SED News: The NVIDIA-Hugging Face Deal, China’s Proxy Economy, the Open Weight Surge

SED News is a monthly podcast from Software Engineering Daily where hosts Gregor Vand and Sean Falconer break down the biggest stories shaping software engineering, Silicon Valley, and the broader tech industry. In this end-of-summer episode, Gregor and Sean turn to a busy season of mergers and acquisitions, including NVIDIA https://www.nvidia.com/‘s reported $12.9 billion acquisition of Hugging Face https://huggingface.co/, Dynatrace https://www.dynatrace.com/‘s near-billion-dollar deal for AI observability platform Arize https://arize.com/, and Temporal https://temporal.io/‘s rumored raise at a $12 billion valuation. They also cover a run of large funding rounds, from AI security startup HiddenLayer https://www.hiddenlayer.com/ to AI personal assistant Instinct and restaurant software platform Owner.com https://www.owner.com/, all against the backdrop of a deluge of new model releases. The main topic digs into China’s “transfer station” economy, the sprawling proxy market that gives developers cheap access to frontier models officially banned in the country. Drawing on a China Talk report, they walk through the tactics at play, from “one fish, three meals” credit farming to silent model swapping, and unpack how this pipeline of captured outputs and human traces may be fueling the recent surge in high-quality open weight models. Gregor and Sean also examine where Meta now sits in the landscape after shelving Llama and pivoting toward its Muse family. As always, the episode wraps up with a few standout Hacker News threads, including a critique of how log-scale charts obscure the real cost gap between open weight and frontier models, and a look at “invisible companies” as an under-the-radar investment strategy. Sponsorship inquiries: [email protected] mailto:[email protected] The post SED News: The NVIDIA-Hugging Face Deal, China’s Proxy Economy, the Open Weight Surge https://softwareengineeringdaily.com/podcasts/sed-news-the-nvidia-hugging-face-deal-chinas-proxy-economy-the-open-weight-surge/ appeared first on Software Engineering Daily https://softwareengineeringdaily.com.

• 0:11Speaker 00

  Hello and welcome to SED News. This is the monthly format of Software Engineering Daily. We're myself, Gregor, and we also have Sean Faulkner.

• 0:21Speaker 01

  Hey, hey Gregor.

• 0:22Speaker 00

  Yeah, we have the two of us taking you through a spin of the main tech headlines. We then dive into a deeper topic in the middle. And then we finish up with Hacker News highlights where we pick some of our favorites from the last couple of weeks as well. But yeah, both Sean and I have been super busy. So that's why this release is going out, I think, about a week later than it should. So apologies to anyone that was hoping to see one last week. But yeah, so yeah, Sean, you're at what have you been up to?

• 0:51Speaker 01

  Other than the usual road warrior travel schedule, school started again for my kids. So my kids are back in that routine. So we're kind of back in the thick of it. And I'm a little bit nervous because my summer has been, there's been so much travel involved. And we haven't even hit like conference season yet. Like, oh my gosh, I don't know what's going to happen, but at that point. But it has been a great time this summer and I'm glad that we always figure out a way to make these sessions work, which I really enjoy. So despite us living on opposite sides of the world and sometimes having busy work schedules, we're able to find time to do this.

• 1:23Speaker 00

  Yeah, absolutely. For anyone interested, yes, what Sean's 4pm right now, my 7am over in Singapore. Yeah. So yeah. But this seems to be the slot that works.

• 1:33Speaker 01

  Yeah, exactly.

• 1:35Speaker 00

  Yeah, no, on my side, yeah, talking of conferences, yes, we've got the Superbase Conference coming up the, I think, second of October. So yeah, this is just crunch month, especially for that. Anyone that's maybe familiar inside a tech company that has their own conference. It becomes a pretty huge deal, sort of all eyes on that. Unfortunately though, there are always other things, other things that come up.

• 1:56Speaker 01

  It's not like your day job stops.

• 1:58Speaker 00

  No, exactly. It's like conference crunch plus all the other things that's.

• 2:02Speaker 01

  Yeah, ours is coming up in two months too. So people are heads down on that right now as well. And they have been for months probably.

• 2:08Speaker 00

  Yeah, exactly, exactly. Yeah. So yeah, I did actually go to Stripe, do their like a kind of, as a few companies do, they kind of have their headline conference at usually in the Bay Area or the US at least. Which Stripe did in May, and then they do like a sort of world tour where they hit some of the bigger cities around the world, like London, for example. But Singapore does get them, which is great. Like we sort of, Singapore tends to be the hub for Asia for these kind of world tour conferences. So went to Stripe's conference here and caught up with the folks there, which was yeah, pretty super interesting. And some of our other team were like on the stage talking about things. So yeah, exciting times. But yeah, let's roll into the headlines. We're doing a slightly different one this month, which is sort of an end of summer special.

• 2:50Speaker 00

  Yeah, I also feel like it's definitely end of summer, just the sort of feeling of end of summer, even though it's very, still very hot in a lot of parts of the world and etc. But end of summer special of sort of M&A that's been going on. So mergers and acquisitions, there has been a lot, there's been a lot of activity. We didn't cover a ton of M&A through the summer, actually. No, that wasn't happening. Just, we just didn't focus on it too much. So yeah, this is kind of fun.

• 3:14Speaker 01

  Yeah, I think we talked about any scale and scale.

• 3:16Speaker 00

  Oh, yeah, yeah.

• 3:17Speaker 01

  And.

• 3:18Speaker 00

  Any scale and scale, scale.

• 3:20Speaker 01

  Scale AI.

• 3:20Speaker 00

  Scale.

• 3:23Speaker 00

  But no, I don't think we're going to touch on them today. We've got far more things that we can touch on, which is great. Yeah, I think the, as I was going to bed last night, yeah, the news was dropping that Nvidia were definitely definite as it can be, but definitely buying Hugging Face. This, this had been rumored, I think, last week. But yeah, it's been sort of announced for about 12.9 billion. So yeah, this is kind of interesting. I did just actually jump straight to the Hacker News comments on this one to begin with. And of course, a lot of, we'll get into it in a second. What is Hugging Face? I mean, for anyone that doesn't know, etc. But I think a lot of the community comments were, oh, they're just gonna, they're doing this so that they can piss off the big frontier model people, basically.

• 4:05Speaker 00

  I think that was a bit of a hot take from some of the Hacker News.

• 4:09Speaker 01

  Yeah, I don't understand why Nvidia would want to do that because they make a lot of money from those people. In my mind, like, it's really, I mean, I guess if you don't know what Hugging Face is, it's, it's become really like the default hosting and distribution layer for sort of, not OpenAI, but open sort of AI models. And they have a lot of developers, there's almost like 20 million developers there. And there's over 3 million models and there's data sets and all kinds of stuff on there. And actually, funny enough, I had my first ever blog post published on Hugging Face this week. So in some ways I give myself credit for, you know, this acquisition going through, but.

• 4:44Speaker 00

  You got it off the line. That's basically what you're saying.

• 4:47Speaker 01

  I'm waiting to see my share of the 12.9 billion. But, but yeah, I mean, I think that potentially this is a way that they can own essentially distribution of the hardware that all these models are going to run on. And then they also are sort of owning the community layer. And I would think, you know, if you think of a company like Google, especially in the early days, you know, a lot of the moves that they made where sometimes they would give away particular pieces of software or they gave away like Android operating system stuff, it's because it was in their best interest for more people to be on the internet. Because the more people are going to be on the internet, then being like the front door of the internet at that time, it's more search, more ads, more revenue, and it's like flywheel spin. So they could afford to give away stuff because in a lot of ways, it's like a customer acquisition channel for them.

• 5:33Speaker 01

  And I think you could kind of, you might be able to make a similar argument here where the more people who are building models, using models, running models is ultimately great for Nvidia's business because the more that happens, essentially the more money they make. So why not?

• 5:49Speaker 00

  Yeah, absolutely. And yeah, I mean, when I said sort of pissing off the big frontier models, I mean, that is a very hot take. It's more sort of a hedge, really. It's like, which could be construed as pissing off, but busy saying, hey, we don't, that's what I think people are also looking at. If they get too much completely interlocked with the circular financing and so on. But there is a whole world out there that has nothing to do with the frontier model creators. And that basically all, most of it lives on Hugging Face. So then it does make a lot of sense for Nvidia to, to be very much there. And make sure it doesn't get, I don't know, bought by somebody else or, because clearly they were open to acquisition. So who else could have bought this? Probably quite a lot of people. So.

• 6:29Speaker 01

  I mean, it's probably a great landing for Hugging Face because I don't know, I think the acquisition was really about buying essentially a community of users. And it was, you know, buying a net new revenue stream or something like that, right?

• 6:42Speaker 00

  Yeah, yeah, that also makes sense. Yeah, let's see. We did actually have Hugging Face on the podcast on an episode, I think it was last year sometime. So go check that out as well. But yeah, very, very interesting, huge, huge move. And it's slightly unexpected, I think. I think when it was rumored last week, it definitely came as a surprise to people. But yeah, hopefully Nvidia is a good home for them as opposed to any other company that could have come along and done this. So yeah. Okay. So moving on to another acquisition, Arise, is that how you say it?

• 7:11Speaker 01

  Yeah, that's right.

• 7:12Speaker 00

  Nice. Okay. So yeah, Arise, that's, I think you've been covering this one, Sean.

• 7:16Speaker 01

  Yeah. So Dynatrace obviously signed an agreement and acquired Arise for a little bit under a billion dollars, so nine hundred and fifteen million dollars. And if you're unfamiliar with Arise, and which probably a lot of people are, although I see their billboards around the San Francisco Bay Area. They are in the AI observability and evaluation platform space. So if there's other companies in the space like Fiddler and Galileo, there's a whole host of them. But really like some of the core capabilities around like tracing LM apps and agents, what they're actually doing at runtime, detecting hallucinations, scoring that output, monitoring model agent behavior in production.

• 7:52Speaker 01

  Like this is a huge space, like it's very hard, especially when you're running agents or multi-agent systems at scale and enterprise, like just having like visibility into what they're doing. And if you're in a regulatory or regulated industry, you also need to know what they're doing to comply with certain regulations and so forth. So essentially they're in that space, but the acquisition makes a lot of sense. Like, I don't think there's a world where agent observability is like a separate category of product over traditional application observability. Like eventually these things are just going to merge. Because like a business and enterprise doesn't want to have two different observability platforms that have to run for the software anyway.

• 8:31Speaker 01

  Like they'd much rather deal with one vendor. So Dynatrace is one of the well-known players in observability for applications and infrastructure services. So adding this to their toolbox makes a lot of sense. And I think one of the things that will be interesting is part of Arise's differentiation was being kind of vendor neutral and being able to plug into any model or framework. And I think one of the concerns I've seen people raise is would they continue to be neutral in, now that they're part of this like larger entity of Dynatrace? And will that mean that they'll bring some of that stuff, be less essentially neutral in terms of the way that they work? I have no idea, but that's kind of like one of the concerns that was brought up. But I think we're going to see a lot more of this consolidation.

• 9:13Speaker 01

  Obviously, like there's a huge amount of M&A activity going on right now, but I think there's a lot of sort of moment in time AI companies right now or genres of AI companies where they're solving an important problem, but it's hard to create like a really truly large business on a specific problem. So either they have to grow into something bigger, like more of a, like a bigger platform, or they'll get absorbed essentially by existing platforms and from adjacent areas.

• 9:41Speaker 00

  How old did you say Arise was?

• 9:42Speaker 01

  They're pretty young. I think their last fundraiser was in 2025 for what they did a series C for like 70 million or something like that. So they haven't raised that much money. So pretty great. I'm sure they were, everybody that works there was probably pretty happy because they, I don't know exactly when they started, but I would imagine they're too old. I could probably look it up, but they probably haven't sunk that much time into something. And they haven't raised that much money to get almost a billion dollar exit. Yeah. So they were officially founded in 2020. So.

• 10:09Speaker 00

  Yeah, especially in this space. Yeah, it probably would have been hard for them to kind of end up in some crazy multiple in the billions. So yeah, this, this does feel like a nice landing for them.

• 10:20Speaker 01

  Yeah. And I'm sure they were probably at a place where, I mean, I don't know. I'm just kind of guessing, but like the, given that their fundraiser was over a year ago, they're probably at a place where do we go for another funding round, which would probably need to be a fairly large round. And then you're kind of cutting off an exit like this at that point and you're signing up for potentially another like six years on this journey to get to like something really big. So you're cutting off a lot of the potential exits and landings. And then on top of that, you're signing up for another half decade commitment to the company. And sometimes that just maybe doesn't make sense. Like it might be overall like a better outcome for everybody. And I'm sure they'll still be able to continue to work on like the vision of what they're doing in the company, just within a larger organization that can better support it.

• 11:05Speaker 00

  Yeah. So yeah, next one is Temporal. So again, that was, you were covering that one, Sean.

• 11:11Speaker 01

  Yeah. So Temporal, they've also been on SED. I interviewed their CTO a couple of years ago. They might've been on multiple times even at this point, but. So this is not them being acquired, but right now the rumor is that they are looking to raise $500 million at at least a $12 billion valuation. So nothing's closed yet. It could all of course change, but that's the rumor that's out there. And that's more than double the $5 billion valuation they had just six months ago, where they raised $300 million series C. So if you don't know Temporal, they are an open source durable execution platform. They also have a cloud sort of managed offering for this.

• 11:49Speaker 01

  But it lets essentially developers run, write long running, multi-step workflows that automatically can survive crashes and retries and timeouts and failures. Without you having to do state management and retry logic all yourself. If you look at a lot of companies, they'll have some either hand-rolled or some version of durable execution. Sometimes it's Temporal, something else. But they're originally really popular for things like payment processing, order fulfillment pipelines, where you don't want that thing to fail. But AI has been this huge tailwind in this space because AI agents are essentially inherently long-running, multi-step processes that are also very failure prone.

• 12:29Speaker 01

  And you don't want to, every time it fails, have to start it all over again and then repay for those tokens. So having something like a Temporal to be able to pick up basically where you left off and retry and handle all that logic for you without you having to do it yourself is really, makes a lot of sense. I think this concept of durable agents has been a real tailwind for Temporal. And in a lot of ways, like they've been at it for quite a long time. It's like one of those companies where it's like overnight success, but 10 years in the making, like they've basically created, I think this category of durable execution. They put in a lot of hard work to kind of grind. They did a really good job of getting into some of the, the big like Bay Area digital native companies.

• 13:10Speaker 01

  And I think that they were successful there and essentially developers would leave and they would bring that technology to the next company. And they got, eventually they got enough word of mouth network effects where they just now actually cut, almost like getting this like viral growth from it. Cause they have so many network effects of people being really happy with the product and taking it with them.

• 13:29Speaker 00

  Yeah, they seem to have also this year. I mean, obviously I've been through this with Superbase now this year, but they're sort of going from like C to D to E in one year, basically. So like they did series D, Temporal did series D in February of this year, I think. And so this would be not really even six months later, they're looking at their next round at double, I believe that's double the valuation. So five billion on the last round and then, or more than double and 12 billion on this round. So yeah. And I think that context is really interesting of yeah, how Temporal has been around a long time, basically replacing state machines and this kind of thing.

• 14:06Speaker 01

  Step functions and.

• 14:08Speaker 00

  Yeah, just like retry logic effectively that used to be just hand-rolled, I think by every company. And then, yeah, and then now it's only this huge tailwind with AI for all the reasons you just mentioned. So yeah, I think super interesting. Yeah. Next one that crossed the radar was Instant DB. So they are, it's definitely more of an acry hire than an acquisition. At least that's how it's being yeah, reported. But yeah, the team of Instant DB joining OpenAI. So yeah, Instant DB is yeah, open source, basically backend, you got a database off storage layer. But the key thing here is like, this is a very useful thing for agents, basically being able to spin up that layer of infrastructure.

• 14:51Speaker 00

  And yeah, clearly OpenAI don't seem to be doing a lot of the building of that themselves, at least not that we can see in public. Acquisitions seem to be, or acry hires again, in this case, again, seem to be how they're doing it. This is interesting and definitely one where there's a cloud offering. They had a cloud offering, which is I believe going to be mixed by OpenAI. I think, you know, this is when developers get a little bit angsty when, you know, there's a product that they use and okay, the team is going, the team's going to go in-house with OpenAI. But ultimately this product that they, I believe they said they were handling like 400,000 apps. I think this is a bit of a loss, I guess, for the ecosystem at the moment, perhaps if it just gets, by getting just sort of subsumed inside OpenAI. But we shall see, see if there's a better version of this.

• 15:37Speaker 01

  Yeah. I mean, it seems like with OpenAI might be, yeah, some of these moves, who knows what their actual intention is, but it seems like they're kind of trying to build their own agent infrastructure stack with some of these pieces that they're kind of assembling. And, you know, they've been making all kinds of different moves in agents for two years now. And I think part of that is like, how do you capture more of the AI value chain outside of the model itself? And even outside of, you know, what they were originally really well known for, which was like the consumer interface of chat GPT around their models.

• 16:10Speaker 00

  Yeah. One of their investors was Jeff Dean. So pretty big name. And yeah, we've got also some just sort of, I think, general, general news on Jeff Dean. Is that right?

• 16:20Speaker 01

  Yeah. So I mean, the big news over the last month was that a little less than a month ago, at the beginning of August, Jeff Dean announced he was leaving Google after 27 years. I mean, he's like employee number 30. He was most recently chief scientist. He's probably one of the most famous and world renowned engineers in the world. You could probably make a fair argument that Google might not exist. At least in the form that it is today, Jeff Dean hadn't been there with all the contributions he made to, you know, MapReduce, Big Tables, Banner, TensorFlow, Google Brain. Like he's just this godlike figure, the fun Jeff Dean facts and that are like these like Chet Norris memes and stuff like that.

• 17:00Speaker 01

  That I remember it was like, uh, only Jeff Dean knows the final number of pi and crazy things like that. Jeff Dean doesn't get compiler errors. He tells the compiler that it's wrong and fixes it and just stuff like that. So it is quite the thing. I never thought that he would leave. I figured he would just eventually retire. But it looks like he's going off to start his own company and he took three really heavyweight people with him as well. And they're now starting this company called Discovery Loop, which is, they're not, I don't think there's a ton of information out there, but they say it's a public benefit corporation aimed at using AI to automate and massively scale scientific experimentation. Definitely something interesting to pay attention to, see where they go.

• 17:42Speaker 01

  I would imagine them raising money is a very easy conversation. They probably just, here's my resume and slide across the blank check and have people just fill it out. But it'll be interesting to see what happens at Google coming out of this. Clearly that's a pretty big loss. And Google did invest in Jeff Dean's company and maybe there's a certain marketing angle to that to kind of save face and all this when you have such a high profile person leaving. But what do you think, Gregor?

• 18:08Speaker 00

  Yeah, no, I was sort of thinking, okay, what is the, you know, I love tipping points, right? What was the tipping point for Jeff Dean to leave Google? And I guess it's just like, this is, there has been no better time for someone of his stature and mind, I guess. Where you are going to get the back, if like, if you actually want it, you can probably have an amicable conversation and be backed still by Google and even bring people with you and not become some kind of rupture. We've seen this a few times and yeah, I think that's it really. Like there is, there has been no better time that he could now step away from like the mothership.

• 18:45Speaker 00

  And to do his own thing, but I'm sure it will be very closely intertwined with things that Google do anyway. So yeah, I think that's definitely life goals right there. If you can give such a contribution through one company, but then get like all the backing to then go and do your own thing from that company as well. And be able to take some of the, you know, amazing talent that Google still has. Like, yeah, amazing. Yeah.

• 19:07Speaker 01

  Yeah. Just one more fun Jeff Dean fact before we, we go off this is, uh, Jeff Dean proved that P equals NP when he solved all NP problems in polynomial time.

• 19:20Speaker 00

  Yeah, there we go.

• 19:23Speaker 02

  You're shipping faster than ever with AI coding agents, but those agents don't vet the packages they pull in and they don't have security context built in. Ori by Endor Labs fixes that. It plugs directly into your editor via MCP, catching vulnerabilities, blocking malicious packages and flagging exposed secrets in real time. No separate tool to switch to, no dashboard to babysit, security that fits how you actually build. Teams using Ori see 10 times fewer security tickets and six times faster fixes. Free for developers. Get started at www-endorlabs.com/auri. If you're running Postgres in production, you've probably felt the moment analytical queries start fighting your transactional workload. Most teams end up adding a second database and all the pipeline complexity that comes with it.

• 20:07Speaker 02

  Tiger Data, creators of Timescale DB, takes a different approach. We extend Postgres with hybrid row and columnar storage, so one table handles both writes and analytical scans. Native compression cuts storage costs up to 95%. Continuous aggregates keep dashboards live without bash jobs. And it scales to petabytes without you re-architecting. Companies like Cloudflare, Octave Energy, Schneider, Axpo and Flowco run production workloads on Tiger Data today. No stale data, no second system to operate, just Postgres. Managed for you, ready for the workload you're building toward. Try it free at tigerdata.com. Your customer lives in the real world and it's messy. Intermittent connections, mid-onboarding drop-offs, edge cases on devices you've never tested. Mobile apps reflect reality in a way no other surface does.

• 20:52Speaker 02

  Yet, from an engineering perspective, they're the hardest to understand. It's common for mobile engineers to see green backend dashboards, normal error rates, no crashes. But inevitably, somewhere, there's a frustrated user watching your app spin. After a few seconds, they'll lose patience, close the app and turn their attention somewhere else. They may never come back. The worst part? Most observability tools never see any of it. Bitdrift, on the other hand, captures 100% of mobile data, unsampled and in real time, so it's immediately queryable by engineers and AI agents. It's mobile observability built for the real world. Try Bitdrift today. bitdrift.io/ signup.

• 21:31Speaker 00

  So, back to just, I guess we called it M and A and it's more just A and R, acquisitions and raises. So, so back to A and R. Yeah, this one is a little bit, maybe for some, off the radar exactly. But that's why I wanted to bring it in, is a company, Hidden Layer, who are a security startup and they make tools to protect AI models, agents and workflows. And obviously, we have covered so many of these attacks that have happened involving some form of agent or whether it was intended or not. So it's very interesting that, yeah, they've managed to raise 100 million, which actually now for a security startup, that's pretty impressive. There was definitely seemed to be some, I would say, decline in the security company raising space. I haven't seen a lot of big raises from security companies, strangely.

• 22:16Speaker 00

  Even though we keep hitting all these security problems, it just seems that unless the company is basically something AI and then they maybe weave security into it, but this is very much no, like they've been in security the whole time. They're just, they've managed to, I think, pivot that story and the product well into the AI space. And they claim that, they haven't given exact ARR numbers, but they say that they're now in the, ARR is now the tens of millions and like 90% of that was driven by new customers signing in the past year. So again, that's tens of millions in ARR for a security company of their age. Impressive as well.

• 22:55Speaker 00

  It is still pretty challenging to get companies to pay for a, quote, "new" security company. Well, I say nobody got fired for buying CrowdStrike. Maybe somebody did at some point.

• 23:08Speaker 00

  But honestly, I still think nobody gets fired for buying CrowdStrike. That's just the classic case of, yeah, they had a bad month a couple of years ago, but then nobody seems to remember that anymore. So.

• 23:17Speaker 01

  I think it's steep in security. It's like a steep hill to climb because unless you're doing some sort of like, you know, maybe consumer related to security. But you know, most of this is going to be like B2B. And how do you get like a truly large enterprise to trust some smaller player in the market that's relatively new with something like security? But I do think that, given that trust in governance and security around models is like so top of mind for businesses right now. I do think that there's going to be a lot of companies that kind of emerge out of this, new companies that have raises and are addressing like real problems in the space. Cause that's a big part, I think, of what needs to be there to unlock B2B enterprise AI,

• 23:59Speaker 01

  which is significantly wagging behind other areas of AR right now.

• 24:03Speaker 00

  Yeah, and like, I think it's interesting. They want to be known as like an EDR solution, which is endpoint detection response. But specifically for AI, they do things like they scan models that you are using. They actually scan the files to like, make sure that you are using the model that you think you're using. So again, like having spoken about Hugging Face and we're going to, actually, our main topic is a lot about the open weight slash quote, "open source" ecosystem. And so Providence, which is, do you know where this thing came from? This very much weaves into that, which is like, do you know the model? Is the model the one that you really think that you're using? You know, is the, we've seen, obviously, we've seen all sorts of like package manager based attacks, you know, where someone can take over a name of something.

• 24:47Speaker 00

  On Hugging Face, I must say, I'm not fully, I don't go into Hugging Face a whole time these days, but I remember when using it, it was still kind of confusing at, at least back, I don't know, a year and a half ago, two years. It was still maybe confusing, like, is this a model made by anyone reputable with data that I can trust or not? So, enterprise has to be super careful now when they're pushing their developers to use AI and especially open the small amount of actually open source models. Like, yeah, great, but do you know what's, what's in them? So, anyway, I think that's why we've seen this from Hidden Layer. They've managed to 100 million at Series B, which is, yeah, super impressive for a security based startup. Yeah, just moving us along, this one is super buzzy startup.

• 25:32Speaker 00

  Got coverage in Wall Street Journal, but instinct.ai. I think buzzy, I want to just keep touching on that because this was a 250 million, Series B at 2.5 billion valuation. Honestly, I'm sure if there's anyone listening out there knows this product, that's great. I'm still a bit confused what the why here of why, why? This feels more like the team, you know, you've got like Noah Shin is, I believe, the, at least one of the co-founders, he was at Sierra. And yeah, it's like youngish team, like, and it's sort of like an open claw, but polished. It's kind of my read on it. Yeah, I'm curious like why the crazy valuation on such a young company. But, but yeah, I don't know if this crossed your radar as a company, Sean, or not, but.

• 26:14Speaker 01

  No, I hadn't seen the news around this. Yeah. But it's basically from what I could tell, kind of going after the personal assistant, AI personal assistant market.

• 26:22Speaker 00

  Yeah.

• 26:23Speaker 01

  And unless of like an open source, completely vibe-coded thing by a single individual and more, more of a company approach. I mean, I don't know, I'm assuming that they were able to get that valuation in part because of maybe who the founders were and then also in combination with whatever growth they might be seeing. But I also would think that like the total addressable market is massive. Maybe that helps justify some of it, but that does seem quite a large or a very high valuation. And also for $250 million Series B.

• 26:53Speaker 00

  Yeah, but I realized in our notes, I put the wrong URL. So there's instinct.ai, which is not them. What's fun though, if you go to instinct.co, which is them, it is just a website that looks like it's from 1992 and it's.

• 27:08Speaker 01

  Yes, it does.

• 27:09Speaker 00

  It's a white background, it's Times New Roman, it just says instinct. And then it says contact, and then you've got San Francisco and you've got, you know, the comms. It's all written, there's no, there's no signs, comms at instinct.co. Yeah, I'll just do a quick read. It then just has one, basically one about paragraph. And it just says, "Instinct is a personal assistant that understands what you're working on and what's important to you. It connects to your applications and devices, email, messaging, screen, audio, location and more. The interface is simple, there are no new interfaces. It's trained to use a phone and a computer." I'll pause there, there's more. But yeah, very sort of open claw-ish. But I mean, this is the ultimate, I'll just say, this is the ultimate flex right here. You know, hey, we have a website that is just a white Times New Roman

• 27:53Speaker 00

  and we just raised 250 million, 2.5. So yeah, signing the Times perhaps, but.

• 27:59Speaker 01

  They're going for agent readability on the website, right?

• 28:01Speaker 00

  Yes, that's definitely what they're doing. You know, agent doesn't need to know anything more than, than this basically. But yeah, you know what, I'm gonna, if I can in Singapore, sometimes things are not available here. Yeah, so I'll try it out. I was never an open claw person, but maybe this can, I definitely need some form of personal assistant, executive assistant. So I will try this out. All right, and then just a final one. I just popped this in, it was kind of interesting. This is, it's called owner.com and they basically just do kind of like backend software for businesses. Like, it sounds super boring, unsexy effectively. It's an eight-year-old company, but I think it's kind of cool. They've raised 240 million, Series D at 2.3 billion. So actually kind of similar numbers, just different rounds to Instinct. But yeah, I think this is cool.

• 28:45Speaker 00

  It's good to see like how companies, if they really do keep moving themselves forward and taking advantage of like what AI can do, then you can still be raising on these like unsexy business models. So kind of almost the anti to Instinct in this case, yeah.

• 29:01Speaker 01

  That's a hard market to like selling into the long tail of restaurants like OpenTable was able to do it through a lot of just sheer will of beating the street with, you know, sales reps and stuff like that. And back then, when OpenTable did it, they were actually putting like physical computers into the restaurants and things like that. But I think the thing is, is like, if you can do it, because it's hard, it's very sticky. And then you have this like mass amount of customers that probably part of it's like, okay, well, I've captured the interest of this customer. Like, how can I, what are there other things I could sell into that same customer that help them that also help me by generating more revenue?

• 29:37Speaker 00

  Yeah, as you say, it's like, it's very restaurant back office, or like, we say back office. Yeah, it's like they cover everything like point of sale, a branded restaurant app, marketing campaigns, online menu, point is the online menu, like literally a restaurant menu, a reviews engine, all this stuff. So they're kind of trying to cover every single base that a restaurant or some, any kind of food service, I guess, would be needing. So yeah, super verticalized in that sense, but nice to see just something that doesn't, of course, AI comes into it, but it's not just an AI company claiming to sort of solve all of those problems. It's very specific for something I think we can all relate to restaurants. So.

• 30:13Speaker 01

  Yeah, it's encouraging for all those founders out there that might be listening that don't have a pure AI company. Like, yes, you can actually still mill the business and raise money despite what maybe all the headlines tell you.

• 30:23Speaker 00

  Yeah, if you want to raise money, not all do, but if you do want to raise money for something that seemed boring, definitely, definitely still possible. So, yeah, nice. Yeah. I mean, I guess just sort of, we're going to get into our main topic in a couple of minutes. In terms of other news, I think it's just that we've seen an absolute deluge of new models coming in, like my night last night was top of hacker news was MetaMuse Spark 1.3 and Gemini 3.8 Flash. And then now, waking up this morning, what do we have this morning, Sean?

• 30:55Speaker 01

  GPT-6, Astra.

• 30:57Speaker 00

  Yeah.

• 30:57Speaker 01

  Hot off the presses.

• 30:59Speaker 00

  Yeah, very exciting. So we have admittedly not had time to really dig into all the things that Astra claims to be able to do, but certainly from the pretty, I would say like they've gone up a level from the launch website perspective for, for Astra. Again, IPO looming perhaps, but yeah, they're really going all out pitching it against Fable 5 and 5.1 and showing lots of nice charts that it's all the benchmarks are far above. So yeah, we'll probably get into that one in more detail next month. But yeah, any hot takes?

• 31:32Speaker 01

  I mean, I think it's just limited release right now, but I mean, some of the things that I was able to dig into with the press was they talked about how they had to spend kind of extra time shipping safeguards and adding, enhancing the safeguards around it so that the model couldn't find and weaponize zero day on its own. And I think part of this is, you know, trying to really address some of the criticism that we've seen over the last couple of months when it comes to the frontier models. There was, of course, what happened and we covered previously with the Methos Fable of it all, where, you know, it was around for a week and it was pulled. I guess, who knows, we're only, we're less than a day into Astra, so maybe it'll get pulled off the market later this week. But I think they're trying to address some of the concerns there.

• 32:17Speaker 00

  Yeah. So we have spent quite a bit of time on sort of, normally this is like the headline section and we've obviously gone through a lot of tech A and R, as we now call it. The main topic this week, it's actually kind of running on a little bit from last month's where we dived into, we called it the Kimi moment last month, Kimi, this open weight model that literally all the weights you could find on Hugging Face and was really starting to show that an open weight model can rival the closed source frontier models. Something that then popped up in my, actually just from a friend sending me an article about what's called the transfer station market in China. And that's kind of framing what we're talking about today. It's, it's sort of like the

• 33:00Speaker 00

  last couple of weeks is really like models, there's been a couple of weeks where the model map, if you like, has been a bit, has been rewritten in some ways. But I think just diving into how we got here is interesting. We didn't really touch on that so much last month. So kicking off, yeah, we've just seen a, even since we, we talked about Kimi last month, we've seen like a bit of a proliferation and open weight releases. Some of these names might be new to people. So z.ai released GLM 5.3 Flash. We've had Quen dropping a whole bunch of new models and all sorts of new models coming out from all these Chinese foundries, or however they like

• 33:40Speaker 00

  to be called. But what we perhaps didn't appreciate was that for a while, there's been this transfer station economy, which is basically where although you cannot access frontier models from OpenAI and Anthropic in China, there has been a huge economy where basically you are able to access them for very low cost. So I'll kind of go through the three ways they've been able to do it. And then where we just have a little pause there. So basically, the first way, apparently, I love all these kind of naming conventions and like come out of like Chinese proverbs and stuff, but apparently it's called the one fish, three meals. And this is from this article. So

• 34:22Speaker 00

  I should definitely give a shout out. So the blog is called China Talk. It's unattributed because for obvious reasons, I think there's far too much being talked about here that if someone is named on this, that could be problematic for them. So totally understand that. So the one fish, three meals, what is that? So basically, they talk about either you can, meal number one from the fish is you can bulk register accounts to farm free credits, reselling unused quota, corporate discount arbitrage and API maxing. So one $200 max plan gets basically carved up to multiple users. Meal two is what's called model swapping. So users select Claude Opus, but the proxy would silently route to Sonnet, Haiku, or in a sort of quote, worst case GLM or

• 35:07Speaker 00

  Quen, fraudently relabels the output. So like researchers audited 17 API proxies and found that widespread model swapping, basically. So proxy access to, quote, Gemini 2.5 achieved only 30% on a medical benchmark, which is way off the actual official 83%.

• 35:25Speaker 01

  What are they doing with the model swapping? Like, are they essentially capturing the value of Claude somewhere else? So essentially, someone thinks that they're using Claude Opus, they're paying that amount, they get some cheaper basically version of that in response, and then the labs are using the delta between that to do something for themselves?

• 35:47Speaker 00

  Yeah, I think, I think that's exactly it. So yeah, user thinks they're using, let's just say Opus, what, yeah, which was the example given here. They think they're using Opus and basically, yeah, lab goes off and uses, charges them, you know, because they say, hey, I'm a proxy. So it's not, I'm trying to think of a good example. I don't want to muddy the waters with names, but like, I think an example here that at least gets people thinking is, imagine Cursor. It's not Cursor, just to be super clear. It's not Cursor. But like, imagine, imagine Cursor was to tell you, hey, would you like to use Opus or Sonnet? And you pick Opus, and then I get a response back and I think that Opus created that and I pay for Opus amounts. I

• 36:29Speaker 00

  instead get a Sonnet response back and I go off and keep doing my work. However, lab keeps the money that I paid them for Opus and then goes and actually uses Opus to, yeah, and we'll get to sort of why are they doing this at all? But yeah, that's, that was a good, good sort of question.

• 36:44Speaker 01

  Yeah. So they're basically using the excess sort of capacity that's there because they're not actually responding with the high value model to use it for their own purposes.

• 36:53Speaker 00

  Exactly. Yeah. Yeah. Meal number three is the logs are the products effectively. So every request passing through a proxy, like a prompt or response or tool calls or iterations that all does sit on this proxy server. And if it wasn't clear, yeah, these users in China are like, they're using these, they pay a proxy to get access to these models. The proxy tends to go through Singapore. So apparently Singapore data centers are like some of the most heavily used in the, or VPNs heavily used in the world. So yeah, so the logs sit on the proxy server. So then for AI coding agents, they contain all the reasoning chains, engineering decisions, like repository context, human verified correct outputs as well. You know, they're basically capturing what a proxy literally is, you know, it sits in the middle, it captures basically all the ins and the outs

• 37:38Speaker 00

  as if you were Anthropic or OpenAI. Yeah.

• 37:42Speaker 01

  But without the concerns of user privacy or GDPR or anything, it's like, just farm this user data however we find necessary essentially.

• 37:52Speaker 00

  Exactly. Yeah. So, and apparently that, yeah, certainly that last one, like logs that seems to be, I guess, quote, the margin of these proxy products. Like, I think they see that as the highest value, basically. You capture users through.

• 38:07Speaker 01

  It's real human behavior. And then it also, those form essentially like memory paths. It's like, I'm trying to accomplish this task and the traces give you a graph for how to accomplish that. And it's probably not always going to be that efficient. So then, if you mine them, you can take that into training and synthesize essentially sort of better neural pathways for similar types of memories.

• 38:33Speaker 00

  Yeah, exactly. So I think this is like, just, I got to say, I had kind of, I had this this certainly. And so when I was hanging out with a friend the other day, who's a programmer, CTO, and he spends a lot of time in China. He's German originally, but a lot of time in China. And so he's always the guy who just knows what's going on and why and so on. So, yeah, he was talking to me about this and I was like, what are you talking about? And he's like, okay, I'll send you this article. So, yeah. And why does that even kind of play into, especially what we're talking about last month with Kimi and, and just this open weight surge, if you like? Well, it doesn't mean, this is not the entire reason this has been possible, like that would be

• 39:14Speaker 00

  undermining, but distillation, this idea that you can capture the inputs and likely responses from these models, you basically are able to then take that and then train models, open weight models, or create open weight models. That sounds a bit like, oh, but how on earth would you do this? Because it's actually that's super slow. Or I think the volume here is what's always hard to like fathom. The Chinese population and just, I think, the sheer scale of what has been going on here is like absolutely unreal. So, yeah, I mean, when people talked about distillation, I don't think it was fully understood. How do you even do distillation? Well, this is a, a sort of a very quick and interesting look into how that even works. And then, yeah, that is sort

• 39:59Speaker 00

  of a huge reason, I think, that we have seen just a massive proliferation of, of open weight models and like of the quality that we're seeing. Again, lots of very, very smart people in, in China creating these models, but it is, these proxies do exist, this stuff is happening. And it's a bit hard to see why that would be happening without, yeah, one seems to have led to the other, at least in some form.

• 40:21Speaker 01

  Yeah. I mean, in order for you to, the part of the process of distillation is essentially using. A larger model or like a teacher model to generate the final outputs, the answers or confidence scores. So and then you use that to essentially teach the other model. So you wanna make a really good open weight model and you have access to the best frontier model and you're not even paying for it because you're like, you're using the margin that you're saving from deviating routing to cheaper models. And then on top of that, you also have human trace behavior for how they're interacting with agents. Then that's all data that you can use for essentially distillation of reinforcement learning and fine tuning to make these open weight models really, really

• 41:06Speaker 01

  good without having needing access to like, this is a way essentially you can get access to the models and all the data that you need to do this process.

• 41:15Speaker 00

  Yeah, it makes the idea of models or banned. It's very interesting. Like if a model's banned, but then by say the Chinese government, but then equally these proxies are sound like pretty prolific and you wonder actually, are these actually supported by the country as opposed to like, is it a blind eye or is it actively supported? And it makes the story of why a model's banned even more interesting. Yeah, you wouldn't really be able to do this on the proxy level or like the proxies wouldn't have any teeth for a user to go and use them if you didn't need to go do this. So I think it's absolutely genius if I'm just being super honest. I think it is super smart because you always got to think 10 steps ahead of like, well, we don't want, at least that's

• 42:00Speaker 00

  probably I guess the way the Chinese government thinks like, oh, we don't want you to have access to this, but clearly they are very powerful and very helpful. Like, but what's the way we could sort of harness the output of this? It's super interesting. So yeah, hopefully that gives a little bit of backstory to where we are at the moment with the open weight side of things.

• 42:19Speaker 01

  Yeah, it does make you think like, if they're able to do this, you know, these types of things, that it'd be hard to get away with this in the US, I would say.

• 42:27Speaker 00

  Yeah, yeah.

• 42:28Speaker 01

  You know, what does that mean?

• 42:29Speaker 00

  You'd hope it'd be hard to get away with it. Yeah.

• 42:32Speaker 01

  What's that mean for, how does like an open weight model stay competitive with essentially this market that is kind of do, you know, stopping it, nothing to be able to make better and better models? If you're willing to like, essentially circumvent people's privacy, circumvent certain laws to train these models and you're, and then you're a legit company not doing that. It makes it pretty hard to have an equal footing in terms of the competition.

• 42:56Speaker 00

  Yeah, well, I think that's a good segue into then. I was also just for this week generally sort of wanting to maybe look at within the main topic, what is Meta doing? Cause you know, that's a topic, we haven't really, we haven't touched on Meta in a big way in a while. And thinking about it last week, I realized I don't actually even know what Meta are up to, which says a lot because it just means they're not terribly interested from a pure personal coding level. But it is very interesting just to look at them from the landscape point of view. Like I hadn't even appreciated that they did actually shelf the Lama family of models like back in April. Everything's moved to Muse. So there was actually, it was a closed model Muse code. So that was starting to show a bit of a reset and that was under, you know, Alexander Wang, who'd come over

• 43:39Speaker 00

  from scale.ai. And then, but now they're moving back to sort of open weight models. So there's Muse Glimmer, which was under Apache 2.0. So then to your point, yeah, Sean, I think they can compete in the sense that open weight models now have a lot more reputation in the sense of qualitative output. Open weight doesn't mean less quality for certain tasks, perhaps like coding, like you can actually get very, very far on these quote cheaper models. And Meta is not an AI company. Like it's, it's still a social platform company. It's very difficult, I think, for them to shift that image of themselves.

• 44:16Speaker 01

  I don't really know, like, what's their go to market around their models? Like, it's just not the natural.

• 44:21Speaker 00

  I think it's compute, basically. Yeah. Which has come out here. So it's like, okay, we can still release an open weight model, but it's on Meta compute. And so we're just going to go down the traditional.

• 44:31Speaker 01

  Well, yeah, I mean, that's how they monetize it. So it makes sense to basically give the model away because, or give the weights away essentially, because it's almost like a customer acquisition channel. That's how you market it. But Meta is probably not the natural location that people they're thinking about. But even if, I think more and more companies are going to have essentially a hybrid model strategy where they have some closed frontier models from the top labs, but then they also have open weight models and probably a lot of them are not going to necessarily run those open weight models themselves. They'll, they'll go through an inference provider like fireworks or base 10 or whoever it might be to do that. But then, uh, they're gonna want to do

• 45:13Speaker 01

  that. So they have kind of a model agnostic strategy and they can direct certain workloads to different models, depending on like what the, the use case of the workload is. But is Meta a consideration in there? I wouldn't think it would be the natural location that a lot of these companies are going to go. I guess maybe Meta's approach is like, hey, well, if we can, you know, continue to prove that our model's the best in the open weight market, then naturally people will want to use our model because it's the best. But that's going to be tough to do. Like we said, given that all the moves that the Chinese companies are doing.

• 45:46Speaker 00

  Yeah, absolutely. So we are going to have to leave it there. I think hopefully that's been like a little insight. I know we like to try and sort of do a deep dive. I, I hope that's been deep enough. This week we had so much to cover in the first segment, but yeah, go check out, if you're interested. Yeah, it's the article, it's on this, yeah, blog called China Talk. It's called how to buy cheap cloud tokens in China. Very, very, very straightforward. Yeah, that was published in May. There is, sorry, there is a name against it. But yeah, so the kind of the who overall is behind this blog. Yeah, is a little bit less, but this person, they are a research associate at Oxford China Policy Lab and they hold a master's degree from University of Oxford. So I do feel like a lot of the research here is pretty well backed up. So go

• 46:31Speaker 00

  check that out. Yeah, we're going to probably only have time for maybe one each on the Hacker News highlights. But yeah, favorite part of the show. Let's get into it. Sean, what was your highlight from Hacker News?

• 46:44Speaker 01

  Yeah. So I grabbed one. It's kind of still on theme of some of the stuff that we're covering around LLMs and inference and cost. But there was a recent, there's a blog post, it's on openteam.com, intelligence versus cost. And it was posted by the anonymous one on Hacker News. But essentially it's a, someone's like, personal deep dive into, firstly, they're critiquing some of the existing charts that you see of, that people use to sort of compare intelligence versus the cost for LLMs. And like, one of the things they point out is that the chart uses a log scale, which visually ends up flattening the actual massive price gaps. So their argument is that's kind of misleading and it doesn't really show the true story. And then what they do in the

• 47:27Speaker 01

  article is they redraw that on a linear scale and show like what the real world pricing and the gap is between the open weight model and like a Fable 5.1. And essentially their argument is that Fable's seven and a half times more expensive, but you're only getting a small sort of intelligence bump that most people wouldn't even notice. So does it really make sense if you could take an open weight model, run it even on a last generation GPU for, you know, 1400 bucks a month or something like that, unlimited, and you're getting kind of 90% of what actually people need is kind of the argument. Now, I will say that this is someone's personal journey and it's not necessarily scientifically based and things like that, but just interesting kind of food for

• 48:12Speaker 01

  thought that I thought.

• 48:13Speaker 00

  Yeah, no, that's super interesting. Yeah. On my end. So I'm trying to think which one to pick. I'll go with, it is, I mean, it's on Hacker News and I think that's what's cool, but it's not super tech related, but AI does kind of come into it as well. But it's a great article called Invisible Companies. And this was posted by Eton Non-Roe. So what is an invisible company? Well, it's basically the argument that people can actually make good money from finding companies that just are very good in their sector, but are just kind of like completely not the buzzy company to buy. And so if you're a VC and you want like a slightly alternative strategy, then

• 48:53Speaker 00

  you can go and find companies that very much are outliers in like their segment for how much money they're able to produce. But the reason that it's called invisible companies is you actually also don't want them to become visible. You don't want them to kind of get into, become a segment that it becomes hot and people now want to buy companies in that segment. Otherwise, because the whole point is that aim isn't to sell these companies on necessarily, it's to hold them and to make money from them. AI comes into it in the sense of like, well, you can definitely do a lot more research much faster to find these companies. But does that make them more or less visible? That's part of what the article discusses. So yeah, I don't know. I'm

• 49:34Speaker 00

  sure we've got a lot of listeners out there who are just generally like entrepreneurs themselves. So I definitely found this one very interesting as well. And just completely, completely off tech topic. Sometimes it's nice just to read something that's not related to models or AI models or that kind of thing. So.

• 49:51Speaker 01

  Absolutely. Awesome.

• 49:52Speaker 00

  Yeah. And no doom this week, unfortunately. No need to doom running on something, but I, I hope, I'm sure we'll find something for next month. But yeah, I guess just any, any quick thoughts on, on what we might see in the few weeks ahead, Sean?

• 50:05Speaker 01

  I mean, given this, it's not a real hot take, but given all the movement that we, you know, all the M and A, would you describe it? A and R?

• 50:13Speaker 00

  A and R. Yeah, yeah.

• 50:14Speaker 01

  Yeah. I think that's really just the beginning. I think we're going to see a lot going on through the end of the year of consolidation of companies and also large fundraises from various ID of different companies. Cause even I would suspect like a lot of the companies that raised fairly large rounds in 2022, that when the market was really hot and were sort of overvalued and to a point where it was hard to lift the value, like some collection of those companies are probably in a place where they're running out of money and maybe don't have the growth to raise. And then some collection of them have probably done really well and are going to raise money and so forth. So we'll probably see a lot of that.

• 50:50Speaker 00

  Yeah, I was just double checking. I don't see, my prediction last month was that we're going to already see Kimi 3.5. Sadly, no, but we've definitely, you know, as we've covered a lot of stuff in the last, a lot of open weight in the last month. I think my prediction for the next few weeks is I really think we're going to start to see some cost inflection point. I think there's going to be some kind of like, not backlash exactly, but I do think there's going to be some moment where like some big company somewhere perhaps is like, we are, we're not going to use that topic anymore. It's too expensive. We're using Kimi or we're using something. I'd kind of like to see it because I think I would like to see the cost being driven down to some kind of equilibrium. But yeah.

• 51:31Speaker 01

  I would suspect too, one other thing is that I think we'll see a big model announcement from Google sooner rather than later too. Cause they've been way too quiet, other than like the Jap Dean news.

• 51:41Speaker 00

  Yeah. Cool. Well, yeah. Thanks everyone for tuning in and we shall catch you next month.

• 51:47Speaker 01

  Thanks, everyone. Cheers.