That's not the case. It's still you have to have a, yeah, like a $10,000 plus setup to run. run it and it's still not entirely fast, but it is secure to be able to download something and run it locally on your machine. Pretty big. We have to pay attention. All right. We also have to be paying attention to our next news story. That's because Microsoft, fresh off there, I wouldn't say break up with Open AI, but kind of now that the relationship has changed, Microsoft is now

leaning into superintelligence, but not in the same way. So Microsoft is a formally launching. launching a super intelligence program aimed at building safer, human-centered frontier AI models. And they're calling it the human-humanist superintelligence. So Microsoft AI chief Mustafa Soleiman will lead the new division. So the new Microsoft AI super intelligence team with chief scientist Karen Simeon in core MAI or Microsoft AI researchers shifting focus to this initiative. So Suleiman says the goal is a humanist superintelligence designed to be subservient to humans and to keep people at the top of the food chain, emphasizing that humans matter more than

AI. So Suleiman cautioned that results will take time, noting it will likely be a good year or two, interesting, before the superintelligence team produces frontier models. So he didn't say it's going to be a year or two before they get to something that they might deem super intelligence. It's just until they start releasing what they would say is a frontier model. So the move follows Microsoft's prior contractual limits with OpenAI that barred a direct pursuit of AGI.

And with this new agreement, the new Open AI restructure that we've been covering over the last few weeks, Microsoft is now openly competing in the push toward AGI and beyond. So yeah, it was kind of this big question mark in the room. What happened when this restructured came along? Would Microsoft openly go after AGI? And well, they kind of did, right? They're more going with this humanist superintelligence, which I don't know, to me it seems kind of counterintuitive.

Because when you talk about super intelligence, that's in theory when you kind of get past AGI. So right, when you get to right now, we would probably, I think most people would say we have artificial narrow intelligence or ANI. AGI would be the next step. And that's been a moving goalposts for many years, right? I did a show like two years ago looking at the definitions of AGI from like 2005, 2010, 2015, by those definitions were well past it. But by today's definitions, probably not. But right, so artificial super intelligence is essentially when the machines don't need humans anymore.

So to me, I'm almost like scratching my head. So what is this humanist super intelligence? Because if you do truly achieve, you know, super intelligence, that means the models don't need humans, right? It's their self-learning. So a little more on that, Soleiman did confirm the team will explore recursive self-improvement or RSI because performance games will come from it. while also recognizing the risks and the need for clear decisions on when and how systems can learn autonomously. So yeah, that's more of a, you know, future episode to tackle, to tackle, like what the heck is humanist super intelligence? Can, you know, can that even exist?

I don't know. I personally think, you know, super intelligence is probably a scarier thing, a little, you know, more uncertain and a little, you know, making people feel a little. little uneasy, right, when, you know, the AI is just going to make decisions on its own, build its own models, you know, has humanoid forms, right? That's when it starts to get a little like, so I guess you need something like humanist super intelligence. So with humans in control, but I don't know, at a certain point, I think if you achieve super intelligence, I don't know, are humans still in control? Episode for another day. All right, before we get into the rest of our AI news stories, a quick word from our sponsors. It's a problem.

I hear all the time. The gap between the AI champions and everyone else in your organization is sizable. You might have half a team that wants to fine-tuned models by hand, and the other half doesn't know what an API is. How do you get them working together on AI that moves the needle without creating a security nightmare? That's where area really shines. They built one platform with three ways to work. Your developers can go full pro code and build custom agents with Python. Your business analyst can use the low code tools or your domain experts who've never coded. they can use the drag and drop no code builder. Everyone's building in the same secure governed environment.

No shadow IT, no security gaps. And because they're model agnostic, you're not locked into one vendor's ecosystem with area. You can even A.B. Tests different agents against each other. Try different models. And when you're ready, deploy to production on the same platform. Your AI strategy should unite your team, not divide them. Check out area in today's show notes or on our website for a free trial.

Go to AIRIA.com. Get rid of the AI gap and move forward with a more resilient AI ecosystem. All right. Getting back to our AI news, IBM cutting thousands of jobs as it sets its focus on its AI business. So IBM said that it is going to lay off a quote unquote low single digit percentage of its roughly 270,000 global employees this quarter. a move that underscores a shift toward higher growth AI consulting in software, according to Bloomberg reports. So even if it was a 1% cut, which it's probably going to be more than one, that would be 2,700 people. So we are talking thousands, maybe potentially 10,000 or more,

providing a clear sense of scale, just for people that are trying to say, okay, what's a low single digit percentage? So probably somewhere from 3,000. to 10,000 plus. So reports are showing that there's several thousand employees who have already received notices, giving them 30 days to find a new role inside IBM or face termination. An IBM spokesperson told CNBC that while some U.S. roles may be hit, the company expects U.S. headcount to remain flat year over year, signaling internal reshuffling alongside

these cuts. So IBM says the pivot is driven by rising demand in AI. CFO Jim Kavanaugh told analysts that roughly 80% of AI consulting in software clients in the last six months were new to IBM. So that's according to the New York Times. So on the AI consulting side and on the AI software side, IBM is booming. So they are going to be growing in those areas, but cutting some jobs. in others. So bookings for AI consulting and software reached almost $10 billion in the third quarter, up $2 billion from the prior quarter, according to the Wall Street Journal. So IBM reported $16 billion in quarter three, 2025 revenue up 9% year over year. So yeah, revenue, growing, opportunities, growing, but job losses also growing. All right. All right.

Another thing that's growing is Meta's investment in AI. So Meta says it plans to invest $600 billion in U.S. infrastructures and jobs by 2028. Let's see, how many researchers can that buy from Apple? Like 10. So that's a scale that if Realize would make it one of the largest private buildouts tied to AI and data centers in the country. So yeah, this is something kind of had already been reported. CEO Mark Zuckerberg was actually caught on a hot mic back in September at a White House dinner, kind of throwing this figure out.

But meta did confirm it and release some new details. So the company frames the spend as a national competitiveness play, saying data centers are crucial to maintaining America's tech edge and delivering the next wave of AI products. So the $600 billion figure, like I said, matches the number that was floated out in September. And he said he wasn't sure what number you wanted to go with. So it could have been a couple different numbers, but apparently they settled on 600 billion. What a conversation to have, right? When you're talking about hundreds of billions, you're like, I'm not sure where we're going to land.

Let's say 600 billion. So meta says it's existing U.S. data centers have supported over 30,000, skilled trade jobs in 5,000 operation roles since 2010. So the company links the investment also to a push for personal super intelligence. A term, like I said earlier, for AI systems that surpass human cognitive abilities. So nothing, I guess, shocking here because we've been hearing about this number now for a couple of months. but I guess we're going to see what else now becomes official because this had been a huge number thrown out there, right? 600 billion and everyone's like, wait, is this real?

So essentially meta says, yes, it is real. So we would expect some other now official commitments coming from meta. Like what are they doing, right? Obviously, they have some big data centers in the work. They're spending literally billions of dollars sometimes on a couple of researchers at a time. So yeah, we'll see if that actually pays off on the model side. Right. There's been rumors swirling around over the last three to six months after their big shuffle at meta.

They've been cutting, you know, people from prior AI teams and kind of restructuring everyone. So there's been some rumors that maybe they would stop going down the open source route and maybe might go to proprietary models, which would be wild, right? because Open AI, everyone's like, oh, Open AI, you're not open. And well, now they do have open models. So it would be quite a shift in events if Meta ended up going proprietary or closed source. But that hasn't been beneficial. But I mean, we'll see.

Because if you are investing $600 billion, right, where are you going to make the money? Where are you going to make the money? Because Meta's previous open source models have not been. and competitive, right? If you look on the charts right now, even on open source charts, they're not a top model, not even really close, especially with what we've seen coming out of China, right? And then you're saying, okay, well, maybe, you know, meta's going to make a pivot and go proprietary.

Okay, maybe. But, I mean, everyone else is starting to now kind of go in meta's water, specifically open AI, right? Their SORA app has been extremely popular. technically it is a social network. So they're kind of, you know, kind of stomping in meta's backyard. So we'll see if meta also returns the favor and they're like, okay, yeah, we're going to release a model that is, you know, bonkers off the charts.

Maybe they get it figured out with billions of dollars they're spending on people and maybe they end up, you know, charging for it. Time will tell. Also, time will tell if we're going to be running data centers in space because one of the most probably out of left field AI news stories of the week is that Google's Project SunKatcher, yeah, trying to explore AI data centers in space. So Google announced that they're testing a plan to run AI computers on satellites. So according to Google, this could ease local opposition to new data centers while opening a new way to scale AI.

So to put it very simply, because I read this a couple of of times and I'm like, wait, what? So the idea is, and I'm dumbing this down here, the idea is to fly small satellites really close together so they can share data quickly, staying within about a kilometer of each other and often just a few hundred meters apart. So Google says this tight, no thrust formation mostly lets the satellites drift together with small course fixes as needed. So yeah, if you're wondering, like, how can they, uh, you know, fly all of these satellites, that are technically going at light speed, right, even if they're just floating out there in space,

well, it's that whole no thrust formation apparently. But the big hurdle is whether regular off the shelf chips can actually survive in space. So Google right now is stress testing, its latest cloud TPU chips with high energy particle blasts to mimic space radiation. So yeah, Google's, you know, putting their TPU chips, in the microwave, not in the actual microwave, but they're putting them through some space radiation emulation to see if it'll actually even work. And early tests have been promising so far, showing the chips held up longer than expected

before errors appeared, which is a promising sign for using commercial hardware in orbit. So Google's goal is for each chip to last at least five years in space, which is the minimum, they say is needed for this to make sense fiscally. So the first two test satellites could launch by early 2027. And those early tests will obviously be expensive and Google will be burning cash, but more of a long-term investment to see if this will work out. Because they hope that by the mid-2030s, Google expects launch prices could drop to make it actually fiscally feasible, which could make space data centers cost competitive with those on the ground.

Right. So supporters say moving compute to. space could cut local noise, water use, energy consumption, and power strain. But critics warn it could add to cloud clutter in the sky and affect astronomy. Who knows what it could affect? Because this is something that's kind of unheard of. This isn't the first time someone's talking about doing AI data centers in space. But this seems like, okay, it might be turning into a reality sooner than expected with some test flights going up maybe in like a year and a half. Right.

So to go from this being kind of a pipe dream to, oh, we could be less than two years out to Google actually sending TPUs, right? Their version of GPUs to space to run AI. It's kind of wild. You could say it's out of this world. That was a bad joke. All right. Our last big piece of AI news. Yeah, this one is big.

So it is backstop gate. Apparently I'm going to call it that. So a seemingly back and forth flip-flop on OpenAI's messaging around the intersection of the federal government and AI. So here's what happened. It's a little confusing, but I'm going to walk you through it step by step. So it started with OpenAI CFO Sarah Fryer speaking at a wall-stom. straight journal panel that she hoped the federal government would play a role in supporting AI investments. After fumbling around for a word, she used the word backstop, which some interpreted

as a request for federal guarantees. So then right after that happened and the media blew it up, maybe rightfully so, Friar walked back the wording on a LinkedIn post writing that, and I quote, Open AI is not seeking a government backstop for our infrastructure. admits, end quote, in explaining that her intent in saying that was to highlight how building U.S. industrial capacity requires both private sector investment and an appropriate government role, not specifically saying that OpenAI needs a backstop. And then on Thursday, CEO Sam Alman reinforced the clarification on Twitter, stating that Open AI neither has nor wants government guarantees for its roughly $1 trillion of data center.

plans and investments in the next few years, arguing that governments should not pick winners or losers or ask taxpayers, tax payers, to bail out private companies that make bad decisions, right? And Open AI is technically the most valuable private company in the world. And obviously, this has been going on, as I talked about earlier, kind of the big short AI play, everything else. but the comments landed as the AI-linked stocks in the market fell off sharply. Like I said, InVedia was down to 7% over the week, wiping out over $400 billion in market value. Microsoft was off 4%. Palantir, like I said, 13%.

But here's what it boils down to. A lot of people are saying, well, should there be some sort of insurance policy or guarantee from the government for AI, if AI is as important as it seems that it could be. Well, after the financial crisis, the U.S. government did introduce this very backstop for the banking system. Right. So there's a lot of interesting pros and cons, right, of what this might look like.

Obviously, yeah, there should probably be a line between the private sector and the public sector. But even with what the White House is doing right now, right? The White House has like equity stake in a private company, in a private AI company. right or sorry, not a private AI company, a public AI company. So there's obviously been some huge private public partnerships. Like I said, the government has a 10% stake in a public AI company. So I don't think it's all that crazy, right? But obviously the people against this are saying, well, this creates market distortion,

taxpayer risk. There's privacy security concerns, obviously the risk of misuse, just the regulatory feature. but a lot of people are actually for this, some sort of government backing, not for open AI, right, but for the AI industry as a whole, or some sort of more intentional investments proactively versus more of like an insurance policy or guarantee of last resort,

like what happened with the financial crisis in the banking system. And the people advocating for this will say this gives the U.S. a geopolitical edge. It can lead to faster AI breakthroughs and infrastructure, or better public services, et cetera. So it's an interesting take, but I'll leave it at that. Let me know in the, you know, in the podcast comments or live stream, if you think this is an absolutely great idea or a terrible idea. But regardless, Open AI has kind of walked back from this,

although they did put a blog post out in October that kind of hinted at this anyway. So that's why I said it's kind of been this apparent flip-flop because they kind of put a blog post out a couple of weeks ago, which we covered in the newsletter that hinted at this. Open AI CEO, Sam Altman kind of hinted at this on a recent podcast he was at. But then the CFO kind of said it out loud very directly. Everyone lost their noodles and then they kind of backtrack. So it's an interesting story to follow.

But that is it for the big AI news stories. But we're going to end with what's new and what's next. This is our quick fire, fire off round of bullet point rumors, rants, releases that maybe just make our top nine or 10 AI news stories. And there's a lot. So here we go. Let me take a, take a sip. Let's get after it. Here we go. Google released its seventh gen Ironwood TPU. OpenAI announced they reached one million business users, the fastest in history. OpenAI signed a $38 billion partnership with AWS after its updated restructure with Microsoft allowed them to shop around for cloud services.

Speaking of Microsoft, they released their new image model, MAI, Image 1 in Bing image creators. You can go use that and test it out. OpenAI released Sora on Android phones. Google rolled out AI mode on Chrome on mobile. I use that a lot this weekend. There's a recent chat GPT enterprise study that showed that 95% of users say chat GPT saves them hours a week.

Google released a job. Gemini powered Google Maps where you can talk natively, right, for driving directions in natural language. Notebook LM added sheets import from Google Drive, mobile updates, and a ton more. We're going to be covering that. All the new notebook LM updates on Wednesday for our putting AI to work on Wednesday. Snapchat and Perplexity signed a $400 million deal for Snap to use Perplexity. Speaking of Perplexity, they rolled out a performance and feature update.

for its agentic browser comment. Open AIs released a new blog post calling for more serious safety measures as AI's capabilities are growing too fast. Yeah, it's actually a pretty important blog post. If you missed in our newsletter, you should probably go read it. Open AI released GPT5 Codex mini, a new cost efficient version of GPT5 codex. Open AI rolled out Peloton and TripAdvisor apps in chat GPT. so now there's nine in total.

Google rolled out, this is a big one, Google rolled out Gmail, drive, chat, search and more in deep research inside of Gemini. Really cool. So now you can do deep research across all of those different workspace apps. There's been a ton of online rumors about a GPT 5.1 model from OpenAI.

So GPT5 might get an update any week, month, day, hour. Who knows, right, along with Gemini 3. Chad GBT announced a new feature to interrupt thinking models by inserting more context. I'm excited for that one to roll out. So if you've ever used Chad GBT's deep research or, you know, GPD5 thinking pro, you know, you might have to wait five, 10, 15 minutes and you can't interrupt it.

Well, now you can. Sam Altman announced in a tweet that OpenAI would hit $20 billion in revenue this year. Leaks suggest a new nanobanana 2. from Google is in the works. Ooh, spicy. And Amazon sued perplexity over comet AI browser shopping on Amazon, saying, hey, only humans can shop on Amazon, not agentic browsers. All right, that was a ton.

And if you spend hours every single day scratching your head at all the AI news and just getting lost and getting confused, well, join us on Mondays, right? This is what we do on Mondays, bring you the AI news that matters. So spending a little bit more time talking about the top AI news stories of the week and then running down the rumors, rants and what's next and what's new. So I hope this was helpful. And if it is, tell us someone about it, right? We're getting into the holidays, the season of giving. Give your coworker.

Give your friend. Give your neighbor. Give your mother's neighbors, coworkers, dogs, babysitters, best friends, girlfriend. Give that person free AI, unbiased education. from Everyday AI. So if you could share about this, if you're listening on the podcast, please click that follow button, right?

Especially on Spotify. We would appreciate that. Leave us a rating if you could. Tell someone about it. And then when you're done, go to Your EverydayaI.com. Sign up for the free daily newsletter. Thanks for tuning in.

Hope to see you tomorrow and Every Day for more Everyday AI. Thanks, y'all. Meet Firefly AI assistant. Now live in Adobe Firefly, the Allman One Creative AI Studio. Just describe what you want to create in your own words and the assistant handles the rest. orchestrating multi-step workflows across Adobe Creative Cloud apps,

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

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