Back to all episodes

Can 150,000 people save NVIDIA? - This Week In AI

November 17, 2026

Jensen Huang resigned as NVIDIA's CEO on November 12, and a petition to reinstate him has passed 150,000 signatures. Shane and Abhi talk through why the board may have pushed him out, what prediction markets say about a comeback, and what NVIDIA without Jensen could mean for the AI race. Google's Gemini 4 Hydrogen and Alibaba's Qwen4-Max are out, along with a debate about Google naming its models after the periodic table. A new benchmark called Continual-ARC finds that OpenClaw, Hermes Agent and Prime Agent don't really learn from experience, and Abhi explains how it works. Plus Anthropic's pretraining efficiency slide, LangChain's $300M raise, shadcn's UI SDK and Eve 1.0.

Episode Transcript

00:00

Abhi: When you're in a position of power like that, you really can't slip up, dude. No, you can't get caught. If you're gonna do stuff, just don't get caught, man.

00:06

Shane: It's much easier if you just take the ethical and moral high ground. Hello everyone, if you're just joining us, this is Agents Hour. We're doing the news, it is Monday, November 16th, and there's some drama to talk about. There's some news. Let's get into it Let's go ahead and start with the big news of the week. You know, we covered this last week, at least the initial findings or the initial reports, and then on November 12th, the day after we did our last show, Jensen has officially resigned as CEO of NVIDIA. And for those of you, you know, wondering why this is important, well, it was the largest company in the world, I guess, for at some points. And if the CEO resigns, this is kinda has a big impact.

00:56

Abhi: I mean it started with the Chinese spy and then there were some other things coming out like over the weekend essentially.

01:02

Shane: Yeah, and I think we don't have all the answers, of course. I think there's investigations ongoing. One, someone I guess illegally leaked this information about the spy because it was classified. So someone broke a law to leak the information to was it originally a New York Times or Wall Street Journal, whoever did the initial post, right? Then there's obviously some speculation on if I think there's a lot of skepticism, but it also could be. And obviously that wouldn't be good. And then I did hear a lot of talk on why Jensen would resign if the investigation hasn't been completed by the board. And so some of it would be if you're a public company, you have to not have conflicts of interest and potentially sleeping with a customer. Which was, you know, whether a spy or not, is a conflict of interest and if you're on the board, you need to kind of protect yourself, right? You have insurance if you're on the board of these public companies. You need to protect yourself. So there's some speculation that the board is kind of forced to ask Jensen to step aside because they're potentially at risk if Jensen was having some kind of a relationship with a customer that purchased a lot of NVIDIA products.

02:12

Abhi: And they did. Right. So.

02:13

Shane: And they did. And so then this came out over the weekend. There's a petition to reinstate Jensen. The argument here is that the board is trying to protect themselves. They want to limit their liability. So by getting rid of Jensen, they're doing the right thing for the board, protecting themselves, not necessarily doing the right thing in the eyes of the shareholders. And so the goal here, I guess, is to get enough people to where it can go to a shareholder vote to bring Jensen back. And they got 150,000 signatures. A lot of those are shareholders, not all of them I imagine. And of course then you got the prediction markets trying to predict if Jensen will return. And I've seen it, I've seen some different markets go all the way up to 15%. Right now I think it's like eight to nine percent that Jensen will return as CEO.

02:57

Abhi: Dude, I signed that petition. There's like a bunch of employees on there too. Huge percentage of employees. I was also hearing that President Trump was saying, like, hey, we need this, you know.

03:08

Shane: Yeah, I heard that 90%, and I don't know if this is accurate or where this data came from, so take it with a grain of salt. 90% of employees favored or were in support of Jensen returning. So I think he was relatively well liked as a leader. Whether you agree with, you know, the mistakes he made or not, I think people generally that worked for him seemed to think he was a good leader and seem to really want to continue working for him.

03:30

Abhi: Yeah. There's a shock though because of how he runs the company where it's like very through him, let's say, it's gonna hurt a bit if they don't figure it out.

03:38

Shane: Yeah, and I think it kind of hurts the United States a bit. Because if NVIDIA slows down, let's say they have all this momentum, they slow down. Well, of course, other US companies could step up, and I think some will, you know, OpenAI, Anthropic, Google all working on their own chips. But also, you know, maybe China starts to catch up a little bit too. They're also working on their own chips. So that's the risk. And I think if you look at NVIDIA, they were basically kind of propping up the economy a bit because they were funding all these deals to allow Anthropic or whatever to purchase these chips. And basically kind of like moving money around to like make sure that these data centers can be built, these chips can be allocated, and the question is will the next CEO be as bold as Jensen and take as many risks? It is very challenging if you're not the founder to take big risks when you're trying to look at what's in the best interest of shareholders. Oftentimes you get, you know, if you think about going from Apple, Steve Jobs to Tim Cook, right? There's kind of a you know, a little bit more risk averse, but still did a great job, arguably, but didn't maybe innovate the same way that if you had the founder at the helm you probably would. So I think that's the risk that a lot of people are starting to weigh in and the and the markets are responding to that, which is why NVIDIA stock has kind of tanked a bit.

04:49

Abhi: Well, I mean the market has tanked a bit, right? So.

04:52

Shane: The market in general, but NVIDIA even before the the market really started to take a hit. Yeah. It's a bad day to own stocks.

04:58

Abhi: Hey, if NVIDIA goes under, you guys might see me under a freeway just letting you guys know.

05:03

Shane: And then this came out, Jensen is apparently discussing returning and so maybe trying to figure out what he would have to do or, you know, how that might work. It but it sounds like he's maybe starting to think through his options and try to make a play to return back you know, it is funny because of course, you know, you had the whole Sam Altman thing, which OpenAI wasn't a public company, so a little bit different set of regulations, requirements, all that. But the whole Sam Altman thing where he was ousted by the board and then employees kind of like rose up, brought him back in. Maybe the same thing happens with NVIDIA. It'd be a certainly a different scale.

05:39

Abhi: Yeah, it's like the whole world scale.

05:41

Shane: Yeah, but pretty crazy nonetheless. So what do you think? Is it coming back?

05:45

Abhi: I really hope he comes back for my own stock portfolio sake, but like it's just such a weird... It's just a strange case. I also was watching All-In. There's always a conspiracy with a lot of stuff, right? And so politics comes into play again. It's just fucked up.

06:02

Shane: Yeah, I mean there's a lot of speculation that it was like a hit on Jensen because people didn't like the way he was pro AI for everyone, pro open source. I don't know. They said, you know, it could be political, could be motivated by others that are less pro you know, AI for all. I don't know. I don't know if I believe any of that, but you have people coming out and there's always conspiracy theories. They're entertaining. I just, you know, I don't I don't put a lot of stock into them until we see some proof but I do think overall I don't necessarily agree with his decisions, but I do think that there's probably no one I would rather have still being CEO of NVIDIA than Jensen. So That's the confliction that I would be in is I would probably want Jensen back too. I own a little bit of NVIDIA stock, not a lot, but enough to where, you know, I'd like to have him back.

06:47

Abhi: Just shows you when you're a CEO, you can't really or when you're in a position of power like that, you really can't slip up, dude.

06:53

Shane: No. It's much easier if you just take the ethical and moral high ground and and not not slip up. And and just as a caveat for those watching, we're talking a lot about stocks. Nothing that we say is financial advice. So no financial advice here. We're just bullshitting. We're having fun. Alright, let's talk about new models. So, meet Gemini 4 Hydrogen. This was from November 12th. This is from Google DeepMind. It's built for rapid coding, multimodal synthesis, and everyday workflows. It finishes tasks 2.5x faster than 3.8 Flash and surpasses Argon on the Artificial Analysis Intelligence Index. Interesting. So if you want to look at some of the benchmarks, you can see it's compared to Sonnet, it's compared to Terra.

07:39

Abhi: Over the weekend I tried Gemini Argon and it was pretty good. I'm not I was not doing coding work. I was just doing some reasoning type stuff.

07:46

Shane: You know, someone did come out and say Sol is smarter and cheaper. Now I think this is a chart error because if you look at like Hydrogen compared to Sol, it's only like a point difference in the Artificial Analysis Index, but it looks huge. But the cost difference is pretty striking. Right? So you'd argue that why won't you just use Sol instead of Hydrogen? But it's only a point off. Obviously the cost is one thing though. I am curious, what's with the naming. I gotta like know the periodic table now to know which one's the bigger model. Is that the is that the thing? So is Argon the biggest? I had to look it up. I think it's like the 16th element. Yeah carbon, which is like six, right, or something. And then you got hydrogen which is one. So hydro hydrogen, then carbon, then argon?

08:25

Abhi: Everyone has a stick, you know. They all gotta have their stick. Flash was good, you know, but let's say Gemini 4, who like, no one's like no one cares about that shit. But now we have Gemini Argon and Hydrogen. It sounds fancy. It sounds good.

08:39

Shane: Yeah, it's like I need to I'm gonna get one of those periodic table shirts so I can look down and know which model's bigger because.

08:44

Abhi: Yeah. But also separates the model class from whatever we were talking shit about before. Yeah I mean, rebranded.

08:51

Shane: Yeah, it's a little confusing, but I suppose over time everyone's got their thing. So this came out yesterday. Meet Qwen4-Max, your new AI teammate for work, creativity, and discovery. Qwen4-Max is built on a new mixture of experts architecture with native vision and a 1 million token context window. So let's take a look at some of the benchmarks here. You can see for coding, you can see you know a lot of different benchmarks. They're more like work-based benchmarks. You have HealthBench, you have Humanity's Last Exam. It looks like it's comparing it to Qwen3.8 Max, which is the previous one. It obviously scores better in every benchmark. Compared to the old model, which makes sense. And it compares Opus 5.5, GPT-6.1 Sol, GPT-6 Astra. And it is looking at it, good at a few of the tasks, right? DeepSWE, it's apparently number one now. Terminal-Bench, it's a little bit further behind. You know, some things like ProgramBench, which I'm not that familiar with. It's a little bit behind. So at some benchmarks, it's very competitive though. BioMysteryBench. It's apparently top of the charts. So.

09:57

Abhi: Dang.

09:58

Shane: There you go. If you're doing bioinformatics research. I haven't tested it yet. It just came out. I'm always interested to see kind of how these things land. But it's good to see more open weight models continuing to come out and still be competitive.

10:11

Abhi: Now that we're paying for tokens, gotta figure out what the right token mixture is, you know.

10:17

Shane: Yeah, turns out when you have a factory that just burns tokens, subscriptions don't go very far.

10:21

Abhi: No. You gotta.

10:22

Shane: You gotta test out open weight models. Keep things reasonable.

10:25

Abhi: We are DeepSeek users now because of the press.

10:28

Shane: We are DeepSeeking our way to victory.

10:30

Sam: Three editions, one year. The agent world moves fast, but we've got you covered. 2026 is all about coding agents, so we've got skills. Sandboxes and file systems. Get the third edition now at mastra.ai.

10:46

Shane: I wanted to talk a little bit because I saw a few posts throughout this week, but can agents actually learn? So let's look at a couple things. The first is a post from Daniel Lenton. It says OpenClaw, Hermes Agent, and Prime Agent don't meaningfully learn from experience. On Continual-ARC, our new benchmark at @letsunifyai, none of them were able to memorize anything of value, and none were able to generalize to new tasks. Training a small 9B model's weights did both. So the argument here is that the memory systems in these agents couldn't actually learn. It maybe memorized like certain facts or whatever, but it couldn't actually learn and like discernment and the benchmark basically has a set of puzzles. You know, it was trying to like see if it would actually learn it. I guess did you have a chance to look at this?

11:30

Abhi: Yeah, I have a little thing I can share. Professor Abhi now. I'm gonna become Professor Abhi. So this is this article is super relevant to us at Mastra, so I wanted to learn more about it. Because we're doing agent learning as a company. And so I first was thinking like how do agents learn? And learning is like obviously a very loaded term, but first you like learn in the weights, which is like fine tuning, you know, continual learning. Or you could learn through context where you edit your prompts and you have better skills and better memory and you have knowledge primitives and all that stuff that like make your agent more efficient with context. Or you like do you have to learn in the system where you know your tools fucking suck your workflows like do not actually help you. You have terrible retrieval metrics. Like you gotta learn all these things about all these different levels. And you can improve your agent in all the different levels for sure. Like I don't think Especially if you don't know how to learn the weights, you could totally make really good progress just doing the other stuff, right? But when you're talking about the weights, like that's like this, you know, if feels to me like this ambiguous thing that's going on that none of us know how to do. And I want to change that, like just personally. I feel like I want to change that for myself. But There's this thing called ARC, which means like Abstraction and Reasoning Corpus, where can an LLM reason about unfamiliar problems rather than like doing pattern-based matching, which you know, a lot of our a lot of these learning in the context is about patterns. Like so it does the right thing based on what it did before and you kind of like nudge it to do that. In ARC, right, like the problem is completely new and so it needs to figure out what to do about it. So they do it through these puzzles. These are ARC puzzles. And you know, you give it an input and then it tries to do an output. And that's all cool. Same kind of puzzles. I think they look really cool. They're all visuals, you know, they're like visual puzzles and that was dope. So I looked into this Continual-ARC thing. And it's kind of the same concept except when an AI encounters the same kind of problem, did it learn from it last time? And then how do we actually how can it actually progress through this problem like in this current the current iteration of the problem? So it has many steps. First step, you encounter a puzzle. So you give it the same puzzle. And there's this the interesting thing is like how what it does next. Next, it can pay for feedback. I really thought that was an interesting term. Pay for feedback. What does that mean? You can actually do a bunch of shit. One, you just solve it and you have no feedback. So how this works for paying is the run has a budget for what it can do to learn in the run. So first you pay nothing, solve it. You just say fuck it, whatever. If you get it wrong, you get minus one point. So like solving and getting good points is a good and is like is encouraged There's like this teacher, so you can ask for past examples, and that costs like let's just say that costs like two points. So you have to reason are you willing to pay two points to get the example to fix the problem? Or are you gonna submit an incorrect answer to cost yourself point to move on? Or can you bank this later as a similar problem that you want to reference when you know when you encounter it again? I just think it's quite fascinating to pay for feedback. Once you do that you either get the information, submit the incorrect answer, find out what it's right, whatever, and then you save all that stuff, and then you update the model weights, and then you do it again. And if you encounter the task again, if you've already seen it because you paid for it, can you do it again? Can you build a skill that does it the next time so you always do this the right way? If you haven't seen it, then you go through the flow again. All right, finishing up here. So ARC versus Continual-ARC, ARC is the puzzle, like it's they're both puzzles, but in Continual-ARC you're having a sequence of puzzles that you learn over time. In ARC you measure things on individual tasks, but Continual-ARC is like a long horizon, like over many tasks, how are how's the retention? And then, you know, do you get better with experience? Like this is a continual learning type of setup. So yeah, that's Continual-ARC and that's the thing for today.

15:36

Shane: Yeah, and I think it was very interesting that, you know, if you looked at OpenClaw, Hermes Agent, Prime Agent, they didn't really learn, right? But if you just trained a much smaller model, and they I think they used a bigger model to like do the training to like help with the training, like to pull out the examples. But then they trained it and then it was able to actually learn. I didn't catch that if it outperformed the much larger models, but it you could prove that it provably learned and got significantly better.

16:06

Abhi: Sound like real to me.

16:08

Shane: It's kinda like when they solve math problems. It's just like it could be very important potentially, but also not like super practical for the average person day to day. It's not your super practical task.

16:18

Abhi: Yeah. And I still don't understand like once you do a run of ARC, how do you change weights and stuff? I guess we'll learn with everybody else as we go.

16:27

Shane: So this is from Jacob Coxon who was previously at Anthropic, had a bunch of drama at one point for leaving, says that this slide, which was from the Anthropic IPO roadshow, has kind of been making the rounds lately. All this slide is showing is that it's getting easier to train models. It's taking less compute. So they're getting more efficient, significantly more efficient. And this is like a log scale, so it's actually like super exponentially more efficient at training models. Which in a lot of ways is a good thing 'cause it means it's cheaper to train larger models than it was, you know, in October than it was in two years ago by a large scale. And it seems to be accelerating, not, you know It's not just it's like accelerating, but it's also like super accelerating, 'cause this is like a log scale. So it's like very, very exponential. In my opinion, that that's a good thing. It means that Models should be able to get better faster over time. But obviously, if you know some people think this is a concerning thing as well. Let's get into the last part, the quick hits. LangChain raised $300 million at a $3 billion valuation. Nice job, dudes. Congrats.

17:29

Abhi: Makes me happy.

17:30

Shane: Today we're introducing UI SDK. So this is from shadcn. Agents are becoming both the builders and the users of interfaces, so we asked ourselves, where should the abstraction live? Take a Select. There's a pop-up selection logic, keyboard navigation, and focus management, those same behaviors are useful in lots of other components. So we pulled these pieces apart, tested them, and made them reusable. The behavior stays the same whichever framework, renderer you build with or the harness generating the UI. That's the idea behind UI SDK. Give agents the pieces to generate the interface you need. Right when you need it, even interfaces we haven't thought of yet.

18:04

Abhi: They're getting into the CPK territory.

18:06

Shane: Yeah. I mean I think this just the whole idea around Yeah, how UI is changing. Of course, like generative UI I think is gonna become more prevalent. I'm really bullish on MCP apps, it's just like the UI now lives in your whatever. App you're using to chat with your agent, whether it's you know Codex, ChatGPT, whatever, Claude, Grok Bot, you can actually bring your UI to users. I really think that those applications are slowly becoming the new browsers for people, for some folks. Like they're not opening Google, they're just opening up their desktop app and so I think bringing UI to those places is very interesting. And so just like the whole concept around UI and how it's changing over time is very intriguing to me.

18:47

Abhi: Like I feel like MCP apps, I think we should have apps in the protocols and not just MCP. Like if I'm in a TUI, I should be able to render Whatever, but in TUI form, you know.

18:58

Shane: So you think every interface needs its own protocol?

19:01

Abhi: I think for every agent it should have the ability to generate what we're calling apps in MCP, but it's like just part of its protocol. That's probably the future in my mind.

19:12

Shane: All right. Interesting. Alright, let's continue on. Eve 1.0 apparently is going to be released soon too, so That's big news for those of you that are building agents and want to use Eve and not use Mastra for some reason. Congrats to them. Getting to 1.0 is obviously a big milestone. Dude, that's the show. That was quick. Follow us on X at Mastra. Go to YouTube, mastra-ai. You can follow me on X @smthomas3. Follow Abhi at @abhiaiyer. What do you have for us? Parting words of wisdom to leave?

19:41

Abhi: Everyone have a great week. Stay safe with your CVEs and security. Peace.

19:46

Shane: See ya.