The Most Important AI News This Week Is Getting Overlooked

Model releases and show overview

0:00We got a lot of notable model releases this week. Muse jumped back to the frontier with Wonder and Wonderful. Google actually released Gemini 4 Argon. And yet it seems like the two little unassuming models that were announced on Thursday that are the subject of today's show, the ones that are getting the least fanfare of all of these, might actually be the most significant, at least when it comes to the practical reality of using AI inside businesses. The AI Daily Brief is a daily podcast and video about the most important news and discussions in AI.

0:30All right, friends, quick announcements before we dive in. First of all, thank you to today's sponsors, KPMG, Harbor, Granola, and Blitzy. To get an ad-free version of the show, go to patreon.com slash AI Daily Brief, or you can subscribe on Apple Podcasts. And to learn more about sponsoring the show, send us a note at sponsors at aidailybrief.ai. One more shout out, we have just announced the latest super intelligent premium training program. We are soon kicking off our first AI Champions Bootcamp. This builds not only your skills to use AI,

1:03but to support adoption across your enterprise. It's anchored by six live sessions and has a ton of value built in. And you can find a link to it from the top of the AI Daily Brief website.

Anthropic embeds safety evaluators

1:14Anthropic has begun the process of embedding third-party safety evaluators into their organization. Last month, as part of calls to pace the frontier, Dario Amadei pledged to integrate third-party evaluators into Anthropic, giving them staff-level access to monitor safety procedures and reports on incidents. In an update published on Thursday, Anthropic said that they've now integrated the first batch, which is of course a six-person team from Meter. They said the Meter folks have employee-like access, which is their term, including use of company computers, office space, and use of internal tooling. Anthropic added that they're working with SecureBio and Accenture as

1:48evaluators and are in early discussions with other organizations, including Hugging Face, explaining why they support this model of safety monitoring, Anthropic wrote, We want to be publicly accountable for the decisions we're making, but we can't publicly share all the information relevant to our safety work. Some of it could be useful to malicious actors or is commercially sensitive like computer training information. Instead, we're giving third-party evaluators deep access to our information and are having them share evidence and reasoned conclusions about the risks involved in building our models. We aim to close the gap between what

2:18evaluators understand about a frontier AI company's systems and what the company itself understands. Now, as this safety regime gets rolled out, there are two big areas of discussion and frankly criticism that are worth understanding. The first is the independence of the outside safety evaluators. Unlike what some have said, Anthropic claims they do not fund Meter. And frankly, the web of connections is a bit beyond the scope of this particular headlines to unpack. However, Meter does share major backers with Anthropic aligned around the effective altruism space. It's not necessarily

2:49full tinfoil hat, it's just that they come from the same philosophical roots and end up having a lot of the same funders because of that. Now, there are a range of valid opinions that could be held about the arrangement, but it's context for why Anthropic is quickly moving forward with that particular partnership, while OpenAI appears a little more hesitant. The other big issue is how much control the labs have over the publication of safety audits. Earlier this week, Reuters published a report claiming that both Anthropic and OpenAI were seeking strict editorial control over reports. This seemed to be a fluid negotiation at the time of reporting, and the parties seemed to be moving

3:20toward a compromise that involved redacting confidential information and noting when major redactions had occurred. With Meter, Anthropic wrote that they, quote, can't veto publication, but we can require them to redact information if it falls into narrow predefined categories. Meter is free to characterize our redactions publicly and say whether it believes the redactions were justified as long as it does so without revealing the redacted information. The categories mentioned were data about customers, legally privileged information, and information about model architectures. Anthropic has also defined three tiers for

3:51embedded auditors, with Meter the only one in the top tier, defined by employee-level access. SecureBio is currently in the bottom tier as they begin initial work reviewing existing chemical and biological weapons risk already publicly disclosed in Anthropic's August risk report. In that tier, Anthropic requires auditors to seek approval before publishing. Anthropic said that they will evaluate monitoring SecureBio to the second tier, which would get them more access to information and no approval required following the completion of their first report. Accenture has not begun work, but Anthropic says they are aiming to develop a partnership with them that involves red teaming

4:22models, guardrail testing, and alignment assessments. Most of the chatter coming out of this is about the independence of Meter, with embroidery founder Zach Korman summing up a big strand of thought when he writes, for all the AI safety people pushing for independent evaluators like Meter and AISI, there's already a term for this. Security theater. And yet I think for a lot of folks, it's less about where it starts and more about where it ends. In other words, if Meter is the only evaluator, it's kind of like JP Morgan picking a handful of bank evaluators that it likes and saying that's a good enough independent audit. It seems like that's unlikely to ultimately be the

4:56scenario. Even some from the Meter team like Parv Mahajan reminded, I'm pretty excited about this as a prototype for risk assessments, but it's important to remember that this is not a substitute for formal oversight. Meter is an assessor working under a voluntary agreement, not an auditor, because there are no set safety standards to audit. I think ultimately this is going to have to end up in some sort of much more external auditor. In fact, this is probably a line of business for the big four. Right now, of course, the constraint is that there just aren't that many people qualified to actually do this sort of assessment, and pretty much none of them work for those truly independent organizations.

5:29Now, in terms of Frontier Lab alignment around this as a move, many noted that Elon Musk responded to Anthropics' post about the move on X, saying SpaceX will probably do the same.

Cognition token management system

5:40Next up, Cognition is launching a new system to help organizations token max more efficiently. So far this year, we've broadly seen two settings for token spend. There was the feast of the token maxing era in Q2, with unlimited token budgets and a mandate to use as much as possible. Then there was the famine of the token austerity era, when organizations started placing caps on monthly usage. Realistically, most organizations were still pretty generous with their token caps, but it sent the message that token budgeting was the way of the future. Calvin Giroud, Cognition's VP of Customer Engineering, believes both approaches are wrong-headed,

6:13commenting, The limits that a user should have should not be static. They should be dynamic based on the efficiency of your usage and the value that you're getting out of it. Cognition has come up with an interesting way to dynamically adjust budgets for engineers using their coding agent, Devin. Cognition says that they can use AI to examine a session and estimate the equivalent number of hours it would have taken a human to achieve the same amount of work. When engineers hit their limits, the system scores how efficiently they're using their token budget. If they have a satisfactory grade, the limits automatically increase. Giroud said,

6:43This is how credit cards work, for example. You have a set credit limit, and as you prove you are a good steward of the credit card, and you're paying back things on time, you get higher and higher and higher limits. Cognition believes that in the hyper-competitive landscape of agentic coding, giving customers efficiency tools to build trust is going to have a better payoff than encouraging maximum spend. Giroud commented that unrestrained token maxing is, quote, not a sustainable long-term approach, and many of our customers are large public companies that have to answer to their shareholders in public markets around the efficiency of the tools they're using. Cognition is also rolling out a new feature called Dev and Coach, an AI system to give feedback

7:16on how engineers are using their agent. The Coach is designed to pick up a few dozen of the most common errors like making ambiguous requests or selecting an expensive model for a mundane task. The main response from developers was, screw this. Software engineer Jesse wrote, Software engineering is now a factory role. While Amisha wrote, There's something very Soviet about this. Brian Chong wrote, How many hours would a human have needed to build this is measuring the wrong thing. It's incentivizing folks to build the most complex and bewildering thing possible, not the most valuable thing. And yet others got that even if this particular methodology is wrong,

7:51this type of system is just going to be a feature of the world that we live in now. Wrote Weave founder Adam Cohen, Token allocation will be the most important skill in the next

White House AI task force

7:59decade. Over in the government realm, the first meeting of the White House's AI task force was held on Thursday, setting the agenda for AI oversight within the administration. The primary task is overseeing an AI safety report to be published within 120 days, according to the president's executive order. The president's Council of Advisors on Science and Technology has been commissioned to prepare the report, which will cover AI risk in cybersecurity and critical infrastructure, biosecurity and dual-purpose applications, as well as model control and accountability. The task force also agreed to set up a working group to implement the White House

8:31Accord signed by each of the Frontier Labs and NVIDIA. The Accord commits the labs to implement safety controls with board-level oversight, as well as partnering with external auditors. Defense Undersecretary Emile Michael will lead an initiative to establish reporting protocols to enable the labs to inform the government when security incidents occur. FTC Chair Andrew Ferguson, meanwhile, is tasked with creating a Law Enforcement and Consumer Protection Coordination Committee across the DOJ and FTC. Finally, Office of Personnel Management Director Scott Kapoor will work with the public and private sector on the hardening of critical systems, as well as coordinating work at the Department of Labor to study the impact of AI on

9:05the American workforce and economy. One thing that's still up in the air is how the government will enforce compliance, given that the White House Accord is only voluntary or, as the president put it, morally binding. A White House official told Semaphore that one stick could be the threat of export controls, which was used against Anthropic earlier this year. Meanwhile, everyone on X right

Barclays Tesla SpaceX merger note

9:24now is talking about the new Tesla Roadster, which basically kind of flies thanks to some SpaceX thrusters. But that's not the only big Tesla-SpaceX collaboration this week. A research note from Barclays claims that a long-awaited merger could be in the cards before the end of the year. Analyst Dan Levy wrote in a Wednesday note that a merger could be announced ahead of Tesla's annual shareholder meeting on December 9th. Neither company has announced anything and nothing has been filed with the SEC. Still, Levy wrote, while the industrial logic of a combination had been in question, we believe recent managerial comments indicate they see

9:56key benefits of a combination. Levy said that neither company has hired bankers or created a board committee to begin talks. However, he believes the joint TerraFab chip plant would be a key driver. He suggested the merger could be structured as an all-stock deal, swapping Tesla shares for SpaceX stock at around 0.7 to 1. Barclays suggests that this would leave Musk with 37% ownership of the combined company and roughly 56% voting power. So far, the prediction markets don't believe it, with odds of a merger before the end of the year still at 24%. And yet, despite the lack of obvious preparation, this is another deal between Elon and Elon, and Elon is already in. During an interview

10:30last month, he said, Imagine what action one might take when there's so much close collaboration.

Vibe coding Kong Royale game

10:34Lastly today, a really cool follow-up to our story yesterday about how this latest generation of models is starting to turn out to look more like Opus 4.5 in terms of their impact and expanding capabilities. YouTuber Max Blade and his streamer buddies have vibe-coded a Battle Royale game called Kong Royale. Blade called it the world's first AAA vibe-coded Battle Royale, and looking at the screenshots, it seems pretty good. The game has a huge island setting, vehicles, and a giant gorilla as the player character. Now, Blade worked on this idea during the early days of vibe-coding, but said

11:05the models just weren't ready for it. But, armed with Opus 5.5 and a couple of Claude Max accounts, he was able to put it together with a few weeks' work. Blade calls the methodology that they used hive-coding to do the work, which is basically throwing a bunch of agents into his daily livestream chat with human developers who tune in. Overnight, the agents would review the chat, select the best changes, and ship them for a morning build. Blade essentially crowdsourced the development of the game using ideas from his audience. We're now about a year and a half out from Peter Level's Plane Game, which kicked off the idea of vibe-coding full games. It is pretty amazing what a solo dev can

11:38now achieve with a bit of perseverance and an AI subscription. Riley Brown thinks that this social creation is the obvious next step. He wrote, This is obviously the future of video games, streamers and their community building the game together, then playing it. Every month that goes by, as models improve, these games will look more and more professional. Alex Hansen wrote, This is the new era of Bros Nights. Instead of cussing at kids half your age on Fortnite, go get Mountain Dew, pizza, and a whiteboard. You map out your own game or think of a refresh or mod of a game from your teenage years. Give it Opus 5-5 to loop on it every night and a Claude

12:10Max-sub you all split. Bros Nights are so back. That, however, is going to do it for the headlines. Next up, the main episode.

12:22A new study from KPMG in the University of Texas at Austin found that when people work with AI, similar skills don't guarantee similar outcomes. Researchers studied more than 500 early career professionals and found that the best performers consistently amplified the value of AI by guiding, evaluating, and refining its outputs. These top performers, called AI amplifiers, weren't defined by what they knew alone, but by how they worked with AI. Learn more about what separates AI amplifiers from everyone else at kpmg.com slash US slash AI

12:55amplifiers. Every episode, I talk about the competition between OpenAI, Anthropic, SpaceX AI, Google, and Meta. And if you've been listening for a while, you might have a favorite. Maybe you think OpenAI and Anthropic can stay ahead, or perhaps Meta's open source strategy can win out. Whatever your view, every AI lab creates a different investment opportunity. Harbor Capital Advisors' AI Lab Ecosystem ETF Suite lets you invest in the ecosystem behind the AI lab you believe in. Search Harbor AI Lab Ecosystem ETFs wherever you invest or follow at Harbor Capital on X to learn

13:27more. Visit harborcapital.com for a prospectus containing investment objectives, risks, fees, expenses, and other important information. Read and consider it carefully before investing. Risks include principal loss and artificial intelligence-related risks. Harbor ETFs are distributed by Foresight Fund Services, LLC. Harbor is not affiliated with AI Daily Brief, and the funds are not affiliated with, sponsored by, or endorsed by any AI lab. This is a paid advertisement and not personalized investment advice. Investing involves risk, including possible loss of principle. When I'm in a meeting, I'm fully in it. I'm thinking about the iteration and creative back and

13:57forth it takes to actually push a goal forward. What I'm not thinking about is capturing takeaways, tracking to-dos, or any of that. And that's where Granola comes in. Granola is an AI-powered notepad that captures what happens in your meetings and turns it into clean, structured notes, with the decisions and action items pulled out and easy to find. There's no setup and no configuration. It just fits into how you already work. For me, it means I get to stay in idea mode, and Granola makes sure those ideas actually become action. Once you try Granola on a first meeting, it is hard to go without. You can try it totally free at granola.ai slash brief. That's granola.ai

14:32slash brief to get your time back. Blitzy's understanding of massive code bases unlocks autonomous security fixes, modernization, and new features. So what happens when there's no legacy code at all? Greenfield is supposed to be the easy part. Clean slate, no technical debt. But even Greenfield moves at human speed one sprint at a time. Blitzy changes the unit of work from the developer to the project, autonomously planning, building, testing, and validating entire applications from scratch. Hundreds of thousands of lines of production-ready code. One Blitzy customer stood

15:02up a brand new application, 534,000 lines of code, compressing a 65-week roadmap into two weeks. Another shipped an entire application with no front-end engineer. Legacy or Greenfield, the answer is the same. Software at the speed of compute. Build what's next at Blitzy.com. That's B-L-I-T-Z-Y dot com.

Microsoft base model MAI Base 2

15:25Welcome back to the AI Daily Brief. Well, would you believe it, the most important AI model announcements of this week are the ones that people are talking the least about. On Thursday, we got model announcements from Meta and Microsoft. Nowhere among them are you going to see benchmarks that change the state of the art. Artificial analysis isn't going to dedicate pages and pages to these. And yet, for reasons that have nothing to do with the excited hype discourse on X and everything to do with how business usage of AI is changing, Microsoft and Meta's latest model announcements deserve a lot more attention. Let's talk about Microsoft first. On Thursday,

16:00Microsoft's AI CEO Mustafa Suleiman announced a new state-of-the-art base model called MAI Base 2. It's a 2.8 trillion parameter model, making it the same size as Kimi K3. It's likely a little larger than Sol or Opus, but not quite as large as Fable or Astra, although we don't have good information on how large any of those models are. It uses a mixture of experts architecture with 80 billion active parameters, which is again fairly similar to the architecture of Kimi K3. This is also a completely fresh pre-training run, with zero distillation from other models. Now, this is just

16:34a base model, meaning that it hasn't gone through reinforcement learning. But Microsoft is pretty excited about what they've seen so far. During a town hall prior to the announcement, Suleiman told staff, we now have a stronger pre-trained base model than the model that currently powers GPT 5.6 Sol. And in fact, we used one quarter of the compute that OpenAI used to train their model. Now, the comparison against OpenAI's models was done using proprietary information Microsoft has access to as part of their investment deal and wasn't shared publicly. However, in public,

17:07Microsoft compared the MAi 2's performance against all open models over a trillion parameters. Across pre-training benchmarks, MAi 2 outperformed on each benchmark against every model, including Kimi K3, and by a wide margin in some cases. Microsoft said that they managed the result with more than 12x the pre-training efficiency as DeepSeq v4 Pro. The model was trained on 16,000 GB200 Blackwell GPUs across 38 days. From there, the announcement post gets a little technical, but the big takeaway is that Microsoft pulled off a very successful pre-training run at Frontier

17:41Scale. PrimeIntellex Eli Bukhoush writes, Wow, amazing technical blog post about Microsoft's new 3T pre-training with very rare and clear insights on how to train models of this size. Now, in terms of why MAi 2 matters, there's the answer for Microsoft and the answer for customers. Among all the companies training AI models, one could argue that Microsoft has had the most torturous journey to the Frontier, or at least one of the most complicated. Initially, they decided to buy rather than build, investing $13 billion into OpenAI between 2019 and shortly after the launch of ChatGPT in

18:14late 2022. That investment looked prophetic, and as the other big companies bumbled around, Microsoft was able to take advantage of OpenAI's momentum and jump right out of the gate with quality AI. However, it wasn't long before there were some fairly significant bumps in the road. Probably the most notable came in late 2023, when Sam Altman was ousted and then reinstated as OpenAI's CEO, with a short period where it looked like basically the entire OpenAI team would go move into Microsoft. I think at that point, it became pretty clear to Microsoft's leadership that no

18:45matter how much they liked OpenAI and liked working with OpenAI models, simply from a fiduciary responsibility standpoint, they were going to need to have a bit more independence. In March of 2024, Microsoft hired DeepMind co-founder Mustafa Suleiman to become CEO of a new AI division. For the next year, we didn't get much from that internal division. All we got was reports of acrimony, like an October 2024 report from the New York Times that had Suleiman yelling at an OpenAI employee on a video call, essentially accusing OpenAI of sandbagging technological breakthroughs.

19:16Even with all of its big access to OpenAI, Microsoft struggled across 2025. Things started to shift this year, or at least Microsoft's new strategy started to become clear when Microsoft released their MAI-1 family of models. Now on the surface, these models were nothing to write home about. The base model was smaller than frontier models of the era, and it was only okay on the benchmarks. The models were adequate as commodity models for doing routine work, and Microsoft started to deploy them across the co-pilot ecosystem for that purpose. But they were certainly a shadow of what Anthropic had achieved with Opus 4.5,

19:50which had been released almost a half year earlier. However, to anyone watching closely, the models were just the appetizer to the entree, which was the new business model. That model was called Frontier Tuning, and it allowed customers to fine-tune their own models on top of a Microsoft base. Now why this mattered is that this was happening at exactly the same time that enterprises were realizing, by seeing it in practice, that simply subscribing to the most state-of-the-art company and trying to run all of their intelligence needs through the top models had no chance of being anywhere near cost-efficient. It turns out that you can use AI for a lot of

20:23things. And especially as we entered the agentic era, where we began to use code and software to solve all sorts of problems even outside of engineering, the total token bill that was coming due was quite significant. And this was even before you had true mass-scale adoption of these new agentic methods across the whole enterprise. Since then, on the business side, the big questions have been about how you addressed that. Some of its policy in terms of new token caps on how much different people can use and what it takes for them to get more access. Then there's the fact that enterprises are becoming model omnivores now, wanting the ability to move in and out of different

20:54models based on the needs of the particular task at hand, rather than be locked into one vendor ecosystem. And that's where the Frontier Tuning model starts to look a little more interesting. Microsoft CEO Satya Nadella has been hammering the drum that it is imperative for enterprises to own the intelligence layer around the intelligence models, and to not give that layer to the companies they work with. That requires building a more sophisticated infrastructure that involves some combination of model routers, or open models, or locally hosted models.

21:27And it's into that environment that Microsoft dropped its MAI models and its new Frontier Tuning business model. And yet, without a state-of-the-art base model, there's only so far that program could go. And that's where MAI 2 starts to matter. If the Post-Train model actually brings it up close to the state-of-the-art, we could have six U.S. companies competing for the frontier, up from just two a few months ago. And what's more, of those models, Microsoft, with its Frontier Tuning product, is offering an answer to what it perceives to be a key in growing enterprise

21:58challenge. While the other labs, like Anthropic and OpenAI, focus simply on the cost side of the equation by offering faster, less capable, but more efficient models, Microsoft's Frontier Tuning model has an implicit answer to the sovereignty question as well. Now, of course, there are other answers to both the sovereignty and the cost question in open models. But given how embedded Microsoft already is in the business ecosystem, you have to imagine that there are a lot of companies that are going to be much more comfortable working with Microsoft to fine-tune a model than going out

22:28in figuring out the infrastructure to host an on-premise model built on a Chinese base. Except

Meta Muse Spark Open release

22:33that Chinese open models may not be those companies' only option. Meta already made a big announcement this week with Muse Wonder and Muse Wonderful. And while those models absolutely put them back in the frontier conversation among U.S. labs and significantly increase what their consumer-focused Muse agent can do, Smart Money is looking at their announcement from Thursday and thinking that Muse Spark Open might be an even bigger deal. Meta is calling Muse Spark Open the most capable U.S. open-weight model to date. The model has 816 billion total parameters, 42 billion active, and a million token context window.

23:08GLM53, by way of comparison, is a 753 billion parameter model, so on a similar scale. AI CEO Alexander Wang said the model is built for coding and agentic work like long-horizon tasks, hard software engineering, and multi-step reasoning. In terms of benchmarks, Spark Open is somewhere between the top U.S. open models like Inkling and Nemotron 3 Ultra, and the Chinese open frontier represented by GLM53 and KimiK3. Its scores are a bit behind KimiK3, on the coding benchmarks, but Meta claimed open state-of-the-art results for agentic benchmarks like JobBench and MCP Atlas.

23:41Artificial analysis scored Muse Open at 39, which put it tied with DeepSeq V41 Flash in 17th place overall, and 7th place among open models. It's about 7 points off the state-of-the-art for open models, but a solid 13 points ahead of Nemotron and Inkling. Artificial analysis hasn't completed their testing of Beam from Reflection AI, so it's difficult to get an apples-to-apples comparison there, but on the few benchmarks that Reflection has published, Muse Open looks on par if not ever so slightly ahead. And, unlike Reflection's model, Muse Open is fully available across major

24:11inference providers with full weights available to download. Now for Meta, this is a triumphant return to open-weight model training. Early on in the AI cycle, Meta was synonymous with open-weight models, releasing the once-beloved Llama family. Particularly during the Llama 3 era in 2024, Meta was, one could say, nipping at OpenAI's heels on the frontier. At the time, many viewed open models as the optimal strategy for Meta. They didn't need to charge consumers 20 bucks for a chatbot subscription, given that they already had one of the most profitable business models in tech. All Meta had to do was play spoiler by giving away AI for free, preventing OpenAI from winning.

24:45Then came Llama 4. Released in April of 2025, the model was disappointing to people right away, and started looking a whole lot worse when OpenAI released O3 the following week. Meta never released the flagship 2 trillion parameter version of Llama 4, dubbed Behemoth, and by the summer of that year, Mark Zuckerberg had embarked on a full teardown rebuild of Meta's AI division. When Meta started releasing models again in April of this year with MuseSpark 1.0, some assumed that they were done with open models. By then, Meta had committed hundreds of billions of dollars to the AI build-out, and they clearly couldn't give it all away for free.

25:19Alexander Wang pledged to return to open models, but some were skeptical that they would get around to it. It seems, however, they weren't just blowing smoke, and Meta has now returned to the top of the open model ranking, at least in the US. Open model researcher Nathan Lambert writes, Guys, this is like my Super Bowl. I waited so long for US open-weight models to be relevant again. Before today, the US models were still behind Chinese models from as early as February 2026. Still more work to do, but this cut the gap by at least half. He later added, If it performs roughly like what the scores indicate, adoption of this model is going to

25:52go gangbusters. First open-weight model in the US with near-frontier agentic capabilities. Nathan reminded that GLM-52 and Kimi-K3 was that step change from China, but added that US models get crazy adoption. Box's Aaron Levy wrote, This is going to further accelerate AI diffusion because enterprises will be able to mix and match frontier performance for orchestration and high-end tasks while being able to have low-cost and tailored post-trained models for repeatable workflows. Overall, this dramatically increases token consumption and workloads across the stack. And so now we are right back again to the same

26:24conversation we were having about Microsoft, where enterprises who are not so keen to just hand over all their data and all their dependencies to OpenAI or Anthropic are starting to have a heck of a lot more choices than they did just a few months ago. Now, one more interesting nugget on the Muse Spark open story. The model, believe it or not, has first-party support from Apple. Now, Apple and Meta not only aren't friends, I don't even think you could call them frenemies. A fairly big chunk of Mark Zuckerberg's life as CEO has been spent trying to get out from under the thumb of Apple based on

26:55its ownership of mobile users and the App Store. And yet, this new model comes with a new partnership. As part of that partnership, Apple has updated their MLX software toolkit to ensure Spark Open will run smoothly on Apple hardware. And both CEOs of these companies took time to talk about each other. Zuckerberg posted a video of Muse running on Mac Studios and said, Muse Spark Open, waits releasing today, runs great on new Mac Studios thanks to work with John Ternus and Apple. Ternus responded, congrats to the team. Amazing to run a model of Muse Spark Open's caliber and scale this efficiently right from your desk. Mac Studio keeps pushing the boundaries of

27:29on-device AI. Talk about a way to announce that you are going to do things differently as the new CEO of Apple. Now, still, just because running this model on Apple hardware is possible doesn't mean it's easy. The full version of Spark Open requires a terabyte of memory, so you're talking about a pair of tricked-out M5 Ultra Mac Studios. The top memory spec for that model hasn't shipped yet, and a pair is going to cost around $40,000. So while that's not really in reach for even the most devoted of prosumers, it's not unobtainably expensive for small startups that care about running private inference. And because it's an open model, you can, of course, run a quantized version and

28:02cram it into a Mac Mini, as long as you're happy with the trade-offs. The companies weren't coy about their past issues. In fact, in an interview with Axios, Alexander Wang credited former Apple executive Daniel Gross, now at MetaSuperintelligence Labs, for smoothing over the past grievances to work nicely together. Said Wang, we're excited for mommy and daddy to be playing nice again. Now, the idea that most enterprises are going to opt away from the easy, hand-me-a-solution sort of approaches that just getting a Claude or a ChatGPT subscription offer is unrealistic, at least in the short term.

28:33But I do think that a growing portion of companies, especially those that do have some technical capabilities, are going to be looking to design more complex model architectures that allow them to move in and out of different models based on their needs and constraints, which could be performance and capability needs, or they could be cost and efficiency needs. This week, both Meta and Muse threw majorly upgraded models into that fray, and I think businesses in America are better off for it. This is a trend we will, of course, continue to watch, but for now, that's going to do it for today's AI Daily Brief. Appreciate you listening or watching, as always, and until next time, peace!