¶ The Amazon Outage
Ejaaz: Last week, Amazon's entire platform crashed for six hours. No one could shop, Ejaaz: buy anything, they couldn't even see prices. Ejaaz: The reason was because a junior developer had submitted an AI-generated piece Ejaaz: of code which crashed the entire platform, and it cost them millions and millions of dollars.
Ejaaz: Now, Anthropic, the creators of Claude Code, which is what Amazon was using Ejaaz: to create AI-generated code, Ejaaz: also had a similar issue where their entire platform has been suffering from Ejaaz: outages this entire week. Ejaaz: They actually also released a new product called Code Review, Ejaaz: which helps use AI to help fix the code problems that their own model is fixing. Ejaaz: It's all getting incredibly complex right now. And Amazon tried to hide the
Ejaaz: entire thing from a Financial Times reporter. It's all pretty crazy. Ejaaz: And it forces us to answer the new narrative, which is AI-generated code isn't going anyway. Ejaaz: Demand is at an all-time high, but have we hit a wall? Has it become too dangerous to use AI to code? Josh: Yeah, and there's this interesting phenomenon happening. As these AI coding Josh: abilities in general become better, where currently they account for, Josh: what, about 4% of total GitHub commits.
¶ Growing Pains
Josh: The expectation is at the end of the year, they will account for 20% of total GitHub commits. Josh: And there is this increasing reliance on these AI tools, but that creates these Josh: key choke points and points of failure that have a significant effect. Josh: I mean, And this Amazon outage was a huge deal that was pushed by a single person. Josh: And then within Claude Code, they've been having a lot of downtime.
Josh: And the Anthropic, I mean, we were trying to use the programs today and it was Josh: down. The servers weren't working quite right. So there's a lot of these growing Josh: pains that are happening. Josh: And it seems like we're running into more issues faster than people are trying to mitigate it. Josh: So that's where I suspect this Claude Code checking agent comes in that we're Josh: going to cover later. But this Amazon story was pretty fascinating.
Josh: Like this cost Amazon billions of dollars. Ejaaz: Yeah, so let me walk you through the timeline for this, because this actually Ejaaz: isn't the first outage Amazon's experience because of an AI-generated piece of code. Ejaaz: So back in early December, actually November of 2025, they took a really hard-line Ejaaz: approach and a new policy was invoked, which was... Ejaaz: I want 80% of all Amazon's code generated to be AI generated.
Ejaaz: And this was their goal to be achieved by the end of 2026. Ejaaz: Now, this flips, completely flips from a company that I think employs like hundreds Ejaaz: of thousands of engineers and wants them all to kind of like handwrite or hand type the code. Ejaaz: So this is a pretty aggressive flip. Amazon's been laying off tens of thousands Ejaaz: of people. So this kind of is in trend with what they wanted to do. Ejaaz: In the middle of December 2025, they experienced their first outage.
Ejaaz: It was 13 hours back then. Ejaaz: So this is all adding up, by the way, we're talking about 10s of millions of Ejaaz: dollars, then in late 2025. So at the end of December, there was another outage. Ejaaz: And then we have the outages that we're speaking about today. Ejaaz: And the issue that this flags is, although AI is like a really useful tool to Ejaaz: generate code, and it finds bugs, it actually might create more problems than you'd expected.
Ejaaz: Because the issue that they're seeing is junior developers that come in that Ejaaz: don't understand Amazon's code base, just kind of use AI to run like code to Ejaaz: run autonomously, figure out what they want to create, and then they just submit Ejaaz: it without actually reviewing and understanding it. Ejaaz: And if this goes unguarded, it creates and results in issues like this.
Josh: Yeah. And there's a problem that's starting to happen now where agents are creating Josh: code far faster than humans can keep up and check it. Josh: So it becomes this impossibility where if you want to move at the velocity that Josh: AI enables, you are simply unable to keep up with the change log of what's happening. Josh: So you have to defer some sort of trust level to this AI, to its ability to Josh: check itself, to run tests, to verify that it actually works.
Josh: And in some cases, it doesn't. I mean, I'm sure some people listening to this Josh: experienced the outage. Josh: And it wasn't just an AWS outage. This is Amazon, the actual storefront, Josh: where my dad actually texted me. Josh: He was like, am I doing something wrong? Can you place this order for me? Josh: Because it's not going through. Josh: And then I went to check myself and the whole service was down. I went on Twitter.
Josh: I saw Casey Neistat was posting a bunch of photos about this. It Josh: was this big deal because anyone who was trying to order anything from Josh: amazon the entire storefront was just totally offline so Josh: aws the web services runs a significant percentage Josh: of the internet amazon the storefront runs a significant percentage of the e-commerce Josh: and these things have been going down at an increasing rate due to these tools
Josh: like cloud code so they're moving much faster but they're they're breaking things Josh: this is like the early zuck facebook mantra it's like move fast and break things Josh: they're now taking it to the extreme and Josh: As we become increasingly reliant on these tools, we might see this start to, Josh: I mean, permeate even further than just Amazon.
Ejaaz: Well, the weirdest part about this is that AI kind of has moved from this assistive Ejaaz: tool to now being the foundational bedrock for a lot of these different like Ejaaz: services that you've just mentioned. Like AWS runs the entire internet. Ejaaz: Every single business that you interact with online probably uses AWS on the Ejaaz: back end. So if they go down, then your business goes down as well. Ejaaz: That's where we get all these outages.
Ejaaz: Amazon had some pretty severe repercussions. They have now done a complete 180 Ejaaz: on their policy of generating 80% of their code via AI. Ejaaz: And they've now said, if you're a junior developer or engineer working at Amazon, Ejaaz: you now need to get your manager's permission to submit code. Ejaaz: So this is so tough because you've now gone from trying to automate the entire Ejaaz: thing to now making it incredibly worse.
Ejaaz: So you're using AI to kind of generate the code, but then you still have to Ejaaz: rely on a human constraint. and there are fewer managers than they are like junior developers. Ejaaz: So it all just becomes really bogged up as a pipeline to kind of shipping changes.
¶ The Job Market
Ejaaz: So I think this is going to slow down Amazon massively. Ejaaz: But if it's for the result of, you know, saving your company tens of millions Ejaaz: of dollars, fair play. But yeah, it's a pretty hard line approach. Josh: And this all seems kind of connected to the trends that we've been seeing, Josh: right? Particularly around jobs and distribution. Josh: It's like a lot of people are cutting jobs, which means they're increasing the
Josh: reliance on these AI tools. But if the result is that the actual output of these Josh: ai tools is causing a lot of damage it creates this this tension and Josh: And it seems like the AI labs themselves, Anthropic, OpenAI, Josh: they cannot hire talented people fast enough. Josh: They are looking to hire every engineer under the sun who is competent and capable Josh: of building great products.
Josh: But a lot of other companies are deferring that workload to these AI labs, Josh: which is starting to create this interesting dynamic where the job market is Josh: suffering, but the actual productive output of these companies is not to an extent. Josh: And that's operating under the assumption that these code tools will get better.
Josh: But if Amazon and a lot of other companies start putting in these thresholds Josh: that, again, bottleneck and throttle that amount of AI production capability, Josh: like that probably doesn't help the cause a whole lot. Ejaaz: There's also the angle that AWS or just data centers in general are becoming Ejaaz: like a really valuable or national security threat level asset, right?
Ejaaz: So if we look at the current Iran conflict that's being waged between the US Ejaaz: and Iran, Iran actually targeted strikes specifically to Amazon's AWS centers Ejaaz: out in the UAE and Bahrain. Ejaaz: And this caused a bunch of the outages as well. So it's not even just like AI coding level threats.
¶ Geopolitics
Ejaaz: This is becoming a geopolitical weapon at this point.
Josh: It makes sense. Every single story we tell, like this Game of Thrones narrative Josh: that's kind of existed throughout the history of Limitless, Josh: it's just elevating itself higher and higher to the global stage, Josh: where now it's impossible to have any sort of conflict or Josh: large decision without ai being in the middle of it like when you think Josh: about this conflict what are the bigger things it's this like anthropic
Josh: deal with the pentagon they're now striking the aws data centers it's coming Josh: for the infrastructure that's building these tools that are so powerful but Josh: this isn't just happening on the worldwide stage this is happening on the consumer Josh: and commercial level too you were mentioning something about lovable and how Josh: much they've grown recently could you just explain lovable for the people who
Josh: aren't familiar and what's what how big they have grown recently it's crazy Ejaaz: Okay, so typically the people who code, this might shock some of you, Ejaaz: are the people who have learned to code. Ejaaz: They're very technically focused work. But there are people like you and I and Ejaaz: a bunch of our friends that don't necessarily know how to code, Ejaaz: but still want to do that. Ejaaz: And that's what resulted in Vibe coding, right?
Ejaaz: Now, Claude Code and products like Cursor are still very oriented for technical folks. Ejaaz: A platform like Lovable is for people who have zero experience coding, Ejaaz: but still want to code things. So the Lovable platform is actually a really Ejaaz: cool platform where you just type whatever you want. Ejaaz: Like I want to create an app that tracks my fitness reps or whatever that might Ejaaz: be or create or suggest cooking recipes for my type of cuisine.
Ejaaz: And it does so in a couple of minutes. Ejaaz: And it's really intuitive. It makes it super easy. Ejaaz: And they have the additional perk of being able to deploy it live as a website Ejaaz: or an app that you can publish on the app store super easily. Ejaaz: So they have this end-to-end experience. Ejaaz: Now, Lovable, when it started, had like a rocket-type trajectory. Ejaaz: Within the first 12 months, I think they hit $200 million ARR,
Ejaaz: which is just like an insane growth. All of these AI companies are blowing up. Ejaaz: Now, in the last month alone, they've added an extra $100 million of AR. Ejaaz: They were at $300 million. Now they're at $400 million. Ejaaz: So that's like roughly like just over 30% growth. Ejaaz: Just insane amount of demand. And that's the point I want to make.
Ejaaz: Like when I look at these Amazon outages and Amazon restricting their developers Ejaaz: from using Vibe Coding to improve their product, I don't think this is necessarily Ejaaz: a wave that they can stop. Ejaaz: So they should stop trying to implement policies that are restricting people Ejaaz: from using it and instead try to figure out what is a better way to implement Ejaaz: AI-generated code because this thing's not going away.
Ejaaz: Lovable, Cursor, Claw code are all skyrocketing and we're not going to stop that.
¶ Rise of Lovable
Josh: And I think a lot of the, I mean, aside of Claude Code, right, is Lovable and Cursor. Josh: They're kind of model agnostic. They'll use whatever model you put. Josh: So I was looking at Polymarket earlier today because they have markets for which Josh: one of these models is actually going to be the best in which time. Josh: And this is interesting because if you looked at these charts towards the middle Josh: of January, you'd notice that Anthropic was kind of the favorite.
Josh: Anthropic was looking like they were going to have the best model. Josh: But the reality now is that by March 31st, there's a very clear divide that Josh: has happened over the last six to eight weeks. in OpenAI having an 85% chance Josh: of having the best model. Josh: Does that mean there's going to be something new or is that just the current model? We don't know.
Josh: Currently they're at 5.4, which is fantastic. I think everyone is kind of unanimously Josh: decided that it's the best for coding. Josh: But then there's also a fun following market here about Claude going down because Josh: it appears as if Claude is really just not doing as good as it needs to be. Josh: It's not as stable. We were trying to use it this morning. It wasn't working.
Josh: They're clearly experiencing growing pains because they have all these new users Josh: and they're having to throttle reasoning. Josh: They're having to limit the amount of searches that people are able to use. Josh: I'm sure the researchers who use these to train the models are not happy. Josh: And it looks like based on Polymarket, we'll be seeing somewhere between five to nine more outages in Ejaaz: The month of March. Yeah, basically every single day this month.
Josh: Yeah, like it seems like they're having problems. And this is proof. Josh: I mean, you can actually bet on the problems that they're having and make some Josh: money on it, right? This is what, $26,000 of volume. Josh: So thank you, Polymarket, for sponsoring this segment. And I guess now we can Josh: get onto the solution, which is what Anthropic is starting to roll out.
¶ Anthropic's Code Review
Josh: But they're doing it in a way that's a little less than desirable. Ejaaz: Yeah. So the main instigators of this entire AI coding debacle is Anthropic. Ejaaz: They created Claude Code. Ejaaz: Amazon owns 21% of Anthropic, which is just crazy to say out loud. Ejaaz: And so they use Anthropic's model, Claude Code, to do all their AI code generation.
Ejaaz: But Anthropic decided, hmm, instead of trying to make our AI coding models better, Ejaaz: let's just release a new product called code review, which will review your AI generated code. Ejaaz: So you now have Claude generating the AI code and then also reviewing the code Ejaaz: as a separate product that you pay for. Ejaaz: And it's all for the beautiful price of, I think it's like 15 to 20 bucks. Where is it? Yeah. Ejaaz: 15 to 20 bucks per review, which some might say is quite expensive.
Josh: And they scale based on the pull request complexity. So that number can get higher. Josh: And what I find it funny is they're selling the problem where they're selling Josh: you a model, you use it to code, it's going to create problems. Josh: And then they're selling you a separate package for the solution where, Josh: oh, you have problems in our code.
Josh: Well, here for another $25, $50, whatever it may be, we'll actually do a code Josh: review and we'll fix the problems that we've created. Josh: And that creates this uniquely disincentivized incentive where there's no real Josh: reason for Claude to want to fix problems when they could just sell this package Josh: on top to remove the problem.
Josh: And this comes in the face of chat gpt Josh: and open ai doing the polar opposite where they're actually offering these Josh: for free and here's rohan from open ai if you Josh: want ai code review but don't want to pay 25 per review check Josh: out codex review it leverages frontier codex models finds complex issues and Josh: 100 usage based most runs should cost one dollar or less so here we have another Josh: head-to-head collision happening where chat gpt open ai is going head-to-head
Josh: with cloud code and And it seems like Cloud Code is the favorite for a lot of people, Josh: but I've noticed a lot of the more technical people on my timeline who are building Josh: pretty hardcore things have been really relying on 5.4 and using codecs much more. Ejaaz: Yeah, so to set the precedence, OpenAI and Anthropic now are neck and neck at Ejaaz: building the best coding models.
Ejaaz: And Codex 5.4 from OpenAI has taken the lead. A lot of engineers are now saying Ejaaz: it's way better than Opus 4.6, as you just mentioned. Ejaaz: In terms of these security code review products that each company offers, Ejaaz: OpenAI is instinctively cheaper. Ejaaz: If you have a $20 a month subscription, you now get access to this code review Ejaaz: tool for no extra cost. So it's automatically a better tool to use.
Ejaaz: Now, in terms of quality, in terms of how good it is at code review, Ejaaz: we don't know because they haven't made any of their datasets public. Ejaaz: Anthropic has, and it's pretty damn good. Ejaaz: Also, if you compare Anthropic's product to traditional SaaS vendors that have Ejaaz: these kind of AppSec security tools, it is still way cheaper. Ejaaz: A lot of people were in my comments, actually, because I tweeted about this Ejaaz: saying it was cheaper. They were like, no, it's not.
Ejaaz: I could use this SaaS tool. And I'm like, yeah, but you're not factoring in Ejaaz: the engineering hours, they spend like 30 to 60 minutes per bug. Ejaaz: That adds up to about $100. Ejaaz: So it's still technically cheaper, but I would rather use OpenAI's tool, Ejaaz: which is just way better.
¶ OpenAI vs. Anthropic
Ejaaz: Now, Anthropic itself are facing so many outages for Claude. Ejaaz: You mentioned earlier that, you know, Claude is out for you this morning. Ejaaz: It's also out for me this morning. And I think this is because OpenAI basically Ejaaz: has all the money in compute to subsidize this cost for a tool like this, Ejaaz: whereas Anthropic is struggling. Ejaaz: They're adding a million users Ejaaz: to their general database or to the general user base every single day.
Ejaaz: And they're not able to subsidize people's compute anymore. So they're just Ejaaz: shutting people's access off to it. And it's kind of frustrating to see, to be honest. Josh: Yeah, and the way you have to think about this is each one of these companies Josh: has a limited amount of compute. Josh: And that compute takes care of everything. It has to serve the customers. Josh: It has to serve the developers, the researchers.
Josh: And a lot of times when you're developing these new technologies or you're training Josh: new models, the training run consumes a tremendous amount of GPU power. Josh: The researchers want a lot of GPUs to run tests and run trials on, Josh: but you still have to find extra resources to serve all of the users who are querying this model.
Josh: Every single day. And I was recently listening to Dario on, I think, Josh: the Dwarkesh podcast, where he was talking about how they think about how many Josh: GPUs to order into the future. Josh: Because they go to Jensen, they go to NVIDIA, they say, hi, Jensen, Josh: we would like to place this amount of GPU purchases for this year.
Josh: And what he was mentioning that I found interesting is that they can't over Josh: order because if they're off by just a small percentage, the incremental cost Josh: of those GPUs will far outweigh the growth, the revenue that they get from growth. Josh: And what it sounded like he did is he was just taking the current growth and Josh: mapping it to the future to hedge themselves against this collapse from debt Josh: that they would have from over-ordering.
Josh: But the reality is, is that since that episode was recorded and now they have Josh: gone fully vertical, like that curve has steepened significantly more and they're Josh: going to have to figure out a way to solve this because this is not an easy Josh: thing. These orders get placed years in advance.
¶ Self-Improving AI
Josh: They're projecting for a certain curve and they're getting a different one. Josh: So is this a short-term thing? Josh: Is this going to be durable? I don't know, but it is noteworthy that Anthropoc Josh: is having some growing pains. Ejaaz: Okay, so we have two problems here. We have an amazing AI model that can code, Ejaaz: and that is inevitably going to be the future. Lovable's demand proves that.
Ejaaz: But then we're also running into the issue where we have all these security Ejaaz: flaws and we could lose tens to hundreds of millions of dollars. Ejaaz: Amazon demonstrated that, right? So we're at this conundrum. Ejaaz: How on earth do we solve this? Do you know what the solution might be, Josh? Ejaaz: It might be AI. It might be AI itself. Josh: That checks out. Ejaaz: You have to use AI to solve both of these individual AI problems, I'm afraid.
Ejaaz: So, Andre Carpathy, the godfather of AI, as I like to call him, Ejaaz: released this really cool experiment that is just completely blown up, Ejaaz: which hints at what the future solution might be to solving this problem that Ejaaz: we just explained on this episode, which is called auto-research. Ejaaz: Now, the best way to think about this is he created an AI model that acts as an AI researcher. Ejaaz: But the coolest part about it is it autonomously does research.
Ejaaz: Now, typically, when you use an LLM, you need to prompt it. And then it says, Ejaaz: I don't really know what this means. Ejaaz: Can you tell me what it means? And you have to keep working back and forth. Ejaaz: This thing runs completely on its own overnight, and in some cases, for days at a time. Ejaaz: He's done this overnight when he slept for like 9 to 12 hours, Ejaaz: and then he's done it again for two to three days.
Ejaaz: And how the AI model works is he sets an objective. He says, Ejaaz: I want you to try and improve this part of your model. Ejaaz: So he's talking to the AI model and saying, you need to improve yourself in this particular way. Ejaaz: Go away and figure it out. The model then runs an experiment every five minutes. Ejaaz: If the experiment comes out with an improvement, it caches it in. Ejaaz: It improves its own model weight. Ejaaz: It improves its own internal insights, right?
Ejaaz: If it doesn't, it discards it Ejaaz: and it tries again. And it runs these experiments on and on and on and on. Ejaaz: In this experiment, I think he ran around 150 experiments overnight. Ejaaz: And it improved itself pretty marginally, as is demonstrated by this graph. Ejaaz: The loss aversion just went crashingly down.
Ejaaz: So it suggests that maybe Anthropic and Open Air, instead of having to rely Ejaaz: on releasing new tools every day, constrained by humans, they could just let Ejaaz: the AI improve itself and everyone's hunky-dory, maybe.
Josh: This seems like an early form of this takeoff situation in which AIs become Josh: self-improving, because I haven't really seen this architecture before in such a Josh: open and easily accessible way i mean this is Josh: all open source this i mean andre says here it's 630 lines Josh: of code and traditionally it's on the human to iterate on Josh: the prompt to improve the prompt to improve the Josh: training code and to work together with the ai but this actually
Josh: does it completely and entirely on its own it goes Josh: off it runs its own experiments it ingests those experiments updates Josh: its values that it has updates its view on the world and then creates a new Josh: set of experiments and runs this iterative process and what we saw in that chart Josh: is that there was these incremental improvements that andre just didn't see Josh: and again it's hard to overstate how much of a powerhouse andre is in terms
Josh: of ai engineering and understanding do Ejaaz: He has two decades of experience in.
Josh: This and like ai what like wasn't even Josh: around two decades ago like he was the guy who was building the damn thing Josh: yeah um so for him to come along and for Josh: the first time ever experience this novel breakthrough in which an ai Josh: isn't able to improve his code in ways that he didn't initially see Josh: it's like when ai played go for the first Josh: time or when ai is playing chess and it makes these erratic moves you would
Josh: not have known that it was optimal but the model is Josh: is so precise has so much context and so Josh: much intelligence that it's actually able to make these optimizations that Josh: humans otherwise wouldn't have seen and this is one of the earlier instances where Josh: even andre got humbled and the code that he Josh: was running was significantly improved overnight by this Josh: self-recursive thing and the most interesting thing about this is it's
Josh: open source it's available to run on your laptop on Josh: your pc on a computer right now and i Josh: have to imagine that ai labs are going to take this and run with Josh: it at scale where if you're training a model why would you not and and this Josh: model using this auto research can do better than andre i'm sure it could do Josh: better than the average mid-tier like prompt engineer at anthropic or chat or
Josh: open ai and actually improve the entirety of the company so this feels like something that is Josh: starting from the bottom where in terms of it's like openly accessible it's Josh: open source and then we'll move its way up through everything and become this Josh: like auto-aggressive self-improving loop it's incredible i want to try it on Josh: something we need a problem that's hard enough to try this on where you can
Josh: truly let it run for a series of days and then come back with the answer that's Josh: better than anything you could have ever imagined well maybe research Ejaaz: On how to make limitless the best and number one podcast in the world maybe i i'd be down to run this.
Josh: That's probably we should try that we should try that we can make some assumptions Josh: on what it would it would say right It would say, well, we probably need to Josh: convince all of our listeners to go subscribe to the podcast on your favorite podcast player. Josh: Give it five stars, share it with all of their friends, and then make sure that Josh: they comment things they either like or don't like about it to keep the engagement Josh: rate high. Those things will help out the show.
Ejaaz: That actually sounds pretty great. And if you're listening to this, Ejaaz: remember, you are no more than organic LLM. Ejaaz: So if you want to process that prompt that Josh just specified, Ejaaz: please go do it. It helps us out a massive amount.
¶ Conundrums
Ejaaz: But just to round up and wrap up this episode, it's this weird conundrum. Ejaaz: AI is either zero or 100 at this point. Ejaaz: It's either deteriorating your product and causing massive outages, Ejaaz: or it's doing something like this and helping the world's leading expert of Ejaaz: AI research actually do a better Ejaaz: job, which is just, there's no middle ground. It's either zero or 100. Ejaaz: And it's just super exciting to see where all of this is going.
Ejaaz: I tweeted this out the other day. Ejaaz: My mind, I don't know about you, Josh, is just in a fog at this moment because Ejaaz: I can't keep up with everything that's going on. Ejaaz: Something's switched over the last 30 days where Ejaaz: The exponential curve has just like racketed up. I don't know whether this is Ejaaz: because I'm in my own echo chamber or not, but it just feels like we are getting
Ejaaz: new model releases every single week at this point. There's so many new Frontier AI labs. Ejaaz: Jan LeCun, who is the former head of AI research at Meta, has now started his Ejaaz: own thing and he's focusing on world models. And I'm like, what is going on here? Ejaaz: It's just, it's so much, but you're going to hear all about it on Limitless. Ejaaz: Josh and I are working 24-7 doing this. We're releasing four new episodes every week, up from three.
Ejaaz: The viewership has been crazy the comments on our last episode Ejaaz: where we talk about uploading a fruit flies brain into a laptop the banger of Ejaaz: an episode go check that out comments are like four times what we normally get Ejaaz: we're responding to every single one we are the most dialed show on ai that Ejaaz: you can potentially watch uh please like subscribe leave us comments give us feedback and josh yeah.
Josh: And important to know like the four episodes is just a Josh: testament to how fast we're going um and just the need to cover Josh: more things so again you will not miss Josh: anything if you watch the show and we will continue to scale to match Josh: the the necessary demand to like sufficiently cover all of this and that's what's Josh: happening i mean it's just downstream of the craziness we are very much in this
Josh: singularity event in the vertical part of the curve so enjoy it like i'm just Josh: trying to take it all in i can take it for what it is get excited about it share Josh: with everyone so yeah thank you guys all for watching as always and we'll see Josh: you in the roundup tomorrow