Resources:
Join the discussion on LinkedIn: Got something to say? Let us know on LinkedIn and network with other AI leaders
Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineup
Connect with Jordan Wilson: LinkedIn Profile
Start Here Series in our Inner Circle Community: Join for free access
New AI Developments Reshaping Business Cost, Access, and Competitive Dynamics
A recent analysis from the Everyday AI podcast spotlights transformative shifts taking place across artificial intelligence, with direct implications for cost strategy, technology selection, and market positioning. For organizations making core decisions on AI investments and digital infrastructure, this week’s developments point to concrete opportunities and strategic risks.
AI Export Controls: Compliance and Service Disruption
Recent US government export restrictions have created immediate operational challenges for AI model providers and downstream users. Strict directives now prohibit access to leading Anthropic models—Fable Five and Mythos Five—for all foreign nationals, including foreign-national employees within the US. This forced Anthropic to disable public access to these systems globally to ensure compliance, highlighting the potential for government action to disrupt mission-critical AI tools without warning 03:28.
Business leaders should note the trigger behind these restrictions: reports of vulnerabilities and associated national security concerns. Enterprises relying on external AI systems must evaluate continuity plans, internal compliance processes, and vendor risk exposure, especially as government scrutiny rises for critical suppliers 04:16.
Open Source AI Models: Cost Efficiency Accelerates Mainstream Adoption
A key development centers on the narrowing performance gap between open source AI models—particularly those emerging from China—and proprietary US incumbents. DeepSeek V4, a Chinese open-source AI model, has entered consideration by major US technology leaders seeking to radically reduce compute costs for their commercial AI offerings. The difference is substantial: open-source models such as DeepSeek reportedly offer per-token rates (~$0.87 per million tokens) that are roughly 18 times less expensive than market leaders like Anthropic (~$50 per million tokens) 07:10.
The significance is twofold: open models reduce vendor lock-in and offer improved cost control, but introduce additional compliance and risk management obligations. Political scrutiny around data security and the integration of non-US models increases regulatory complexity 08:20.
AI Vendor Strategy: Microsoft Considers Open Source Alternatives for Copilot
New reporting indicates that Microsoft is preparing to shift its Copilot CoWork AI assistant to a usage-based pricing model, while also exploring the integration of lower-cost self-hosted AI models like DeepSeek V4 06:06. This strategy could mitigate unprecedented AI operating costs, highlighted by recent reports of nine-figure API bills for major enterprises using proprietary models 07:45.
The implication for buyers is clear: increased transparency and flexibility in AI licensing and usage are on the horizon, with major providers willing to bet on non-traditional partners to keep enterprise costs manageable. However, switching to open source solutions also requires internal investment in hosting, optimization, and ongoing safety engineering 08:06.
AI Performance Benchmarks: Open Models Closing the Gap
The latest model release from a Chinese startup, GLM 5.2, has drawn attention for attaining state-of-the-art results in highly specialized coding and engineering tasks. The model, equipped with 753 billion parameters and optimized for 1-million-token context windows, now challenges proprietary models including GPT-4.8 and Claude Opus 4.8 in benchmark tests 12:30.
Organizations prioritizing AI for software development or text-heavy workflows may find that open weights models—now available under permissive licensing—deliver comparable output to top-tier closed systems at a fraction of the cost. However, the text-only design of GLM 5.2 (non-multimodal) makes its application more suitable for coding and technical analysis rather than image or audio processing 13:16.
AI Pricing Models: The Shift from Token Maxing to Token Efficiency
Rapidly changing economics in AI are driving businesses to re-examine token usage and agent workflow design. A trend toward “token efficiency” has emerged, with recent reports noting that nominally low-cost AI subscription packages were subsidizing actual compute expenditure by several orders of magnitude 15:17. As provider subsidies taper, organizations must scrutinize the operational cost of their AI deployments, especially where autonomous agents or continuous workflows can balloon compute requirements 15:41.
Enterprise AI Governance: Calls for US-Led Standards at Global Summits
AI governance has taken center stage at international policy summits, with calls for the US to lead a coalition defining technical standards, risk management, and cross-border access to advanced AI and underlying chip technology. Explicit focus has been placed on withholding the export of cutting-edge AI chips and systems to key global competitors while prioritizing risk scenarios such as cybersecurity, bioterrorism, and state-level intelligence 17:03. These discussions reinforce the “geo-compliance” factor businesses must account for when choosing core AI vendors or considering market expansion.
Hardware and Health Tech: MidJourney’s Unexpected Pivot
One of the week’s most unconventional yet potentially impactful stories involves MidJourney’s transition from generative image software to the development of a full-body ultrasound scanner. This device, partnering with Butterfly Network for hardware modules, aims to offer spa-based, non-invasive, MRI-quality imagery to the consumer market—starting with a flagship location in San Francisco 21:39.
Although the initial commercial focus is outside traditional healthcare infrastructure, this approach raises the prospect for AI-enabled preventative health tools, particularly for segments willing to invest in routine, non-insurance-based diagnostics 24:06. As body composition mapping (outside FDA clearance) drives early adoption, further regulatory steps would be needed for direct clinical application.
Competitive Dynamics: SpaceX and Cursor’s 1.5-Trillion-Parameter Model
Following its acquisition by a major private enterprise player, Cursor previewed its plans to launch an in-house 1.5 trillion parameter AI model, leveraging access to over 100,000 GPUs through its parent company 27:45. Cursor’s vertical integration of proprietary model development and code-hosting infrastructure, including a Git-native AI agent platform, signals an evolving competitive threat to entrenched code repositories and AI tools.
This move positions integrated AI-native platforms as viable challengers to traditional software development ecosystems, especially in developer-heavy, engineering-centric firms where workflow automation and cost-of-ownership matter 30:09.
Anticipating the Next AI Model Rollouts
Multiple state-of-the-art models—OpenAI’s GPT-5.6, Google’s Gemini 3.5 Pro, and an anticipated re-availability of Anthropic’s latest—are all rumored for imminent release, potentially within the same week 34:42. The introduction of longer context windows (up to 1.5 million tokens), improved agent performance, and advanced technical comprehension will expand the use cases for enterprise AI deployments and raise expectations for system performance 33:44.
Building an AI Adoption Roadmap
Recent events highlight the growing complexity—and opportunity—embedded in AI strategy decisions. From open-source model selection, risk exposure due to regulatory controls, aggressive new hardware for health and wellness, to pricing models where cost transparency is becoming mandatory, these trends present actionable signals for organizations navigating the AI landscape. Being prepared to re-evaluate partners, technology stacks, and cost structures is critical as the pace of AI innovation and governance continues to accelerate.
Topics Covered in This Episode:
- ChatGPT Pulse Sunsetting and Tasks Comeback
2. ChatGPT Scheduled Tasks Features and Access Tiers
3. Claude Design June Update Overview
4. WYSIWYG Editing and Design System Imports in Claude Design
5. Claude Design Export Options and Third-Party Integrations
6. Google Vids AI Avatars Upgrade with Veo 3.1
7. OpenRouter Fusion Multi-Model Synthesis Feature
8. Claude Code Artifacts for Team and Enterprise Plans
9. GLM 5.2 from ZAI Open Weights Model Overview
10. GLM 5.2 Benchmarks and Enterprise Use Cases
11. OpenAI Codex Record and Replay Feature Explained
12. Codex Record and Replay vs. Traditional RPA Tools
Episode Transcript
Jordan Wilson [00:00:16]:
While the US government in Anthropic continued trying to make up and be friends again, the rest of the AI world didn't stop. And it was actually some smaller players that were making big splashes this week, potentially signaling more competition at the top for the big four of Anthropic, OpenAI, Google, and Microsoft. That's because open source may finally be having its chat moment as a new Chinese model might actually be closing the gap between proprietary US models and open source or open weight options. Oh, and MidJourney made some news as well. Yeah. That MidJourney that you probably haven't heard from in years. That's because they may be trying to disrupt the medical imaging space. Yeah.
Jordan Wilson [00:01:05]:
There's a lot going on if you don't follow every single day. And don't worry, I do that for you. So welcome to Everyday AI, and this is our Monday segment where we bring you only the most relevant news updates. I tell you what matters, what doesn't, and how you can use that information to grow your company and career. This is Everyday AI. So if you're new here, yeah, we do this every single day, daily. Unedited, unscripted, live stream podcast, and free daily newsletter. So, if you are new, please make sure to subscribe to the show, on Spotify or Apple Podcasts.
Jordan Wilson [00:01:40]:
I'd appreciate that. And then make sure you go to our website at youreverydayai.com. Sign up for the free daily newsletter each and every day. We recap the highlights from today's show as well as give you all the other AI news you need to know just in a quick digestible form. Alright. So without further ado, let's go over the biggest stories of the week that you might have missed, and I'm gonna break it down for you simply. Alright. Let's start at the top because, yes, at least at the time of this recording, we still don't have access to Fable five or Mythos five at whole.
Jordan Wilson [00:02:16]:
That's because the Trump White House in Anthropic are still trying to make things work. So, actually, a little bit of a controversial statement maybe, from president Trump as he told the Axios show on Friday that he considered anthropic previously and or its CEO, Dario Mahdi, to be a national security threat, though he now believes the situation is improving. Yeah. He did say that you might go wanna listen to the interview yourself. We'll be linking it in today's newsletter, but he definitely say that he considered them a few weeks ago, whether he was talking about anthropic as a whole or Dario Amadi to be a national security threat, though it's now, he says, it's improving. So The US Government, if you haven't been following along, the US government recently imposed strict export controls on anthropic barring any country outside of The US and foreign nationals within The US from accessing, Infropic's most recent models called Fable five in Mythos five. So Infropic has completely disabled in response to that. They've disabled public access, to those models.
Jordan Wilson [00:03:28]:
Fable five, which is the version of mythos five with more guardrails. So yeah. But it's been down and gone now for almost a week since that export control. So the directive bars any foreign nationals, including, foreign national anthropic employees from accessing, these frontier systems. But because anthropic can't instantly verify the exact citizens, citizenship of every single user on its platform. It was forced to yank both those models offline globally to remain compliant, and that's been about a week. So according to Axios in this new interview and story, the Trump administration's concerns were triggered by the Amazon reports detailing the vulnerability in Anthropics technology. Right? And a lot of reports said that was their CEO, Andy Jassy, and that led to the urgent discussions with the company leadership.
Jordan Wilson [00:04:24]:
Trump said he was initially alarmed by Anthropic's response, but was reassured after Amodi, acted quickly and responsibly calling him nice and smart after beating him at the g seven summit this past week. So the Pentagon has labeled anthropic a supply chain risk, a designation typically reserved for foreign adversaries, and it was the first time a US company had gotten that label, from the US government. So Trump did not rule out using emergency powers under the Defense Production Act if INTROPIC does not comply with new standards, but stated he doesn't think that step will be necessary this time. Trump revealed that it was a competitor and part owner identified later as Amazon that alerted his administration about anthropics, practice. So we went over this, in detail. FYI, if you missed the story, let me just go ahead and pull the, the episode number. That was from Thursday, episode eight zero one. So if you want more on, the background on those developments, you go listen to that episode eight zero one.
Jordan Wilson [00:05:30]:
But at least for now, mythos five still offline, but the latest here, president Trump sat down with an interview with Axios. So, yeah, if you missed that, definitely worth checking out. Alright. Our next story, and this seems to be an underlying theme this week on open source models, apparently being considered at the highest levels. So Microsoft is moving its Copilot Cowork AI assistant to a usage based pricing model. So, kind of a couple pieces of news here. Number one, that Microsoft is going usage based on Copilot Cowork instead of just including it or going, with, you know, a certain number that's included in a monthly subscription. That's number one.
Jordan Wilson [00:06:17]:
Number two is it's generally available whereas before it was in preview. And number three, the most recent news according to reports is that they're actually looking at models like China's DeepSeq v four, a version that Microsoft may fine tune, just for Copilot Cowork to bring down costs. So according to reports, the company is exploring a lower cost self hosted AI model as an alternative to the more expensive anthropic models that are currently powering Copilot Cowork by default. And Axios did report that Microsoft is considering, among others, deep, DeepSeq v four, a Chinese developed open source model, which could dramatically reduce costs, potentially charging just about 87¢ per million tokens compared to Anthropic's $50 per million tokens. So, yeah, not the best at math, but that's, like, 18 x cheaper, or something like that. So Microsoft expects to announce its final model choice and deployment details within weeks with new, with any new option to be hosted on Azure to keep customer data within Microsoft's cloud infrastructure. So the shift responds to the high compute cost that we've been hearing about over the last month or two, of agentic AI tools. Specifically, we've seen a lot of reports from enterprises, you know, getting clawed bills that are pretty high.
Jordan Wilson [00:07:45]:
Right? We saw that report, couple couple weeks ago. We don't know if it was legitimate or not, but apparently, an enterprise company had a, 9 figure, bill from Anthropic, which is a lot. Right? Hundreds of millions of dollars, because they weren't keeping, an eye on usage caps. So that's why open source models like DeepSeek v four can cut costs, but they do require companies like Microsoft to handle additional hosting, optimization, and safety engineering, shifting some burdens away from commercial API providers. So there are obviously a ton of political and regulatory concerns as DeepSeek's Chinese origins have attracted scrutiny in Washington, raising questions about data security and compliance for US companies. Microsoft has stated that any deep seek based solution would be optional and subject to additional safety and compliance filters aiming to address enterprise risk and regulatory requirements. So industry observers are watching for Microsoft's final decision, noting that total, cost of ownership will depend not just on per token prices, but also on model performance, reliability, and the need for robust monitoring and safety systems. So this one, it is kind of intriguing.
Jordan Wilson [00:09:07]:
I'll say that to say the least, with Microsoft making this move and specifically according to reports looking at DeepSeek. So why is that noteworthy? Well, Microsoft obviously has good relationships in huge ownership stakes in both anthropic and OpenAI. Right? So, traditionally, it's been the GBT models from OpenAI that have powered, Copilot. They recently started offering some Claude models as well, but the Copilot CoWork was running Claude models by default. So CoWork is kind of the original technology from Anthropic that Microsoft essentially have their own version that was powered under the hood by Anthropic's Copilot technology. So what's interesting here with Microsoft looking at model providers like DeepSeek is, well, OpenAI and Anthropic have launched some serious concerns in, regulatory, steps that they've taken, you know, going to the US government, accusing DeepSeek and other Chinese companies of distilling their models. So that's what's gonna make this really juicy. Pretty, if I'm being honest, a pretty bold, step here from Microsoft even though it is just according to reports, to do this considering, two of their biggest and most important partners in OpenAI and in Propic probably don't have the best view of DeepSeek for those exact reasons because they've shown, right, Anthropic came out with a little bit more of a report how DeepSeek and other Chinese companies are essentially taking all of their hard work to make these open source models.
Jordan Wilson [00:10:46]:
Alright. So our next AI story staying in the same vein. Yeah. We have a new king of the open hill. That's Chinese startup, z AI, is making a ton of headlines this week with their new open weights model, g l m five two. So z a I has released that model, g l m five two, a massive 753,000,000,000 parameter language model, now available with unrestricted MIT open source licensing. So GLM five two is designed for long horizon autonomous coding and engineering tasks featuring a highly stable 1,000,000 token context window, allowing it to handle extremely long documents and complex workflows. So the model is immediately accessible for download on Hugging Face.
Jordan Wilson [00:11:38]:
But, yeah, you kinda have to be an enterprise with a little bit of compute to be able to do that, because it is a large model unless you're gonna run a highly quantized version of it. But you can also use it through Hugging Face, through the z a I API and in over twenty third party coding environments. So, the new model, which is pretty interesting here, it does look like it's very bench maxed FYI. Right, that's when a company maybe puts a little too much priority on getting certain benchmarks. But still, it outperforms most other open source competitors, and it is even now challenging the top proprietary models on several industry benchmarks, including SuiteBranch Pro, which I think is thankfully on its way out of being an important model for software engineering. Yet still, it is batting on some of these benchmarks, including SweetBench Pro, Frontier Suite, and others around g p t five five and Anthropics Claude Opus 4.8. So with Fable five off the shelf. Right, I mean, o GPT five five and, Opus four point eight are the undisputed leaders.
Jordan Wilson [00:12:52]:
So, pretty significant even if it is maybe a little bit oversaturated and a little bit bench maxed, that in open source model, I I I won't say they've closed the gap, but they're starting to close the gap on, Frontier models. So there's obviously some use cases that you would absolutely not want to use, GLM five two mainly because it's not multimodal. So it's text only. So that's a big downside. Right? So for someone like me, I'm constantly, doing multimodal inputs, you know, whether that's in Google's AI studio, whether it's in, you know, and profit Claude, codecs, chat GPT, etcetera. Right? So you can't have multimodal inputs or multimodal output. But with text only, you know, GLM five two does seem to be a competitor, and it is technically state of the art even beating, in a couple benchmarks. I kid you not.
Jordan Wilson [00:13:45]:
Fable five, Opus four eight, and GPT five five on some more niche, kind of front end coding design, some certain benchmarks out there. So, pretty surprising, considering that even the best open source models prior to this were generally, I would say, about four to eight months behind the frontier counterparts. So still, GLM five two is far behind mainly in terms of capabilities because it's not multimodal. But for text only, you know, at this point, it is only, like, two ish, two to three months behind, which is fairly impressive. And I think it does change the narrative, especially when, you know, kind of all of the rage in the second quarter and the summer so far has been shifting to token efficiency. I do think most companies, the biggest decision that they're gonna make in 2026, similarly similarly to what Microsoft is doing, you know, looking at switching over from some anthropic models, to DeepSeek is, well, most companies are gonna have to start looking at their, their, whether you're talking about token budgets, your AI spend, but, you know, switching over, from token maxing to token efficiency, which we did a story on that in our start here series. So make sure you go check that out. But essentially, right, some of these agents run-in loops nonstop, and as companies have kind of started to, pluck away the, you know, them kind of subsidizing.
Jordan Wilson [00:15:15]:
Right? Because you see different reports. You know, a $200 a month plan, if you were paying API cost, might cost actually, if you're paying for those tokens anywhere from, you know, 10 to 15 to $20,000, you know, for $200. So, you know, essentially, these companies have been eating, those costs just to get users in and happy and sticky and to, you know, get it rolled out into the organization. So a lot of companies are gonna be have to be looking at token efficiency. So, pretty big news both with the last two stories. Microsoft looking at open alternatives to power its Copilot co work as they switch to usage base and with the new z a I five, GLM five two model. Definitely one worth keeping an eye on. Alright.
Jordan Wilson [00:15:59]:
Our next AI news story, the CEOs of Anthropic and Google DeepMind, and OpenAI met with g seven leaders and top US officials at the g seven summit in France. So the main news out of here, at least when it comes to an artificial intelligence standpoint, was a push for a US led coalition on artificial intelligence. So, Dario Amadi and Google DeepMind CEO, Demas, Demas Hassabis urged the creation of a US led international coalition to set rules and standards for AI, setting urgent risks and the need for global cooperation. So president Donald Trump, US secretary of state Marco Rubio, treasury secretary Scott Bessent, and commerce secretary Howard Lutnick represented The US at the meeting highlighting the high level attention on AI governance. We also had Canadian prime minister, Mark Carney, who expressed support for US leadership informing an AI standards coalition signaling growing international alignment on the issue. So the call for action comes after the US government imposed export controls on Anthropic's latest AI models, Fable five and Mythos five, which led the company to disable access due to the national security concerns that the government labeled. So Omodi recommended that international cooperation should cover structured access to advanced AI models and trade of chips and critical components, specifically excluding China from these exchanges. So the tech leaders emphasize the urgent need to address AI risks in cybersecurity, bioterrorism, and intelligence, warning that unchecked AI could lead to major disasters if misused.
Jordan Wilson [00:17:44]:
OpenAI CEO Sam Altman advocated for a global forum to set standards for testing, risk analysis, and international collaboration, while OpenAI's Chris Lehan noted broad recognition that The US should take the lead. So the big development out of g seven summit. Well, number one is that AI was front and center. Right. So maybe it was just timing circumstantial timing, with the mythos, pull down that this happened in the middle of it. So, you you know, especially when it comes to the rise of Chinese open source models. Obviously, from an adversarial perspective, the g seven, collectively wants to keep an eye on what China is doing in AI, how they're using this technology because like I've been saying for years, ultimately, this will play out one way or another on the battlefield. Right? This is, I think, why, the g seven and world leaders, are starting to pay more and more attention to AI, not just because of how it can be used, in cyber warfare, but also on the actual battlefield, which is, actually goes back to the initial, kind of feud between Anthropic and the US government for how Anthropic did not want the military to use its models.
Jordan Wilson [00:19:02]:
So, the the other big takeaway here is obviously seemingly some international support, for The US to take a lead in this g seven AI coalition, which makes sense. Right? A lot of times, other world leaders, you know, might not want to, you know, step aside and let The US take a lead on something like this. But, obviously, with the, the biggest and most powerful systems in the world residing in The US, it does make a lot of sense. Alright. Here's one that maybe didn't make as much sense. That's MidJourney. They've got an interesting pivot as I take a sip here of my coffee. AI moves too fast to follow, but you're expected to keep up.
Jordan Wilson [00:19:53]:
Otherwise, your career or company might lag behind while AI native competitors leap ahead. But you don't have ten hours a day to understand it all. That's what I do for you. But after 700 plus episodes of Everyday AI, the most common questions I get is, where do I start? That's why we created the Start Here series, an ongoing podcast series of more than a dozen episodes you can listen to in order. It covers the AI basics for beginners and sharpens the skills of AI champions pushing their companies forward. In the ongoing series, we explain complex trends in simple language that you can turn into action. There's three ways to jump in. Number one, go scroll back to the first one in episode six ninety one.
Jordan Wilson [00:20:37]:
Number two, tap the link in your show notes at any time for the start here series, or you can just go to starthereseries.com, which also gives you free access to our inner circle community where you can connect with other business leaders doing the same. The start here series will slow down the pace of AI so you can get ahead. Love the headline here. I I believe this was from The Verge. They said Midjourney goes from generating cat images to full body ultrasound scans. Yeah. That's where we're at. So if you missed it, pretty, I mean, this could either end up being one of the biggest pivots or one of just the most confusing.
Jordan Wilson [00:21:22]:
But Midjourney CEO David Holes has revealed the company's first hardware product, marking, obviously, a very significant shift from its well AI well known AI image generator. So here's what they announced at their event late last week. So they announced the mid journey scanner, which is a full body ultrasound device designed to capture detailed internal images of muscle, fat, bone, and organs using a ring of sensors. So Holst claims the scanner aims for image quality comparable to an actual MRI, though the scans should only take about sixty seconds according to the company. So this pivot, well, you might be like being like, okay. How did they literally go from, you know, one of the first AI image models, you know, generating silly images, you know, especially early on that didn't look very good. Right? But Midjourney, I think was, you you know, the big player early on before Dolly and all these other players. But before Midjourney, their CEO, David Holes, actually had a relevant background.
Jordan Wilson [00:22:31]:
So he spent a decade as the CTO of Leap Motion, which is a company entirely focused on advanced hardware, optical spatial computing, and tracking human not, anatomy. And the other reason why this pivot maybe makes sense is, well, because mid journey doesn't have any investors. So they can do what they want. So presumably, aside from there's a lot of lawsuits out there that are still pending, so we'll see where they, land or settle in those. But presumably, Midjourney has a nice stack of cash, and they don't have, you know, let alone companies like, you know, SpaceX and Cerebras that just went public. Obviously, Anthropic and OpenAI will be going public, fairly soon. So that changes what you can and can't do and even in the time leading up. But Midjourney, as a company that's essentially bootstrapped in, at at a time at least, they were printing money, they can make these kind of pivots.
Jordan Wilson [00:23:26]:
And with Hulse's background, it maybe makes a lot more sense than people, initially thought when they just saw the headlines. So let's talk a little bit more about the device. So it uses 40 butterfly ultrasound on chip imaging modules and two petaflops of processing power, developed in network, in partnership with the Butterfly Network. So the scanning process involves the subject stepping into a shallow pool of light and descending into water, where thousands of underwater sensors use ultrasonic waves to create three d images of the person that is in the machine. So here's the other interesting part. Well, you're not gonna well, maybe you will, but at least that's not the plan to find these things. I mean, a doctor's office. That's because MidJourney First plans to open a spa in San Francisco's Union Square by the end of next year, featuring 10 scanners, a gym, saunas, cold plunges, and hot tub equipped scanning rooms.
Jordan Wilson [00:24:29]:
So yeah. Not something that they're, you know, gonna be rolling out to, you know, AI equipped doctor offices. It seems like they're gonna be opening spas with these scanners. So the company is currently focused on body composition maps, which do not require FDA clearance. That's the key there. But future medical applications would need regulatory approval. So users will be able to share their scan library with doctors and AI health tools, and MidJourney promised to prioritize data privacy as the launch approaches. So Holst envisions the scanner becoming a faster, safer alternative to an MRI without radiation or magnets in hopes for future FDA device classifications to enable broader data collection.
Jordan Wilson [00:25:18]:
So, yeah, a strange one here, but the more you think about it, it kind of makes sense. Right? MRIs are expensive. They're not easy to get into. So we'll see, how this is initially received, and we'll see I think the telling thing will be how this is received outside of San Francisco. Because let's be honest, there's a decent chance no one knows how much this is gonna cost, but this just seems like it's made for tech pros. Right? Let's be honest. Let's it it it it sounds like it's made for people with, you know, who, put, you know, use stacks of cash to, you know, as a booster seat, essentially. So I think people who have, you know, millions of dollars of disposable income who live in Silicon Valley might see something like this and say, oh, absolutely.
Jordan Wilson [00:26:10]:
Outside of San Francisco, I mean, we'll see how this catches on. Right? I'm from Chicago. I don't necessarily see this catching on. Again, it depends on price. Yes. There's all these other things. The, the spa, the plunge pool sounds great and all. But I don't know how many people would be interested in, you know, in in their kind of thought on this is the more that you scan, the more scans you get, the more likely you are to pick something up and detect a potential, disease or abnormality sooner, which makes total sense.
Jordan Wilson [00:26:42]:
But I just don't know the amount of people who can afford something like that or if there's an appetite outside of tight tech circles for people to be going, I don't know, monthly or even quarterly, to get these scans. But regardless, hey. You gotta swing for the fences to hit a home run. You're not gonna put runs on the board by bunting every single time at bat. Alright. Speaking of swinging for the fences, this one did not get a lot of coverage, but I decided to make it a main story because I think it's really important. So cursor announced a new well, they didn't actually announce it, but they kind of previewed it that they're working on their own from scratch 1,500,000,000,000 parameter model, and they are releasing a GitHub competitor. So two big pieces from, cursor.
Jordan Wilson [00:27:34]:
So days after officially being acquired by the now public SpaceX, cursor dropped some major announcements at their compile conference. So those are that they're launching a new made from scratch AI model with over 1,500,000,000,000 parameters trained on scratch on more than a 100,000 GPUs, making it potentially one of the largest and most powerful models to date. So that's what happens when you, are now owed by SpaceX and you have access, to all these, you you know, compute factories, Colossus one and two, that, Grok slash x a I slash SpaceX. Right? They're now just one thing they've invested heavily in. So, Cursor says this new model is designed to be a generally intelligent assistant moving beyond code generation to handle complex engineering tests, like planning, testing, and interacting with user interfaces. So a couple reasons why I think just the model is important. Well, number one, their last model that they up, that they came out with that I believe was based on, Kimmy. So it was just a fine tune or built on top of an open source model.
Jordan Wilson [00:28:50]:
So composer 2.5 was actually a really good model. It was the first time that cursor, I think, had a model that a lot of people were starting to consider. And as we start to shift into the super app phase alright. So make sure you go listen to episode seven ninety nine, where we talked about AI super apps, and I did mention cursor is a player, now. You like, you have to look at the cost efficiency. So, you know, essentially, with composer 2.5, you got about maybe 90% of the power that you would get from something like Opus, four eight at, like, 10% of the cost. So with cursor going from scratch and now having that relationship with XAI, Grok, SpaceX, I do think we have to start paying a little bit of attention, a little bit more attention, to Cursor, especially because I do think on the harness side, it is one a and one b in terms of I do think still, codex on the harness side one a and cursor one b. They have a great, harness in the, in the cursor platform.
Jordan Wilson [00:29:58]:
So, cursor's CEO, Michael Truel, emphasized the scale and fresh training approach allows for better control over model behavior and supports a wider range of engineering workloads. The other big news is they did announce their origin feature, a new Git native code hosting platform built for AI agents, which aims to challenge Microsoft's GitHub by supporting thousands of automated code operations daily in featuring automated conflict res resolution. So Cursor's acquisition gives it access to SpaceX's vast computational resources with analysts saying the combination of Cursor's reinforcement learning expertise and SpaceX's compute power could quickly make Cursor a top competitor to Claude Coe, codex, and even GitHub. So the GitHub play is interesting because we saw reports about a year and a half ago that OpenAI was developing their own, GitHub, kind of, competitor, but it never came to fruition. So maybe that was one of those, you know, side quests that was tabled. So cursor maybe could start putting itself into the conversation. We'll see if we start shifting from talking about the big four of anthropic Google, OpenAI, and Microsoft, and maybe cursor might thrust itself up into that first tier. Alright.
Jordan Wilson [00:31:22]:
Our last big piece of AI news. Yeah. It seems GPT five six from OpenAI could be days away. So according to reports, leaks, sleuths online, all that stuff, OpenAI is preparing to release its new GPT 5.6 family of AI models as soon as next week. So according to reports and leaks, you might see a couple different variations of the g p d 5.6 model, including the standard, a new mini, and pro versions that could debut together. Early access to g p d five pro has reportedly already reached some pro subscribers who report noticeably better understanding on technical prompts, though certain web development issues remain unresolved. So yeah. If you've been on, you know, Twitter at all over the last, like, five days, you'll see a lot of people.
Jordan Wilson [00:32:24]:
And I need to do a better job of, you know, coming up with these, use cases that I actually save. The thing is, I try to not take on as many of the visual use cases, because I think there's plenty of people out there. And I don't think that for the most part, these are things aside from, you you know, creating things like decks, PowerPoints, you you know, word documents, you you know, business dashboards. I think a lot of those, are easy to compare. But I think, you know, what a lot of people are doing is just like, you you know, three g s, you know, three d worlds, you you know, coding games. I don't think for the majority of our audience that's, something that people are excited about. But that's why I think we've seen all these reports that apparently, OpenAI may be testing, a b testing, GBD five six, because people have saved, oh, here's what GBD five five pro did when I said, you know, create an interactive world model that you can walk through, you know, just using three g s or something like that. And people have those kind of saved and all of a sudden it's, wow, like, three, four times better.
Jordan Wilson [00:33:27]:
So a couple more details. The new g p d five six is rumored to push the context window to 1,500,000 tokens up from 1,000,000 tokens in g b t five five, allowing for significantly longer and more complex conversations and coding sessions. So developers who think or believe that they have access to g b d five six pro say that it outperforms anthropics mythos models in agentic coding tasks, raising the stakes in the competition for the AI powered software development. So, make sure to keep an eye on the newsletter, this week. And, obviously, if it does come out, you better believe we will be having a show because there's a good chance that, maybe as soon as today, maybe by the time that you, you know, hear this podcast that we might have Fable five from anthropic back. Again, according to reports, we could have GBD five six as soon as this week. And we did earlier, from Google themselves at their IO conference last month. They did say that next month, which would be this month in June, they did say that they're going to be releasing their Gemini three five pro model.
Jordan Wilson [00:34:42]:
So in theory, this week, we could get access or get access again to three state of the art models. So, pretty big week. Buckle in. Make sure you got your tokens ready or your API, budgets in check because there could be a lot of new model access rolling down the pipeline. Alright. So that is it for our main AI news stories. So let's quickly go over everything else. This is the what's new and what's next.
Jordan Wilson [00:35:11]:
Just stories that didn't make our top cut. Some rumors, some leaks, but let's go ahead and go through them quickly. Bullet point style. Alright. Here we go. So Amazon, Nvidia, and AMD invested $310,000,000 into Odysee ML to advance real world AI understanding. Anthropic upgraded claw design, adding custom design systems, code sync, and export features. That one, really good.
Jordan Wilson [00:35:41]:
FYI, OpenAI is reportedly preparing to integrate the library feature from chat g b t into codecs. I would like that. I spent a lot of time trying to find all those files. XAI released their Grok Imagine video 1.5 model. It looks pretty good. Adobe expanded their Firefly AI assistant to other, tools in their stack including Premiere, Illustrator, InDesign, and frame.io with some new features. OpenAI released unified usage analytics and spend controls for chat GBT enterprise. Some new leaks show that Anthropic is working on a schedules feature for its always on upcoming Claude Conway agent.
Jordan Wilson [00:36:24]:
Google rolled out their ask ad manager beta, an AI agent for custom insights and troubleshooting for Google ads. Like I said, G b d five six and Gemini 3.5 are likely around the corner. Codecs, this is the big one, released their record and replay workflow capture for Mac, so excluding some countries. And I do believe I'll have to check again at the, the newsletter. Most of the times either on Friday or Monday, I'll ask you guys what you want for our Wednesday show, which is a hands on demo. And I do believe that the new codecs, record and replay, one, which I was surprised. I thought Claude Design was gonna win our poll, but it seems like everyone wanted the new, record and replay feature, which is really cool. Google Gemini co lead, Noam Shazir, left for OpenAI after Google acquired or aqua hired, him and his group from character AI.
Jordan Wilson [00:37:20]:
This is a pretty big one. Right? So one of the original authors on the attention is all you need, paper that led to the transformer. Right? So one of the OGs of AI leaving, Google Gemini after they spent $2,700,000,000 to join OpenAI. Another big get for OpenAI this week, former White House AI adviser, Dean Ball, joined OpenAI to lead Frontier AI policy team. So some big hires for OpenAI after Anthropic landed, Andre Karpathy a couple of weeks ago. Speaking of Anthropic, they released their Claude code artifacts, which allows teams to share live coding artifacts for collaborative and interactive sessions. So very similar to codex sites. X a I oh, already covered that one.
Jordan Wilson [00:38:09]:
Copilot co work is generally available. Like I said, downside though is usage based pricing. And last but not least, Google released the open knowledge format. So kind of an alternative or what could in theory be an alternative, for in, a Jentic communication system. So instead of markdown files, you might have okay f's if a lot of people, adapt this. So that's it. That is a wrap for the AI news that matters. So y'all, I'm telling you, this is happening so fast and the pace is only going to pick up.
Jordan Wilson [00:38:41]:
Alright. So you probably shouldn't be spending, like, four or five hours every single day. Right? Taking all this work home for you. That's what I do. So make sure I try to keep these podcasts thirty ish minutes, right, even faster if you listen on two x. So make sure you subscribe to the podcast, but you really need to be reading our newsletter. So make sure you go do that at youreverydayai.com. So thanks for tuning in.
Jordan Wilson [00:39:04]:
And FYI, if you are new here, Mondays, we do our AI news that matters. Wednesdays, we go hands on with a kind of live demo with our AI at work on Wednesdays. Friday, we go over Friday features, which are new features that you can actually go use today, and I kind of give you the one zero one. And then Tuesdays and Thursdays, we rotate what those shows are. So if you knew if you're new here, that's kind of our weekly lineup. So thanks for tuning in. Hope to see you back tomorrow and everyday for more everyday AI. Thanks, y'all.