Sam Nadler (00:00:06): A vampire. A vampire! All right, so here’s what it says. The spirits have spoken—woo hoo! What a spooky sight. Can you talk with the fake teeth in?
Jordan Metzner (00:00:15): Not that much. Sure.
Sam Nadler (00:00:17): Haha. This week we’re breaking it down! Welcome everyone to Built This Week, episode nineteen—the special Halloween episode. Built This Week is 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.
Sam Nadler (00:00:43): I’m Sam Nadler, co-founder here at Ryz Labs. And each and every week I’m joined by my friend and business partner, the vampire himself—Jordan Metzner. What’s up, Jordan?
Jordan Metzner (00:01:01): Yo Sam, happy to be back. My wig’s falling off, everything’s a mess over here in vampire land—but it’s good to be back for another spooky special. We’ve got a lot of treats today.
Sam Nadler (00:01:20): Haha, awesome. How does AI impact the life of a vampire?
Jordan Metzner (00:01:25): It keeps me up all night.
Sam Nadler (00:01:28): Love it. Before we jump into today’s docket, don’t forget to like and subscribe! New episodes drop every Friday on YouTube, Spotify, or Apple Podcasts.
Sam Nadler (00:01:40): We’ve got a fun one today—we’re going over a Halloween-inspired product we built called Mirror, Mirror on the Wall. And we’ll also talk about Google’s new tool for vibe coding, plus some fun AI news—including Nvidia hitting five trillion dollars.
Jordan Metzner (00:02:16): Yeah, we made some Halloween treats last week, but I thought we’d step it up this time since this episode drops October 31st—Halloween day! Let’s jump right in.
Sam Nadler (00:02:43): Cool. So this was built using Google’s new AI Build Studio—we’ll talk more about that later—but let’s start with Mirror, Mirror on the Wall. Let’s see how this goes…
Sam Nadler (00:02:56): Okay, here we go. Mirror, mirror on the wall, who is the smartest of them all? “A single mind, a single name, that is a fool’s and simple game. The brightest thought, the keenest art, can only ever know apart. For wisdom shifts like desert sand—the smartest fool won’t understand.”
Sam Nadler (00:03:23): So there’s my mirror, mirror on the wall! You can see it responding in real time, pulling in my words. Let’s refresh and ask another question.
Sam Nadler (00:03:37): Mirror, mirror on the wall, who is the most valuable company on the stock market today? “The immortal world of stocks and gilded halls—one great name above all other calls. No longer does a bitten apple reign, but windows watching through the sun and rain. Its cloudy kingdom spreads a vast design where all the world’s ambitions intertwine.”
Sam Nadler (00:04:14): Haha, okay, it got it wrong—it said Microsoft instead of Nvidia. But you get the idea. It’s a bit laggy, but imagine this inside a vertical mirror—all powered by Gemini using real-time speech-to-text and text-to-speech.
Jordan Metzner (00:04:32): I think it’s cute! You live on a street with a lot of Halloween activity—imagine setting this up with a webcam and monitor so kids can talk to the mirror. That’d be wild.
Sam Nadler (00:04:59): That’s actually a cool idea! I could probably do that—monitor, camera, speakers… maybe I’ll pull it off.
Sam Nadler (00:05:10): Thanks for the mirror demo, Jordan. Now let’s move on to our tool of the week: the Vibe Coding tool inside Google AI Studio.
Jordan Metzner (00:05:28): Yeah! So we covered it last week as news, but now we’ve both been using it. It’s basically Google’s own build tool—AI.studio/build. You can prompt your idea, and Gemini will start building.
Jordan Metzner (00:05:50): There are over sixteen starter ideas—like banana-powered apps, voice apps, image animations—all sorts of fun stuff.
Jordan Metzner (00:06:11): I actually built Mirror, Mirror using it. But let’s do another one for Halloween—something different. What if we make an app that guesses my costume using the webcam?
Sam Nadler (00:06:34): Let’s do it!
Jordan Metzner (00:06:36): Okay, just hit “Build,” and Gemini does its thing. You can even add difficulty settings to make it a game or a timer to increase excitement.
Jordan Metzner (00:07:23): All right, I’ve got my vampire teeth in—fixing the wig—let’s go. “The spirits have spoken—woo hoo! What a spiky sight! With those gleaming fangs and silver locks, you sink your teeth straight into the elegance of a vampire. Ready to charm the blood out of everyone tonight!”
Sam Nadler (00:08:06): That’s hilarious.
Jordan Metzner (00:08:08): Yep! All built in Google AI Studio—using the webcam and Gemini. They even themed it Halloween-style.
Sam Nadler (00:08:27): Okay, let’s try another idea—something more practical. What if we build a route optimization app for trick-or-treating?
Jordan Metzner (00:08:40): I love it. It could take a neighborhood map and optimize the route to hit the most houses within a one-hour walk. You could set parameters like total distance, kid’s age, radius, etc.
Sam Nadler (00:09:21): Exactly. I’ll prompt it: “Create a route tool that optimizes my route for trick-or-treating to cover as many good homes as possible within one hour in any U.S. city. Include age, preferences, and best candy scores.”
Jordan Metzner (00:09:59): Sounds great! And since it’s on Google’s stack, we can use Google Maps integration. It’s even suggesting weather data and user profiles to track favorite houses for next year.
Sam Nadler (00:10:44): Haha, yes—predict which houses give the best candy based on last year’s data. Genius.
Jordan Metzner (00:11:07): Okay, building now. Let’s test it in Beverly Hills. Eight years old, candy quality six out of ten, spookiness five out of ten… “Find my route.” “Ride the flats of Beverly Hills. Start on Crescent Boulevard. A spooky stroll for one hour—wide, well-lit streets, generous candy, perfect for families.”
Sam Nadler (00:11:51): Haha, love it.
Jordan Metzner (00:12:14): It even adds safety tips! Trick-or-treat planner complete.
Sam Nadler (00:12:41): This thing is incredible. You can even iterate by prompting new features—like note-taking for houses that give full-size candy bars.
Jordan Metzner (00:13:04): Haha, exactly. “Add AI features,” “improve route visualization”—you just click, and it generates.
Sam Nadler (00:13:26): So what makes this tool stand out?
Jordan Metzner (00:13:47): The step-function improvement is that Gemini is already integrated. With others—Bolt, Lovable, Replit—you need API keys and backends. Here, AI is native. That’s huge.
Sam Nadler (00:14:25): Let’s run it again—set quality to six, test in Nashville.
Jordan Metzner (00:14:47): Boom! Sylvan Park—my old neighborhood. It nails it. Great area, flat, safe, lots of houses. Spot on.
Sam Nadler (00:15:24): That’s amazing. So in minutes, we built multiple Halloween tools for basically zero cost.
Sam Nadler (00:15:49): All right, let’s jump into the news. Big one first—Nvidia!
Jordan Metzner (00:16:22): Yep, Nvidia crossed five trillion dollars in valuation this morning—after breaking four trillion just days ago. They even announced a deal with Nokia.
Sam Nadler (00:16:41): Wild. Nvidia keeps selling the “shovels” for the AI gold rush, and demand is insatiable.
Jordan Metzner (00:17:03): Totally. I think they could hit ten trillion if AI keeps scaling and Nvidia stays the top chip supplier. Not financial advice—but still.
Sam Nadler (00:17:29): Yeah, Jensen Huang and the team are crushing it. Nvidia’s at the center of this AI revolution, and it’s not slowing down.
Sam Nadler (00:18:18): Next story—Elon Musk launched Grokapedia, a kind of AI-powered Wikipedia rival.
Jordan Metzner (00:18:50): Yeah, initial reactions say it’s biased—but Wikipedia has always had bias too. Since large models can source information directly, maybe AI encyclopedias are the next step.
Sam Nadler (00:19:07): Right, Wikipedia’s always been the default record, but now LLMs can generate accurate summaries instantly.
Jordan Metzner (00:19:24): Exactly. Elon said it’s meant to be an “alternative to woke Wikipedia.” But it makes sense—why not build your own knowledge base for Grok? The challenge will be SEO and staying relevant.
Sam Nadler (00:20:01): True. Wikipedia’s crowdsourced model adds trust, though. Without that, can Grokapedia be seen as a real source of truth?
Jordan Metzner (00:20:19): Anything written by humans is biased anyway. At least with AI, bias is uniform across the dataset instead of random per article.
Sam Nadler (00:21:03): Fair point. Interesting shift—we’ll see if Grokapedia becomes the new standard.
Sam Nadler (00:21:49): That’s it for episode nineteen—the Halloween special! We covered Mirror, Mirror, built tools with Google’s Vibe Coding, and talked about Nvidia and Grokapedia.
Jordan Metzner (00:22:17): Happy Halloween, Sam—and to everyone listening! This is the first Halloween where you can literally vibe-code your own “Mirror, Mirror on the Wall.” I might set mine up outside the house.
Sam Nadler (00:22:36): Haha, love it. Hope everyone has a fun, safe Halloween—and maybe use Jordan’s tool to plan your candy route.
Jordan Metzner (00:22:58): And we’re almost at fifteen thousand YouTube subscribers! Please like, subscribe, and share. See you next week on Built This Week!
Sam Nadler (00:23:04): This week, we show you how a fresh idea can be a clever treat. See you next time.
