🍜 March Madness goes AI..

A $1 million bet on Houston to win March Madness?

March 20, 2025

In partnership with

Welcome, Noodle Networkers.

Nvidia just dropped some major AI announcements at GTC 2025 🤖, unveiling next-gen chips, a new robotics platform, and an AI physics engine. But despite all the hype, Nvidia’s stock took a dip—because apparently, revolutionizing computing is not enough unless you also promise investors a self-learning money printer. Maybe next year. Meanwhile, one tech CEO is putting AI’s predictive skills to the ultimate test with a $1 million March Madness bet 🏀. His AI-powered bracket is all-in on Houston, while a seasoned sports gambler is backing Duke. If AI wins, Wall Street will start replacing analysts with chatbots. If it loses, humans can enjoy one last victory before the machines take over. Over at Walmart, AI is now helping with product sourcing 🛒, predicting what customers will want before they even know it. If this goes well, we might never run out of essential items again. If it goes too well, AI will start auto-ordering things for us, and we will all wake up one day to 600 packs of ramen at our doorstep. The University of British Columbia is banning China’s DeepSeek AI from campus 🔒, citing security concerns. Whether this is a necessary move or just another AI Cold War moment is up for debate, but one thing is certain—students will now have to write their essays the old-fashioned way: begging ChatGPT instead. And in the AI investment world, researchers are warning that billions could be wasted on dead-end projects 🚧. Apparently, for every groundbreaking AI model, there are a dozen that achieve absolutely nothing—kind of like a Roomba that spends all day spinning in circles instead of cleaning.

Is AI truly building a smarter future, or are we just throwing massive amounts of cash at robots that cannot even find the WiFi? Let’s get into it...

In today’s AI digest:

  • Nvidia’s GTC keynote unveils four big AI surprises 🤖

  • AI could help win March Madness brackets with a $1M bet on Houston 🏀

  • Walmart develops an AI assistant for product sourcing đź›’

  • UBC bans China’s DeepSeek AI over security concerns đź”’

  • AI researchers warn billions may be wasted on dead-end projects 🚧

Read time: 7 minutes

How long do you want our daily issue to be?

Login or Subscribe to participate in polls.

WHAT’S HAPPENING TODAY

Use AI as Your Personal Assistant

Ready to save precious time and let AI do the heavy lifting?

Save time and simplify your unique workflow with HubSpot’s highly anticipated AI Playbook—your guide to smarter processes and effortless productivity.

(source: barrons)

1. Blackwell Ultra: The AI Superchip That Eats Data for Breakfast

Nvidia unveiled the Blackwell Ultra, the next-gen AI processor that’s 50% faster than its predecessor, the NVL72. In non-nerd terms, that means it can crunch AI models faster than a Tesla goes 0-60.

  • What it means: This chip is so powerful, your AI assistant might start answering your questions before you even ask them.

  • Huang’s energy: â€śWith Blackwell Ultra, we’re not just raising the bar; we’re sending it into orbit.” (He didn’t actually say this, but c’mon, he totally could have.)

2. Vera Rubin AI Server: Because AI Never Sleeps, But Needs a Faster Bed

Named after the astronomer who helped prove dark matter exists, the Vera Rubin AI server is 3.3 times faster than previous models and built for handling massive AI workloads.

  • What it means: It’s basically an Olympic sprinter for AI processing.

  • Reality check: At this rate, AI servers are evolving faster than my ability to remember my own passwords.

3. Nvidia Dynamo: The AI Personal Trainer Your Models Didn’t Know They Needed

Nvidia Dynamo is an open-source inference software system designed to help AI models scale efficiently. Imagine an AI coach yelling “FASTER!” at your neural networks while they pump digital iron.

  • What it means: AI training is about to get way more efficient, so expect chatbots and virtual assistants to go from “uhh, let me think” to “got it, boss” in record time.

  • Ideal use case: Hopefully, it’ll also teach my AI fridge not to judge me for eating ice cream at 2 AM.

4. Meet "Blue": Nvidia’s Disney-Collab AI Robot (That Probably Dances Better Than You)

Nvidia partnered with Disney Research and Google DeepMind to unveil “Blue”, a next-gen robot powered by a new physics engine called Newton. The goal? Smarter robots for entertainment, industry, and maybe even making your morning coffee.

  • What it means: Nvidia’s AI-powered robotics just leveled up, and if this thing starts doing backflips, Boston Dynamics is officially on notice.

  • Hottest take: I swear if this robot gets cast in a Marvel movie before I do, I’m filing a complaint.

Why It Matters: Nvidia isn’t just leading the AI race—they’re paving the road, building the cars, and selling the fuel. These breakthroughs mean faster AI models, smarter robots, and possibly the first GPU-powered digital therapist to help us cope with how fast AI is advancing.

March Madness

(source: thestar)

The Digest: A $1 million bet on Houston to win March Madness? That’s exactly what Alan Levy, CEO of 4C Predictions, is doing—except he’s not relying on gut instinct, lucky socks, or the wisdom of your coworker who “totally called that upset last year.” He’s letting AI make the picks. Levy’s AI model is crunching data like a sports-obsessed supercomputer, analyzing stats, coaching tendencies, and player performance to outsmart professional gambler Sean Perry. If this AI pulls off a perfect bracket, Vegas oddsmakers might start sweating.

Key Details:

  • Houston’s Got Game (And AI Knows It) – The No. 1 seed Houston Cougars are riding a 13-game win streak and just snagged the Big 12 Championship. AI says they’ve got what it takes—but they still have to survive March Madness chaos, where even the best teams can get dunked on by a squad no one’s heard of.

  • AI’s Playbook > Your Office Pool Strategy – Levy’s AI isn’t just picking based on win-loss records—it’s scanning advanced metrics, coach strategies, and even referee biases. Meanwhile, your coworker Steve is picking teams based on uniform colors.

  • Man vs. Machine: The $1M Showdown – If AI outsmarts Perry’s human-made bracket, it could prove data-driven betting is the future. If it fails? Well, at least the AI won’t have to explain the million-dollar loss to its wife.

Why It Matters: March Madness is gloriously unpredictable, but if AI can consistently outguess humans, sports betting might never be the same. Of course, we’ll believe it when AI successfully predicts a 16-seed beating a No. 1 again. For now, let’s just hope AI doesn’t start trash-talking—nobody wants a chatbot flexing about its bracket picks.

Walmart develops an AI assistant for product sourcing đź›’

(source: pymnts)

The Digest: Walmart just introduced "Wally," an AI assistant designed to help merchants source products faster, analyze sales trends, and basically do all the boring spreadsheet work so humans don’t have to. Think of it as a super-powered retail sidekick that never takes a lunch break or calls in sick.

Key Details:

  • Wally: Data Cruncher Extraordinaire – Wally can instantly analyze mountains of sales data, identify trends, and diagnose why certain products are flopping or selling like hotcakes—which means fewer “Why aren’t people buying this?” meetings.

  • AI With Customer Service Skills – If something goes wrong, Wally can raise support tickets and troubleshoot issues faster than your IT guy responds to emails.

  • Predicting the Next Big Seller – Wally doesn’t just report on trends—it predicts them. So if Walmart suddenly starts stocking AI-recommended fidget spinners again, don’t be surprised.

Why It Matters: Walmart is using AI to make retail sourcing faster, smarter, and way less tedious. Merchants can spend more time making strategic decisions and less time digging through spreadsheets like digital archaeologists. Let’s just hope Wally doesn’t start negotiating its own employee discount.

University of British Columbia

(source: colonist)

The Digest: UBC just slammed the door on China’s DeepSeek AI, banning it from university networks and devices over privacy concerns. Apparently, DeepSeek collects data more aggressively than a grandma collecting coupons, and UBC wasn’t about to let it snoop on students’ late-night essay panic sessions.

Key Details:

  • DeepSeek’s Data Grab – The AI reportedly logs personal info, keystrokes, and possibly even your worst typos, making it the digital equivalent of an over-attentive librarian.

  • Security? What Security? – DeepSeek’s encryption is about as solid as a wet paper towel, which means bad actors could potentially access your data faster than you can say “cyberattack.”

  • UBC’s Firewall Goes Up – The university banned DeepSeek on all university devices and blocked it on their network, meaning students now have to find another AI to do their homework for them.

Why It Matters: With AI tools becoming essential for research and learning, institutions are having to decide which ones they can trust. UBC’s ban sends a clear message: if your AI is snoopier than a reality TV producer, you’re not getting in.

AI waste

AI researchers warn billions may be wasted on dead-end projects 🚧

(source: futurism)

The Digest: AI researchers are throwing up red flags, warning that billions are being funneled into AI projects that might be going nowhere faster than a treadmill on max speed. Despite Big Tech's obsession with scaling up, experts say that just making AI models bigger won’t magically create Artificial General Intelligence (AGI)—which means a whole lot of investors might be paying for a rocket that never leaves the launch pad.

Key Details:

  • More GPUs, More Problems? – Microsoft plans to spend $80 billion on AI infrastructure in 2025, because apparently, the solution to everything is just “add more GPUs.” But researchers argue that scaling alone won’t crack the code for human-like intelligence—just like stacking more toasters won’t get you a spaceship.

  • VCs Are Betting Big… Maybe Too Big – Investors poured $56 billion into generative AI last year, hoping to fund the next AI gold rush. But without clear breakthroughs, some fear this might end up looking more like the dot-com bubble—except with way more chatbot-generated press releases.

  • AI’s Energy Hunger = “Go Nuclear” Mode – Training AI models is so power-hungry that companies are literally looking at nuclear energy to keep data centers running. Yes, we’re now at the point where building AI is competing with national security for electricity.

Why It Matters: If AI research is hitting a wall, we could be staring down another “AI winter”—aka a funding freeze that could make crypto’s latest crash look like a minor inconvenience. To avoid this, tech companies might need to rethink how they build AI instead of just tossing money into a GPU-shaped bonfire.

THE NOODLE LAB

AI Hacks & How-Tos

The Digest: Artificial Intelligence (AI) has revolutionized fitness by offering personalized workout plans tailored to individual goals and preferences. Platforms like Planfit utilize AI to create customized routines, helping users achieve their fitness objectives more efficiently.​

How-to:

  1. Access Planfit:

  2. Set Up Your Profile:

    • Input personal details such as age, fitness level, goals, and available equipment.​

  3. Receive Personalized Workout Plan:

    • Based on your information, Planfit's AI generates a customized workout routine.​

  4. Follow Guided Exercises:

    • Access detailed exercise guides with videos to ensure proper form and technique.​

  5. Track Progress:

    • Monitor your workouts, achievements, and muscle recovery through the app's intuitive interface.​

Explore More: Use Follow-up Mode to dive deeper into a specific industry and fine-tune your niche selection. We just covered AI-powered niche discovery in a workshop—access it at the top of this email! 🚀

Trending AI Tools

  • Nichely: An AI-powered topic discovery tool that helps users uncover endless topics within any niche, enabling content creators and marketers to find untapped opportunities.

  • Krikey AI: A platform that converts videos into 3D character animations using machine learning, providing a browser-based video editor with AI-driven voice dialogue features.

  • GMind AI: A productivity tool designed to assist professionals with research, ideation, and web content creation, featuring content generation, research assistance, and business support functionalities. ​

  • Flux: A text-to-image model developed by Black Forest Labs, enabling users to generate high-quality images from textual descriptions, suitable for creative projects and content creation. ​

  • AiSciVision: A framework that specializes large multimodal models in scientific image classification, aiding researchers in analyzing and interpreting complex scientific images.