We Built an AI Trading Bot for Prediction Markets — Here’s What Actually Happened

Jordan Metzner: I think this puts NVIDIA at a position where not only are they Tesla's vendor for, you know, the chips inside the vehicles, but they're also now Tesla's competition.Theme Song: Built this week, breaking it down. Built this week, we show you how. A fresh idea, a clever tweak, you locked in true. Built this week.Sam Nadler: Hey, everyone, and welcome to Built this week twenty twenty six, the podcast where we share what we're building, how we're building it, and what it means for the world of AI and startups. I'm Sam Nadler, cofounder here at Ryz Labs, and each and every week, I'm joined by my friend, business partner, cohost, Jordan Metzner. How are doing today, Jordan?Jordan Metzner: Hey, Sam. Happy New Year. Happy to be back. Brand new episode. Got lots going on in the AI world, and, yeah, just so much to talk about.So really happy to be back here.Sam Nadler: Yeah. So we'll jump into the docket. But before I do, don't forget to like and subscribe. This is our first episode of 2026, but we've just eclip the 25 episode mark. This is our twenty sixth episode, so give us a like and subscribe.We're about to cross over 18,000 subscribers on YouTube, and we'd love your support. So our docket, we're gonna cover a tool you built for I I think it's an automated trading bot on Kaushi. And then we're gonna get into one of our favorite AI tools, NotebookLM, to show kinda how we use it for onboarding new employees, and there's tons of different use cases that are that are great for learning. And I'm sure there's tons of others, but learning and just overall educational opportunities. And then finally, the the hottest AI news, couple articles about NVIDIA, couple articles about Anthropic, and that'll be it for for this week's episode.Jordan Metzner: Awesome. Let's jump right into it then. I should.Sam Nadler: Alright. Let's see what you built for Kalshi, which has been, you know, I think last year, all over the news. It's been a exciting, I guess, new way to gamble. I think that may be the probably inappropriate word that they don't like using, but let's see what you built.Jordan Metzner: So, you know, obviously, prediction markets have been really hot craze in 2025 moving into 2026. And, you know, not only have we seen obviously the proliferation of of Calshi and Polymarket, but we've seen, you know, other brands starting to come out and drive their own prediction markets. I think Robinhood has a deal with Calshi. I saw Coinbase is launching a prediction market. I know Fanatics is was talking about launching a prediction market.I believe FanDuel and DraftKings, they've talked about launching prediction markets. And why are all these guys launching prediction markets in the first place? Well, I think one thing that, you know, Cauchy and Polymarket were able to figure out was that, you know, the style and format of these prediction markets has allowed them to circumvent state local regulations in regards to casino gambling online and allowed them to be to be regulated by the CFTC. And so that's kind of been like the first really big change because historically, you know, I live in California, and online sports gambling was not available in California through FanDuel and DraftKings and Fanatics and some of these other players. And yet, Calci is available.And, you know, effectively, the behavior is almost identical. You really feel like you're just gambling on sports. The format is a little bit different. The way it's laid out is a little bit different. The fact that you're, you know, supposedly playing against other players in the market is a little bit different.But for the most part, I mean, it does feel pretty close to the equivalent of gambling. And, you know, for a while now, I've thought about building like a, you know, a trading bot or a chat bot or some kind of bot where I could try to figure out kind where was some opportunity inside of inside of these markets to make money. Now, of course, like, I'm not the only one doing this. I mean, I think you could just go into Twitter and type in Calcchi bot or, you know, any of these other ones. And, you know, we've seen incredible types of, you know, run ups.I think, you know, with with the Maduro evasion invasion last week, we saw, you know, some traders obviously make a ton of money there. Obviously, saw traders with some type of inside information predicting, you know, who was the most searched person of the were of the year, you know, what was the Spotify artist of the year. So we've seen a ton of kind of whether it was known or unknown, insider trading or not. But we've seen some some definitely some mega moves on top of all of these markets. And I think, you know, during the break, I had some free time and and, you know, I just thought, you know, why don't I give my hand a try at making my own trading bot.So, you know, one, I'm not really a good gambler and I don't know a lot about sports. So I think that puts me in a good position because I honestly have no idea who is gonna win the soccer game today between two teams in like the British Premier League. I don't know any of that kind of stuff. And so I figured, you know, I'm gonna build a trading bot. So, you know, first let me give you some parameters of like, you know, what I thought would be important if I was gonna build a trading bot.So the first thing I think that's important to me is like, need feedback. And I need as close to real time feedback as possible because during development, I wanna know if like, you know, my algorithm works, if I'm right or wrong. And so, you know, any type of long dated trade is really not a good option for me. So who's gonna win the World Cup? You know, who's gonna be, you know, the nominee for the next presidential election?Like, those types of trades, they don't really work for me because they're so long dated and so far out. Now they could be good trades for other people, but I just don't really have the time to to figure that out. And so what I'm looking for are trades that that basically are are gonna expire soon. And I don't care when they started, but they're gonna expire soon. Right?So I'm looking for live trades, and normally the best types of live trades happen to be sports because there's a sports going on all the time. Now, again, like I said, I don't really know anything about sports. So, you know, how how should I make these trades? What is my algorithm? How does it work?Right? And so what I did was kinda built an algorithm, obviously, using, you know, Opus 4.5, using Cursor, using Cloud Code, using a few different tools. You know, I I think you know this, Polymarket's not available to traders in The United States. So Polymarket wasn't available to me. Calcchi is available to me and has a fully robust API, so I chose to do Calcchi.I put about $250 into my account, and you can see my current account's at $2.43. So I'm down about, a little less than $6.60, so about $7. And in total, made about 12 trades. And of those 12 trades, I actually won eight of the 12. So I was actually at about 66% success rate in those trades.And yet even with that, I was still negative on my trades. I had a negative yield here. It says, like, overall of 2.6. I didn't trade all the $250. I probably only trade, like, $20.30, $40 at most.So I would say, like, overall relative percentage wise, it's probably even higher as a as a brute percentage. But I think what really killed me was the transaction fees, which I really wasn't predicting as well and didn't have it in my original algorithm. So, anyway, nonetheless, let me show you a little bit of of what's happening. And so what's happening right now is and you can see a change in real time, but essentially, there's a there's a live alert feed that is constantly looking over all the active trades that are happening on Kalshi. And you can see if we go down here, you can kind of see some of this different activity happening.Now I can see, you know, spikes in volume. I can see spikes in whale behavior, essentially like high level bets, essentially high high dollar amounts. I can see I can see changes of behavior. I can see all this kind of stuff. And all that, what that does is it kind of gives me options of of trades that are happening, you know, right now.So here we can filter the trades out by by a whale has made a position or coordinated or even just something that has like a lot of, you know, activity here. This says coordinated entry because it says seven different traders piled in on a yes bet on that particular trade. And so what you can see here, there's all different types of trades of all different types of things. You know, UNC Wilmington versus Northeastern, who's the winner here. Right?So what you can see is this is a whale trade. Somebody bet 236 contracts at 35¢ that UNC Wilmington will will beat Northeastern. And honestly, I don't even know what sport that is. If it's college basketball, football, I I have no idea. But in general, all these trades are available on Cal Street in real time.And so what I did is I looked at, you know, kind of trades like this, and, you know, I would essentially bet based on its recommendation. So this says, you know, bet $92 on yes. I think I was, like, doing it more fractionally, so I probably would have bet $9.20. ButSam Nadler: Is the strategy mainly to follow the trend? Like, you're either following a whale's bet or following a high volume signal. Is that, generally speaking, what what your strategy is?Jordan Metzner: Yeah. I think, you know, one thing I think that makes prediction markets so interesting is it feels like there's insider trading. It feels like there's people trading that have information or access to information that wasn't previously available. And so because of that, and because of your ability and visibility into seeing the market trades, You know, when you go to a Las Vegas sports book, you know the line, you might know a few other pieces of information, but do you really know how many people are betting? How many people just placed a bet in the last ten minutes, five minutes, twenty minutes?You really don't. And what these markets do is they offer some sort of transparency in that respect to allow you to see how the behaviors are happening on top of those trades. And so, yeah, I can look at whales basically like high high dollar amount trades. I look at coordination. So essentially like when there's a high coordination like many people making the same trade.So here you can see seven traders. Here you can see seven traders. Here you can see seven traders. So you know what it's seeing, hey, a lot of people are placing the same bet. Maybe there's some kind of behavior there.And then the last one is just like high, like in general. Like here you can see like when it's on high, essentially, you're seeing like big shakes and move ups between the different between the different markets. Right? So, you know, a bunch of people are moving into this, so that might be like an an active thing as well. Okay.Cool. So that was my dashboard, but let me kinda jump into my trading view page here. So here you can see the trades are gonna come in. I have a confidence score on every trade. I even have a QR code that you can pop up and scan this QR code, and that will actually take you right to the trading page so you can make the trade on your phone.You can see there's 36 live events. We can we can filter by sports or politics or crypto, economy, entertainment, etcetera. We can sort by the reasons, the codes for why these different things are happening. You can also sort by different confidence scores, ultimate ROI, a bunch of different things. So, you know, you can say, hey, I'm looking for, you know, a confidence score of above 80 in sports.And as you can see and not just whale signal, but let's put all signals. So you can see right now, the confidence score of the highest events are only 75. And to me, that's like not a good enough confidence score. In fact, I wanna say a confidence score of 80 or above or even 90 would be optimal. And so, yeah, you can basically use this to make your different trades.You know, this is a a multilevel trade. Here you can see these are kind of some more complicated trades as you can see. So these are like this is a 20¢ trade, but if you're correct, you win $10. Right? So the multiple here is really, really big versus kind of like a who wins, which game if we look at this like, you know, Cremenisi, Calgary winner.Right? You see, like, you bet a dollar to get $9 on the yes here. So my guess is that if you bet on the no, it's probably has a 90% chance of winning. Alright. Great.Let's look at some of the trades I made here. So I deposited some money in my account. I made some trades. I canceled some trades. I focused on some sports here.You know, here, you can see kind of like when I won and when I didn't. This this UI is probably not optimal to see kind of when I won and lost, but, you know, essentially here you can see I'm here's a Serie A game from Italy. Here's from France, and all different games all across. You know, this is Newcastle versus Crystal Palace. And, you know, if we click on the trade here, you can see in this trade, I bet yes, and I got zero earnings, which means I lost.Here I bet yes, and I did win. I won $9. So you can kinda see like which trades I won here. I won four dollars, here I won $0. Anyway, the UI is probably like a little bit suboptimal, but the long story short is that even though I won the majority of my trades, I guess like over 50%, 66%, I still lost money.And I think I lost money for two major reasons. One, the transaction fees are incredibly high, and when I'm taking low margin trades, of like low risk settlements, I found that the transaction fees kinda take away my margin. And then two is that, you know, these are bots trading on bots trading on bots. Right? I'm trading based on the behavior of the whole network, and how many other people are trading on the behavior of the whole network.And so, you know, essentially, you have this, like, kinda consensus algorithm that's happening over time where, you know, consensus drives more consensus drives more consensus. Even if your plan is to do the opposite of the consensus is, then there's a new consensus, and that becomes a new consensus. And so, you know, the good thing about sports is that they're very well defined. Right? A a game starts and it stops.It has a winner and a loser or a tie. And so those things are like pretty well definable. There was some news this week that, you know, poly market didn't wanna pay out on the Maduro trade because they said it wasn't officially an invasion. And so you start to look at some of these events, these political events, and you start to say that actually maybe there's some gray area in the definition of these types of events, whereas in sports, you know, there was no gray area of who won or lost or how many points were made, etcetera. So in general, it was a very educational experience.I learned a ton. And, you know, just to bring it all around full circle, I think, like, my biggest takeaway is that, you know, the only real way to make money on these markets is to have inside information. And without that, I think, you know, you're you even even with algorithm, even with all this data, I think there's just there's no real way to gain some type of alpha or some type of advantage against the rest of the market, at least for me. You know, maybe I could spend more time trading on it. I've seen people spend a lot of time improving their trade bots, but, you know, the market is continuing to change.More people are coming in, more people are coming out, and it still remains to be seen kind of how this will all end. But but, yeah, I just thought it was a cool, fun, little project. I'll probably end up taking my money out of here and, you know, go away with my $7 loss and and call it a day. But, yeah, happy to answer any questions, Sam. I built all this.Obviously, it's all connected directly into Cal Sheets. It's got a back end, front end, the whole thing.Sam Nadler: No. I think it's you know, I think the key would be, like, what, you know, what signals could actually, you know, drive those earnings, but I think that's obviously the hard part. Super cool tool, Jordan. I mean, I think it's topical and, you know, just shows what you can do with AI even though it didn't you know, your 12 trades didn't really work out in your favor. It's just you know, I'm sure someone who maybe has that ability, that gambling experience could take something what you built and turn it into some a tool that, you know, makes money into sleep.So love what you did here. Just to transition, we're gonna transition into our AI tool of the week, which is a tool you actually showed me, and it's called NotebookLM. It's a tool from Google. And what I've really liked about NotebookLM and I you've used it for a few different things than I've used it, but it's almost for me, it's been able to generate really nice education materials. And I think the educational use case of Notebook LM is pretty incredible.So I personally have used it. You know, historically, when we've onboarded new employees over time, we've gotten you know, in in some ways, they've been organized. In other ways, they haven't been organized. But a lot of the shared learning has been just Google Docs. Google Docs of, hey.This is what happens when we do this. This is what you should know about this scenario, etcetera, etcetera. And, you know, there's there's a lot of great things about reading Google Docs as you join a new team, but, you know, there are certain things that can be conveyed better in either an infographic. You know, I love the audio use case. I've heard from you know, I was talking to one of my daughter's teachers about how students are using LLMs, and this wasn't really specific to her age group, but a lot of them are taking their notes, putting it into Notebook LM, and generating an audio overview.So they're basically taking their notes from the week, putting it in a podcast, going to workout, and listening to their podcast or a walk or whatever. And I just thought that's such a cool way to learn. I did take one of our, you know, standard onboarding Google Docs and create an infographic if you want me to share, but it looks like you're building as we speak using LLM. I didn't see the prompt, but is there a way to win consistently on Kaohsi with a trading bot? So you can take what it gives you.You can chat with it, and then I guess you could build a playbook or flashcards or a slide deck based on the output. So we'll see what it says. But yeah. Yeah.Jordan Metzner: So, you know, we can yeah. We can we can definitely add in and post. Hopefully, we can show up on the screen what you built with Nobelk LM, your onboarding document, because I think that's pretty cool. But, yeah, I just asked it to teach me about Kalsi Trading Bots, basically. And so, you know, what it did is it went on the Internet, it looked for some trading bot documentation.It found a YouTube video about Kalsi trading bots. It found the documentation from Kalsi's website about how their APIs work. And it even found a paper that says unraveling the problem is like forest arbitrage and prediction markets. And so it's actually finding white papers and it's bringing it into the mix. And with all of that, it's kind of preparing this, you know, nine page slide deck for me.But one of the questions I asked it is, you know, after I built all this, is there any way to win consistently on Kalshi with a trading bot? And it says, win consistently usually relies on mathematical arbitrage or providing liquidity rather than pure gambling on outcomes. Yeah. Obviously. And so, you know, these arbitrage opportunities, whether they're like cross platform arbitrage, that's like a very famous one.So kind of finding where if you vote yes on Polymarket and no on Calci or no on Calci and yes on Polymarket, that there's a small delta arbitrage there that you can kind of capture that yield. Like I said, I believe fees are too high to actually capture those yields because the fees have to be lower than than the delta between the the arbitrage. There's also some they call it same platform arbitrages. I actually don't believe those truly exist. Those are like market inefficiencies within their own market, and they're honestly like pretty pretty low.I have seen some arbitrages of like time arbitrage where people have been able to for example, it might be a soccer game and someone's in the stadium, and so as soon as the goal happens, they're able to make the trade before they're seeing that goal being put onto the scoreboard on a Calhouni or or Polymarket. But I believe that, like, time arbitrage shall go away. I mean, if these markets have that type of risk, they'll probably be able to tap that to patch that up pretty quickly. And I think overall that's what this is saying is that like there really is a yield because like eventually there's enough volume on these markets that they become pretty perfect markets. And you know, then it just becomes kind of the same as like trading stock options.Right? And you know, I don't think I have a good ability to write an algorithm to be a good stock option broker, stock option option trader. And, you know, it seems that, you know, after learning this these markets, there there's not excess, you know, yield available. Right? And if there is excess yield, they're taking it back in fees.And so I think that was, like, kind of the most interesting thing for me, which is that, you know, while CalSheet and Polymarket may say they're not a casino, they sure feel like the house, and it sure feels like the house always wins.Sam Nadler: Well, if you don't mind, let me share really quick, and I'll show you, like, this quick infographic. I have it pulled up. I just think the educational use case of is is really powerful. And so just let me just keep in mind, this was literally a doc that I a Google Doc. It was just plain text.It was a hiring journey as one of our major businesses is staffing, and we onboard recruiters. And within that one Google Doc, generate an infographic, and it, you know, walks you through no. I haven't really reviewed this to see if it's it's perfect, but, you know, yeah, obviously, phase one is sourcing, phase two is offer and onboarding, and phase three is post hire support. And I just thought this was a really cool illustration. You know, took five seconds to create.And, obviously, with a little bit of chatting with notebook, there's just so many ways you can either generate infographics, podcasts, slide decks, you name it. And I just think it's a really cool tool that people should be aware of. And this is one use case, your use case. I wonder if your deck is ready, but it's it's I, you know, I find it better for this educational content versus Google AI Studio, Boldt or Claude or Open ChatGP TV.Jordan Metzner: Yeah. One thing I, you know, I found Notebook LM. See here, it's still generating the slides. But one thing I thought that was really cool is that I've seen it being used from people of all ages, and that's kind of one unique use case. I've seen, like, you know, doctors and scientists use it as well as, you know, like you mentioned, kind of elementary school and college students using it as well.You know, as you know, it's like original version was kind of creating a podcast based on any subject matter, and it seems like you can still do that with this kind of audio overview. But, I mean, look, Google's added video overview, mind maps, reports, flashcards. You can even create your own custom quiz. We're making a slide deck here as well as data table. So, you know, I I think, obviously, Google will obviously continue to develop this.It's been one of their hits in the AI world. And, yeah, it's just a fun place to to to use as a tool to to get more educated and learn about new topics. And, you know, thankfully, Google's given this away for free, but, you know, hopefully, this will be part of the Google Suite or maybe some other premium features over time.Sam Nadler: Cool. Alright. We got some hot news. We're gonna start overall with NVIDIA and their entrance into the autonomous vehicle space with the launch of, I don't know if I'm pronouncing this correctly, Al Alfa Romeo. So it's, you know, I think it's direct competition to Tesla's self driving.My understanding is they're partnering with Mercedes. And within, you know, this year, they're gonna have their NVIDIA chips inside the Mercedes vehicle. Was it the e class or the c class? I I can't for can't remember. And, yeah, I think it's a really interesting opportunity for NVIDIA.Obviously, it's a long road to master autonomous driving, and Tesla's been at it for, you know, longer than a decade, I think, it feels. But what are your thoughts on on the strategy and and their entrance into the market?Jordan Metzner: Yeah. So, I mean, obviously, this is pretty crazy news, at least from, like, outside perspective, you know, not knowing what Elon and Jensen have been talking about. But, you know, one, obviously, I'm a huge Tesla fan, as you know, and I have a Tesla and full self driving at least since fourteen point two has been really, really incredible user experience. I think this puts NVIDIA at a position where not only are they Tesla's vendor for, you know, the chips inside the vehicles, but they're also now Tesla's competition. And I also think it what's kind of interesting is that, you know, NVIDIA knew that if they wanted to sell more chips inside of vehicles, that they were gonna be the ones that were needed to do the development.And that is actually kind of a new and interesting thing because, you know, historically, Tesla's been buying these chips from NVIDIA and building their own full self driving stack. And I think, you know, if you would have asked me before this announcement who who's Tesla's competition in full self driving, we probably would have named other auto manufacturers that are trying to build self driving vehicles. And I think we probably would not have said that NVIDIA, who is a vendor of Tesla, is going to be Tesla's competition here. And so I think, like, that's probably the biggest change, which is, like, you know, in areas in which NVIDIA believes that they're the consumer of these chips is not able to fabricate the technology in which the chips are being used, that they have gone ahead and gone and developed software simply just so that their their customers can buy more chips from them. And I think that is like super interesting.Now the flip side of that is kind of what Elon has said publicly about this, which is, you know, they've gotten maybe 99% of the way there, but, you know, Tesla truly believes that like the last, you know, 1% is a is a very, very long tail, and it'll be very difficult for these manufacturers to get there. Now what does this mean for NVIDIA? What does this mean for Tesla? It's really hard to say. I think we all know that NVIDIA is sitting on a ton of cash.Does NVIDIA go buy an auto manufacturer and become Tesla's competitor? You know, does NVIDIA become the provider of first source to many auto manufacturers of full self driving? You know, I think I'd mentioned this to you a few weeks back, but, you know, with the latest Tesla update, I was truly and I still do believe that Tesla will sell their full self driving vehicle technology to other auto manufacturers. It seems so obvious. But it's if that's the case, now they're they're gonna compete against NVIDIA, which, you know, might be willing to give it the software for free if they buy a certain chip.Right? And so that actually might have some, you know, impact on market dynamics, etcetera. But I also think, like, you know, at the end of the day, a lot of people do do not understand and have not experienced the quality of Tesla's full self driving in its latest iteration. And I think that when people feel it, when people experience it, and like you mentioned just a few weeks ago, Sam, when you had that first Waymo, I told you that they were driverless. You knew they've been driverless.You knew that. Right? But once you're in the car and you're driving and you feel the turns and you feel the experience, everything just kinda changes, you know? And I think the same is true for the Tesla self driving. You know, we haven't seen large populations of people who don't have Teslas, like, experience it, use it, feel it.You know, Tesla's Robo Taxi is coming out to market now, but, you know, you always tell me about how, you know, your dad hates driving. He never likes to drive long distances. He never likes to drive at night. Well, know, what happens if he has a Tesla and he's full self driving all the time? You know, how much better does that make his quality of life?Or maybe he's just willing to use it at nighttime so that he you know, he's he's willing to maybe go out to dinner when previously he wouldn't. So obviously, this is a new day, and and a lot of things are happening. I can't wait until I can like actually just like, you know, completely ignore my steering wheel and and gas and brake and just, you know, kinda be on my laptop while I'm inside the car.Sam Nadler: Or take a quick nap.Jordan Metzner: And hopefully, we're not that far away. Yeah. Maybe even take a nap. Exactly. Yeah.Watch a movie. Watch it. You know? Do whatever. SoSam Nadler: Oh, yeah. That LA traffic always on me is a benefit.Jordan Metzner: Yeah. I don't think anybody likes waiting when they don't need to. But, yeah, if you're in traffic and you can watch a movie and you fall asleep or, you know, get some work done. Right? So, yeah, obviously, huge opportunity here.Let's just jump right into the next NVIDIA story because I think it'sSam Nadler: also incredibly interesting. But, yeah, NVIDIA launches the new Rubin chip, and I know you have a bit more insight on why this chip is actually different, but it seems like it's like the next wave of or will enable the next wave of massive AI models.Jordan Metzner: Yeah. So, I mean, look. Obviously, they they being NVIDIA is trying to, you know, consistently improve the technology of their chips. They are obviously building, like, you know, better packs and better formats and new implementations. This new chipset is called a Rubin or a Rubin Pod.And I think each side inside each Rubin is 72 GPUs, and then I think inside of each GPU, there's more than a thousand plus chips. So we're talking like thousands and thousands of of GPUs at the core that are able to power these things. And, you know, as the market leader, we're gonna continue to see new use cases that were not available previously. So, you know, we're gonna see increased inefficiency. We're gonna see increase in new applications, and hopefully some new discoveries, whether that's scientific, mathematical, or even things that are fun like video generation, image generation, self driving, and whatnot.So overall, it just seems like NVIDIA is kind of, you know, essentially kind of they're striking on all cylinders. But, yeah, it's you know, I mean, I did not expect this coming out of NVIDIA at the first of the year. And, yeah, the game's heating up and, you know, competition is probably good for the market, and hopefully, this will this will continue to put pressure on Tesla to to continue to improve their product.Sam Nadler: Yeah. Just to be specific, it's from this article, it says the Rubin architecture will operate three and a half times faster than the Blackwell architecture on model training and five times faster on inference tasks. I mean, you know, that's that's a pretty big increase in performance. So, you know, I I think to me, obviously, it was expected that I assumed, I guess, for me that NVIDIA would keep pushing the boundaries on on new chip development, but I think it's really exciting. They're entering into this autonomous driving space.Super bullish about the next several years with NVIDIA. And, yeah, anything else in the news or NVIDIA? Anything else you wanna cover?Jordan Metzner: I think just wrapping up, we didn't really talk about Anthropic and Claude, but just, you know, really quickly, Anthropic raised a new round this week. There's a new version of Claude code. But, you know, after sitting with my developers this week, you know, everybody is just enthralled with the quality of work coming out of Anthropic and the quality of Claude code, the quality of Opus 4.5. And, you know, this feels like the early days still. I think, you know, another Opus version, you know, I e something a little bit better than 4.5 will really be amazing.We're seeing a ton of tools written on top of Cloud Code. I think I told you about the the Ralph Wiggum plug in that allows you to do things like that. There's a vibe code check-in plug in. I've seen some Cloud Code observability tools. So it's it's really getting to a cool space where not only is is Cloud Code and Anthropic and their models and becoming kind of standardized in the market, but we're starting to see some really cool tools on top that just allow developers to do even crazier things.And again, like I said, Sam, like, this is just the early days. We're in January of the beginning of the twenty twenty six. But I think in three to six months from now, we're gonna say, wow. That felt, know, really antiquated to what's available on the market today.Sam Nadler: Cool. Totally. Well, thank you, Jordan. It was great to see your Kalshi tools. Great to hear your take on NVIDIA.And don't forget to like and subscribe. New episodes every week. We got some great guests lined up already throughout pretty much the next eight to ten weeks. So see you next week, everyone. Any final thoughts?Jordan Metzner: Yeah. Thanks everyone for listening. You can check us out at builtthisweek.com. You can check us out on YouTube under built this week or any of your favorite podcast platforms. Coming into 2026, we've got a ton of new entrepreneurs in the AI space and even in the non AI space joining us.So we got some really exciting episodes coming up, and thanks everyone for listening. Don't forget to like and subscribe, and see you again next week. Bye bye.