BUILT 2 SCALE | RECEIPTS OR REGRETS | End of Year Special

Welcome to our brand new segment: Receipts or Regrets. A receipt is something you keep because you're proud of the call. A regret? Well, that's obvious.

Matty and Scotty look back at 32 episodes of bold predictions, hot takes, and occasionally terrible life choices. With help from ChatGPT o1 (the only model that could handle all 32 transcripts), here are the calls that aged like fine wine and the ones that aged like milk.

Let's dive in.

Robots in the Home by 2025: Slop Receipt

Scotty called it early: robots in homes by 2025. The Chinese came through with $20K humanoids you can actually buy. Tesla and Figure didn't deliver, but the ball went in. Wrong pocket, but still counts. Slop receipt kept.

Make Every Australian a Millionaire: Escalate

Australia has $20 trillion in raw materials needed for AI infrastructure. Scotty proposed inviting global companies to build data centers here in exchange for giving Aussies a million dollars each plus free compute for life. The government? Still no AI strategy. No AI czar. Nothing. This isn't a receipt or regret. It's escalate and shout from the rooftops.

AI in the Avocado: Big Receipt

Guzman y Gomez IPO'd at $45 per share with a $5 million valuation per store. Scotty said there must be "AI in the avocado" for that multiple to work. Today? Stock down 55% to $20. Lacks AI in the avocado confirmed. Receipt kept.

Talking to Anna from Sesame AI in Bed: Regret

Matty got caught talking to an AI voice assistant under the sheets. His wife walked in. "Who are you talking to?" "It's a bot!" didn't help. New rule: no bots in the bedroom.

Steve Irwin Tech Talk in Dallas: Big Regret

Scotty invoked Steve Irwin while doing a tech talk in Dallas to lean into his Australian accent. Tough crowd. Too soon. Too much of a stretch. As a now embedded Austin local, even more cringey. Won't be doing that again.

Qantas: Split Decision

Scotty called out Qantas for no Wi-Fi on international flights in 2025 when Starlink exists. Regret. Matty? Qantas fanboy. Status points, flexi tickets, business class upgrades for $3K. He's keeping the receipt. Built 2 Scale will be taking separate flights.

Brett Adcock 200x Apple: Regret

Brett said Figure AI would be worth $800 trillion (200x Apple). Three years in, no product, lots of parties. Figure AI revenue? Near zero. Apple's revenue? $416 billion. Scotty's calling regret until Brett hires those two HR managers.

Limitless Pendant Meets Zuck: Regret

Matty bought the Limitless AI pendant. One year late, terrible battery, no Find My feature. Then got an email changing privacy terms. One hour later? Meta acquired them. Now Zuck has all his data, including the time he argued with his dog Hank and the AI thought Hank was a difficult coworker. Regret.

Dual Carriageway: Apple and Google Ecosystems: Regret

Matty self proclaimed he'd run dual ecosystems. Two laptops, two phones, two lives. Result? Paid $300/month for Google Ultra with no features and watched the Android Gemini phone camera take 10 seconds to open. Converted to Mac. Everyone not on Mac is wrong. Receipt on Mac, regret on Google.

Peak Waymo vs. Tesla: Receipt

Both called Tesla's long game over Waymo's robo taxi approach. Elon can produce a robo taxi for a tenth of the price. Economics win. Waymo might retrofit other OEMs with their tech, which is smart B2B play for the lefties and Euros who won't touch Tesla. But best product wins. Receipt.

First Year ARR is Nonsense: Big Receipt

The bubble frothiness of first year ARR announcements was too much. Monthly subscriptions reported as annual recurring revenue without proof of retention? Nonsense. ARR should only count annual upfront payments in year one. Y Combinator ARRs should exclude other YC ecosystem revenue. Receipt kept.

OpenAI Wants to Be the Apple of AI: Receipt

Called this in April before Jony Ive joined. Now it's obvious. If OpenAI doesn't get hardware right, they're in trouble. Apple might acquire them. Sam as CEO of Apple by end of 2026. Receipt banked.

Coders Shouldn't Be Called Builders: Rant Receipt

Scotty unloaded: Engineers sitting in cafes? Fine. Software architects on Figma? Okay. But builders? No. Leave that term alone. Someone makes a mistake on a construction site, they fall off a scaffold, they die, the builder goes to jail. That's the risk. The pyramids are unsolvable. Not a Figma pitch deck. Sacred word. Keep it.

Just In Time Software: Receipt

Matty predicted AI would write code on the fly to build interfaces and business logic that haven't been predefined. It's already happening. Dynamic software driven by LLMs instead of predefined workflows. The software version of "let's just do it ourselves" is here. Receipt.

The Takeaway:

Looking back at 32 episodes, most calls held up. Robots came (just from China). Tesla's winning the robo taxi race. OpenAI is chasing hardware. And no bots in the bedroom is now a firm rule.

Here's to more receipts in 2026.

Which call surprised you most? What's your prediction for next year?

Keen to stay ahead? Subscribe to Built 2 Scale on YouTube (link in comments)

In the final episode of 2025, Scotty and Matt celebrate 33 episodes of Built 2 Scale by diving into Yann LeCun's ultimate entrepreneurial pivot, raising $3 billion in euros after getting ousted from Meta by Alexander Wang to work on spatial intelligence. They dissect why this is terrible news for Elysium (autonomous homes now have a 10 year delay), celebrate Sergey coding again at Google while the Qantas vs United business class wars rage on, and introduce the year end segment Receipts or Regrets where they review their boldest predictions. From Brett Adcock's 200x Apple claim to robots in homes by 2025, from AI in the avocado to Limitless getting acquired by Zuck with zero notice, they hold nothing back in this year end wrap up featuring domain buying confessions, builder vs coder rants, and why coders should never be called builders.

Built 2 Scale | Episode 33

TIMESTAMPS:

0:00 Final Episode of 2025: 33 Episodes Complete

2:01 Yann LeCun Raises $3B for Spatial Intelligence Startup

4:06 Why This is Terrible News for Elysium Autonomous Homes

7:00 Brett Adcock's Figure AI Christmas Party: Robot Rave with Deadmau5

9:02 Voice AI Bandwidth Solution: Scotty's 30 Year Long Bet

13:44 Human Like Voice vs Fast Intelligence: What Do You Actually Want?

16:51 Sergey Back Coding at Google: The Return of the Founder

21:57 Receipts or Regrets: Year End Prediction Review Begins

23:41 Matt's Receipt: Robots in Homes by 2025 (Chinese Did It)

25:47 Scotty's Escalate: Make Every Australian a Millionaire With Raw Materials

28:52 Receipt: AI in the Avocado, Guzman y Gomez Down 55%

31:01 Regret: Sesame AI Bot in Bedroom, Wife Not Impressed

33:17 Regret: Steve Irwin Tech Talk in Dallas

35:38 Qantas Fanboy vs United Points: The Business Class Debate

40:09 Receipt: Peak Waymo, Tesla Has Long Game Sewn Up

44:19 Regret: First Year ARR is Nonsense, Y Combinator Circular Economy

46:39 Receipt: OpenAI Wants to Be Apple of AI (Johnny Ive Hire Confirmed It)

52:27 Rant: Coders Shouldn't Be Called Builders, Leave Us That One Term

56:26 Receipt: Just in Time Software Revolution Happening Now

58:52 Matt's Dirty Drunk Habit: Domain Buying, Sold Usainboat.com for $20

1:00:34 Limitless Acquired by Meta: Zuck Now Has All of Scotty's Dog Arguments

This Episode Covers:

1. Yann LeCun raising $3 billion in euros for spatial intelligence after Meta exit, choosing Europe where innovation goes to die
2. Why Yann working on spatial intelligence is terrible news for autonomous homes timeline
3. Brett Adcock throwing robot rave with Deadmau5 while still having no product after 3 years
4. Voice AI bandwidth debate: Human like conversation vs fast accurate intelligence
5. Sergey back coding at Google, spending 90% of time teaching rather than sitting on $500M yacht
6. Year end Receipts or Regrets segment reviewing boldest predictions of 2025
7. Robots in homes by 2025: Chinese delivered with $20K Unitree, not Tesla or Figure
8. AI in the avocado: Guzman y Gomez down 55% from peak, now $2B market cap
9. First year ARR is nonsense: Y Combinator circular economy needs to exclude internal revenue
10. OpenAI wants to be Apple of AI: Johnny Ive hire proved the hardware thesis
11. The builder rant: Coders sitting in Starbucks with Frappuccinos aren't builders, leave us that one term
12. Just in time software: LLMs writing code on the fly rather than predefined workflows
13. Qantas vs United business class points arbitrage strategies

KEY INSIGHTS:

1. Yann's strategic retreat: Raising $3B in Europe for spatial intelligence after Meta exit shows classic researcher move to longer horizon tech when pressure mounts. Europe welcomes unproductive research with open arms
2. Sergey's teaching multiplier: Back coding at Google spending 90% of time teaching creates compounding effect as team learns his mental models. More valuable than sitting on yacht
3. Voice AI clarity test: If you want it human like rather than fast and smart, you probably need to talk to actual humans. Best technology disappears into substrate
4. Brett's capital burn rate: Three years in, robot rave parties with Deadmau5, no product. 200x Apple claim ($600 trillion valuation) looking like major regret
5. First year ARR fraud: Y Combinator ecosystem needs to exclude internal revenue between startups. Monthly subscribers times 12 without churn data is nonsense metric
6. OpenAI Apple thesis: Johnny Ive hire confirmed hardware battlefield strategy. Either sets up Apple acquisition with Sam as CEO or prepares for OS layer competition
7. Builder terminology sacred: Someone dies on construction site, builder goes to jail. Kid on cursor in Starbucks is not a builder. Engineers Australia membership for software degrees is offensive to civil engineers
8. Just in time software arriving: AI writing code on the fly beats predefined workflows. Internal bottlenecks now solvable without buying point solutions
9. Qantas points hack: Buy flexi ticket, bid 60K points for business upgrade, 80% success rate. $2.5K total vs $12K buying outright
10. Limitless Meta acquisition: One hour between aggressive privacy policy change and Meta acquisition announcement. Zuck now has data on Scotty arguing with dog Hank

About Built 2 Scale:

Weekly analysis of artificial intelligence business strategy, robotics adoption, and technology market dynamics. Hosted by Scott Wilcox (Australia) and Matt Perrott (Austin), providing actionable insights for founders, investors, and technology executives navigating the AI transformation.

Subscribe for AI business intelligence: /@built2scale

Available on Spotify | Apple Podcasts

#AI #YannLeCun #SpatialIntelligence #BrettAdcock #FigureAI #Google #Sergey #OpenAI #Receipts #Regrets #2025Predictions #Built2Scale #YearEndReview #TechNews

BUILT 2 SCALE | TOOL OF THE WEEK | Episode 32

Every week, Matty and Scotty break down the strategies, frameworks, and tools that separate the world's best builders from everyone else.

This week: Alloy, the prototyping tool that just created a 10x moment for non-technical people in software companies.

Alloy: The "Aha" Moment for Product Managers

Alloy is a prototyping tool that lets non-technical people adjust software on the fly. Take a screenshot of your app, drop it into Alloy, and use natural language to make changes. It looks real, feels interactive, but it's a prototype.

The Problem It Solves

Before Alloy: Take screenshots, drag them into Figma, add arrows and markups, copy-paste elements from other screens. Tedious. Time consuming.

With Alloy: Prompt it. "Hide the side panel and make the drawing full screen. Add markup tools, text, red drawing, pins, and comments. Let users save versions as private or distribute to subcontractors."

Three minutes later? Interactive prototype complete.

Real World Impact

Matty had a US prospect ready to buy, but they needed one feature: drawing markup tools. Instead of saying "I promise the engineers are working on it," he used Alloy to create an interactive demo in minutes. Sent the video to the client. Deal moving forward. Engineers building it in two weeks.

That's the power: show, don't promise.

Who Uses It?

Product managers, sales teams, anyone who needs to visualize changes fast. You can grab a competitor's website, screenshot it, and say "do this, but add our features." It exports to Figma and code (though the code isn't production ready). The value is in design and iteration speed.

The Bigger Picture: AI Native Private Equity

This tool sparked a bigger discussion: businesses are no longer just building software for industries. They're participating in industries as AI native players.

Instead of building a tool for lawyers, start an AI enabled law firm that's better than the rest. Instead of servicing construction, acquire construction companies and apply your automation logic.

This is the new age of private equity: acquire existing businesses with demand, apply AI to solve the logic layer (input, logic, output), and turn 10% profit margins into 30%.

For software companies facing shrinking margins, the pivot isn't just selling tools. It's acquiring businesses and applying your logic to them.

The Takeaway:

Alloy represents a 10x improvement in prototyping speed. But the real insight? AI enables new business models. Don't just service an industry. Participate in it. Acquire businesses, apply automation, and enjoy the upside.

What's your "aha" AI moment been? Have you found a tool that genuinely changed your workflow?

Keen to stay ahead? Subscribe to Built 2 Scale on YouTube (link in comments)

BUILT 2 SCALE | ROBOT RUNDOWN | Episode 32

Every week, Matty and Scotty break down the latest in robotics: from humanoids to specialized automation, and everything reshaping how we work, eat, and live.

This week: Travis Kalanick (Uber founder) is quietly revolutionizing food production, and the debate between humanoids vs. point solution robots is heating up.

Travis Kalanick's Cloud Kitchens: Rethinking Automation From First Principles

Travis is back on X, and he's showcasing something big. Cloud Kitchens isn't just automating food delivery. It's building the infrastructure layer for food production itself.

The latest demo? A fully automated system assembling and bagging 300 bowls per hour using specialized robots for specific tasks. No humanoids walking around. Instead, Travis redesigned the entire process from scratch.

This is the Elon playbook applied to food: don't fit into existing architectures. Redesign the whole thing. Creative first principles thinking applied to manufacturing speed and efficiency.

Humanoids vs. Point Solution Robots: The Great Debate

Here's the question reshaping robotics strategy: Do you want one humanoid doing everything, or multiple specialized robots each doing one thing perfectly?

In manufacturing, Travis is proving the latter. In homes, the question gets more complex.

The Coming Wave of Niche Robotics

Just like how niche coding apps must specialize to compete with giants like Gemini, robotics will follow the same pattern:

The Big Players:

• Tesla (Optimus)
• Figure
• Unitree (China)
• Gens

5 to 7 "Mag Seven" type companies building general purpose humanoids.

The Point Solution Explosion:

Hundreds of niche companies building specialized robots powered by the tech infrastructure from Google, Nvidia, and AI advancements.

Examples already emerging:

• Abby (Melbourne): Companion robots for elderly care, designed to be friendly and colorful
• Ongo: A desk lamp robot with personality that interacts with you (think Toy Story vibes)

These might seem like gimmicks, but they represent the next thousand successful businesses taking robotics mainstream.

The Domestic Space Gets Smart

What's already in your home? Legos. Barbie. Eight Sleep. Furniture. Appliances.

Now ask: when does the intelligence layer and robotics layer get plugged into what's already there?

Just like AI integrated into commercial spaces, we're about to see it plug into domestic life. Expect acquisitions. Expect $150 products on shelves that bring real robotics tech into everyday homes.

Rising Tide Lifts All Boats

The narrative that big companies will dominate and kill all startups misses the point. The technological revolution in AI and robotics is spawning entirely new categories.

Yes, we'll talk about data centers in space. But we'll also celebrate the little products that make robotics tangible and accessible.

The Takeaway:

Robotics isn't just about humanoids. It's about rethinking systems from first principles (Cloud Kitchens) and creating specialized solutions for specific needs (point solution robots). The next wave won't just be dominated by giants. It'll be defined by hundreds of niche players making robotics part of daily life.

Humanoids or specialized robots? What's your bet for the future?

Keen to stay ahead? Subscribe to Built 2 Scale on YouTube

BUILT 2 SCALE | AI NEWS | Episode 32

Welcome back to Built 2 Scale. Every week, Matty and Scotty cut through the noise to bring you the AI developments that actually matter: the moves reshaping markets, the strategies redefining competition, and the shifts you need to understand to stay ahead.

This week: OpenAI's code red moment, Meta's aggressive pivot, data centers in space, and unexpected market effects from the AI boom.

OpenAI's Hardware Play: 40 Apple Engineers and a Code Red

OpenAI just hired 40 Apple hardware engineers. The battlefield has moved to hardware.

The vision? AI models running on network nodes, generating what you need on the fly. No traditional operating system. Just intelligence in real time.

If Apple builds AI into iOS and runs models locally, do you even need ChatGPT subscriptions? That's the existential question OpenAI is racing to solve.

Meta's Limitless Acquisition: The Privacy Policy That Broke the News

Meta acquired Limitless, the AI wearable. Customers got an aggressive email demanding privacy updates or lose access. Fifteen countries were cut off.

One hour later? Meta announces the acquisition.

With Yann LeCun's departure and this move, Zuckerberg is having his own code red. Meta now has Ray Bans, Oakley, and Limitless wearables. They're doubling down on hardware and pivoting away from open source AI.

AI Data Centers in Space: Not Science Fiction

Gavin Baker broke down why space based data centers make sense:

• Energy: Sun is 7x more powerful in space
• Cooling: Space is freezing
• Land: Unlimited vs. NIMBY politics
• Rockets: SpaceX and Blue Origin make it viable

The only bottleneck? Bandwidth. But we've solved it for satellites.

Boom Supersonic: From Jets to Energy

Boom built turbines for supersonic flight. Then realized the same tech can generate electricity for AI data centers.

They raised $300 million from Altimeter and Y Combinator to pivot into energy infrastructure. Great tech, unexpected demand, funded vision.

Kalshi: America's Youngest Female Billionaire

The Kalshi founder (PolyMarket competitor) just became the youngest self made female billionaire.

Prediction markets prove backing opinions with money gets real information. Market equilibrium in action.

Construction Wages Surge 25 to 30%

Data center construction is driving electrician and plumber wages up 25 to 30%.

Private capital deploying for AI infrastructure creates labor shortages. Rising costs create more incentive to automate and invest in robotics. Market forces playing out.

The Takeaway:

The AI race moved beyond models. It's now hardware (OpenAI vs. Apple), infrastructure (space data centers), energy (Boom's turbines), and real world effects (labor shortages). Companies that can't pivot across dimensions will struggle.

What surprises you most? OpenAI's hardware push, data centers in space, or the construction wage surge?

Keen to stay ahead? Subscribe to Built 2 Scale on YouTube

In this episode, Scotty enters crazy season construction while Matt preps for a founder mode Christmas, then they dissect Sam Altman's Code Red response to Google's dominance by hiring 40 Apple hardware engineers. The guys explore whether this signals an Apple acquisition setup or a play for the operating system layer, why Meta's Limitless acquisition with zero notice shows Zuckerberg scrambling without a clear vision, and Boom Supersonic's brilliant $300M pivot from jet turbines to natural gas energy for AI data centers. Plus: Why construction wages jumping 30% accelerates the robotics timeline, Travis Kalanick automating 300 bowls per hour at Cloud Kitchens, and the emerging Private Equity AI playbook of buying traditional businesses and injecting AI to 10x margins.

Built 2 Scale | Episode 32

TIMESTAMPS:

0:00 Crazy Season Construction & Founder Mode Christmas

2:57 Bill Ackman's "May I Meet You" Dating Advice Goes Viral

7:10 Scotty's Stock Picks: Google Hits $4 Trillion

8:28 Jensen on Joe Rogan: 4,000 Emails a Day, No Ice Baths

11:15 OpenAI Code Red: Hiring 40 Apple Hardware Engineers

15:38 Is Sam Setting Up an Apple Acquisition?

21:01 Meta Acquires Limitless Pendant With Aggressive Privacy Changes

23:41 Zuckerberg's Vision Problem vs Elon's Clarity

33:51 Data Centers in Space: Unlimited Solar & Free Cooling

38:52 Boom Supersonic's $300M Pivot: Jet Turbines to Energy

46:33 Construction Wages Up 30% Thanks to Data Centers

52:27 Travis Kalanick's Cloud Kitchens: 300 Bowls Per Hour

1:00:34 Tool of the Week: Alloy Prototyping Changes Sales in 3 Minutes

1:09:09 The Private Equity AI Playbook: Buy Businesses, Inject AI, 10x Margins

This Episode Covers:

• OpenAI's hardware pivot hiring 40 Apple engineers as response to Google's model and compute advantage
• Meta acquiring Limitless with aggressive policy changes signaling Zuckerberg's lack of clear AI vision
• Boom Supersonic raising $300M by pivoting jet turbines into natural gas energy for AI compute bottleneck
• Data centers moving to space for unlimited solar power, free cooling, and no environmental opposition
• Construction wages up 30% from data center demand accelerating automation and robotics investment
• Travis Kalanick's Cloud Kitchens automating food at 300 bowls per hour with point solution robots
• Alloy prototyping tool turning 3 hour design mockups into 3 minute AI powered iterations
• The Private Equity AI playbook: Acquire traditional businesses, inject AI logic, transform 10% margins into 30%

KEY INSIGHTS:

• OpenAI's existential threat: Google has better models, cheaper TPU compute, and 100x more data. Hardware pivot either sets up Apple acquisition with Sam as CEO or prepares for OS layer competition
• Meta's strategic confusion: Zuckerberg can't articulate clear five year vision like Elon does with multi planetary life and truth seeking AI. Scrambling with acquisitions instead of building coherent strategy
• Boom's antenna advantage: CEO Blake Scholl heard AI compute energy bottleneck and pivoted jet turbine tech to natural gas generation. Raised $300M solving bigger problem than supersonic flight
• Space data centers unlock: Seven times more solar in orbit, free cooling, unlimited land, no NIMBY opposition. Elon's rocket monopoly plus Starlink bandwidth makes him infrastructure layer winner
• Wage surge validates robotics: 30% construction wage increases from data center labor demand creates bigger ROI case for automation than any efficiency argument
• Alloy's 10x improvement: Non technical product managers mock interactive prototypes in 3 minutes vs 3 hours with Figma. Game changer for sales demos and client feedback loops
• PE AI arbitrage opportunity: Better ROI buying traditional 10% margin businesses and injecting AI automation than competing in crowded software markets. New age of private equity leverage

About Built 2 Scale:

Weekly analysis of artificial intelligence business strategy, robotics adoption, and technology market dynamics. Hosted by Scott Wilcox and Matt Perrott, providing actionable insights for founders, investors, and technology executives navigating the AI transformation.

Subscribe for AI business intelligence: /@built2scale

Available on Spotify | Apple Podcasts

#AI #OpenAI #Meta #BoomSupersonic #CloudKitchens #PrivateEquity #Alloy #DataCenters #AIStrategy #Built2Scale #HardwarePivot #TechNews

BUILT 2 SCALE | TOOL OF THE WEEK | Episode 31 - November 28, 2025

Every week, Scotty and Matty break down the strategies, frameworks, and mental models that separate the world's best builders from everyone else.

This week's tools:

1. Embrace the Flywheel Effect

Your success operates like a business flywheel. All of your history and context from previous operations feeds into making your new product even better. Each win compounds the next.

2. Learn by Doing

Take a page from James Dyson's playbook: "Don't think too long about doing things, just go out and do them." When faced with a question, just go do it.

3. Curate Your Content Diet

Replace mindless scrolling with high-value content. Be ruthless about your triggers. If needed, start a fresh social media account focused on your domain expertise or entrepreneurship. Make your scrolling work for you.

4. Know Thyself Before You Hire

Ask yourself: What am I good at? Why me? What skills do I need around me? This clarity reveals whether you need a co-founder and helps identify your superpower.

5. Hire for Your Weaknesses

Hire for what is absolutely not your superpower. Early SpaceX had great rockets but was about to die until Elon hired a VP of sales to secure government contracts. Let go of the vine on non-superpower areas.

6. Clarify Your Vision in Writing

Put your vision in writing so new hires can take it away and explain it to their friends and family. As Mark Andreessen says, a CEO's greatest skill is the ability to tell a story. This attracts customers, talent, and capital.

7. Prioritize Foundational Roles

In the first three years of scaling, nail these three categories: Product (CTO), Brand (CMO), and Distribution (Head of Sales or Growth). Everything else can wait.

8. Block Out Thinking Time

Your role as a founder is to set the vision, not be completely operational. Block out sections of days with no tasks to do. Use this time to think about the vision or pivot into solving business problems.

Which tool resonates most with where you are right now?

Keen to stay ahead? Subscribe to Built 2 Scale on YouTube

BUILT 2 SCALE | AI NEWS | Episode 31 - November 28, 2025

Every week, Matty and Scotty cut through the noise to bring you the AI developments that actually matter: the moves reshaping markets, the strategies redefining competition, and the shifts you need to understand to stay ahead.

Google's Gemini 3 Adds $2 Trillion to Market Cap—The Age of Scaling is Over

The AI landscape just shifted from a compute arms race to a battle for ecosystem dominance, custom silicon, and real-world intelligence.

The Vertical Integration Play

Google spent a decade building proprietary TPU chips, and it just paid off. By cutting Nvidia dependency entirely, they can now out compete on cost per token. The new race? Token per watt efficiency. Google just took the lead.

Ecosystem = Moat

Gemini 3 isn't just competitive with OpenAI and Claude. It's natively integrated across Pixel, Google Docs, YouTube, and every product in the Google suite. When your model is "at par or better" AND built into tools people use daily, distribution becomes your unfair advantage.

Real-World Intelligence Takes Center Stage

Gemini 3 Pro understands 3D context, turning sketches into renders and photos into floor plans. It actively "watches" YouTube clips instead of just reading transcripts. The training data advantage? Unbelievable.

Industry consensus is clear: top minds (including Ilya from OpenAI) say "the age of scaling is over." The next frontier demands:

→ Reduced energy consumption

→ Real-world spatial intelligence

→ Physical applications beyond screens

The Three-Layer Strategy

Musk's playbook tackles all constraints simultaneously:

• Real-world data → Tesla fleet
• Energy → Tesla batteries & solar
• Compute → Custom chips with Samsung

This is full-stack AI competition.

OpenAI's Move

To justify their valuation, OpenAI must expand into memory, personalized UI, and consumer apps (payments, shopping). The bet: LLMs "can get into everything in your life."

The Niche-Down Imperative

If you're building on foundational models: specialize or die. Google and OpenAI offer such broad capability that billion dollar companies must carve defensible niches with specialized workflows or get priced out.

Geography Matters Less

Silicon Valley's premium only applies to cutting edge AI research. For companies leveraging models intelligently or scaling GTM? Austin, NYC, Denver work fine.

The Takeaway:

AI competition evolved into a multi-dimensional battle: custom silicon, ecosystem lock in, real-world data, energy efficiency. Companies that can't compete across dimensions must niche down fast.

What's your take? Are we past the age of scaling?

Keen to stay ahead? Subscribe to Built 2 Scale on YouTube (link in comments).

In this episode, Scotty debates whether cricket on office TVs kills productivity or builds culture, while Matt navigates Thanksgiving week shutdowns in Austin where the entire tech economy grinds to a halt. They dissect the seismic shift happening in AI infrastructure as Google's Gemini 3.0, trained entirely on TPUs, proves you can bypass Nvidia's 75% margins while building world class models. The implications are staggering. From vibe coding startups getting bundled out overnight to the "age of scaling is over" consensus among top researchers, they explore whether there's room for multiple frontier models, why Marc Andreessen's "Silicon Valley is everything" take misses the mark, and the critical hiring mistakes founders make in their first three years.

Built 2 Scale | Episode 31

TIMESTAMPS:

0:00 Thanksgiving Shutdowns & Cricket in the Office Debate

4:12 WeWork Economics & Culture vs Productivity Balance

8:14 Google Gemini 3.0: The TPU Strategy That Changes Everything

13:18 Google vs Nvidia: Token Per Watt Economics & Market Impact

16:59 Vibe Coding Apocalypse: How Gemini Beat Lovable in One Day

21:20 Multi Model Future: Claude, ChatGPT & Gemini Strategies

26:46 Ilya's Bombshell: "The Age of Scaling is Over"

31:51 Real World Data: The Next AI Frontier Beyond 2D Training

36:49 Marc Andreessen vs Reality: Do You Really Need Silicon Valley?

42:16 Distributed Teams: Time Zone Hell & The Remote Work Debate

47:02 Tool of the Week: Founders Podcast (400+ Biographies Distilled)

51:38 Screen Time Hacks & Content Diet Optimization

56:26 The 3 Critical Roles to Nail When Starting a Business

1:03:08 SpaceX Lesson: Why Elon Nearly Died Without a VP of Sales

1:06:29 Vision Distribution: Writing It Down vs Giving Speeches

This Episode Covers:

• Google's Gemini 3.0 TPU training strategy and what it means for Nvidia's margins
• Why cost-per-token economics matter more than benchmark scores
• The vibe-coding startup extinction event: Lovable vs Gemini in one day
• Claude Opus 4.5 release and Anthropic's coding-first AGI thesis
• Multi-model future: Room for Google, OpenAI, Anthropic with different strategies
• "Age of scaling is over" consensus from Ilya, Yann LeCun, Demis Hassabis
• Real-world data and spatial intelligence as the next AI breakthrough
• Marc Andreessen's Silicon Valley claim vs distributed global talent reality
• Time zone brutality and why AR/VR won't fix remote work
• Tool of the Week: Founders Podcast distilling 400 biographies into patterns
• The 3 critical roles to nail when starting a business (two frameworks)
• SpaceX lesson: Vision in writing scales, speeches don't

KEY INSIGHTS:

• Google's economic warfare: TPU training creates structural 50% cost advantage vs Nvidia-dependent competitors—forces token price matching while OpenAI pays premium margins
• Vertical integration checkmate: When Google reaches model parity, ecosystem lock-in (Docs, Search, Android, YouTube) becomes insurmountable moat
• The bundling massacre: Gemini beating Lovable in one day signals what Microsoft did to Zoom with Teams—horizontal players will bundle out vertical startups
• Anthropic's focus moat: All-in on coding creates talent magnet and defensible niche while Google/OpenAI serve billions horizontally
• Scaling plateau is real: GPT-3→3.5→4 showed diminishing returns—top researchers (Ilya, Yann, Demis) agree architectural breakthroughs needed, not just more compute
• Real-world data frontier: 2D screen training has plateaued—spatial intelligence from IoT/sensors/robotics is where next breakthroughs happen
• Geographic arbitrage reality: SF only necessary for cutting-edge AI engineering talent—successful AI-enabled companies can thrive elsewhere with go-to-market teams
• Vision must be written: If your vision requires you giving speeches, you'll never scale—write it down so others can distribute it
• SpaceX nearly died from sales neglect: Even Elon's rocket vision wasn't enough without VP of Sales unlocking government contracts
• Hire for weaknesses, not strengths: Technical founders need sales/marketing, sales-driven founders need technical depth—balance matters more than doubling down
• Council of Models concept: Andre Karpathy building aggregator that queries all models, compares responses, synthesizes best answer—future of consumer AI
• Token-per-watt is the only race: Reducing energy per unit of intelligence matters more than benchmark leaderboards for long-term AI economics

About Built 2 Scale:

Weekly analysis of artificial intelligence business strategy, robotics adoption, and technology market dynamics. Hosted by Scott Wilcox (Australia) and Matt Perrott (Austin), providing actionable insights for founders, investors, and technology executives navigating the AI transformation.

Subscribe for AI business intelligence: /@built2scaleAvailable on

Spotify | Apple Podcasts

In this episode, Scotty takes FSD to hurry-mode across San Francisco while Matty battles jet lag with his toddler, and both come back with insights on what really matters in 2025's AI landscape. From crashing VC offices unannounced to exposing the AI consultant charlatans at LA's business summit, they tackle the shift from startup hustle culture to sustainable founder life. Featuring deep dives on ChatGPT's Pulse feature, the explosion of AI slop, why Australia's tech lag might actually be an advantage, and why the startup community needs a "Pensions, Yachts and Families" club for founders who want to build empires without sleeping in the office.

Built 2 Scale | Episode 30

TIMESTAMPS:

0:00 Back in Action: Cross-Continental Travel Updates

0:42 Formula One Pace: Solo Travel vs. Family Speed

1:19 Mad Max Mode: Full Self-Driving Hacks in San Francisco

3:47 Cold Calling Silicon Valley: Showing Up at VC Front Doors

5:05 The Bogan Accent Advantage: AI Can't Fake Aussie Yet

6:22 LA Business Summit Letdown: Celebrity Watching vs. Real Innovation

7:21 AI Consultant Charlatans: 65-Year-Olds Promising AI Transformation

1:03:06 ChatGPT Pulse: Sam Altman's Favorite Product Feature

1:04:24 Wallet Reduction Section: Which AI Subscriptions Are Worth It?

1:05:08 The $300/Month AI Subscription Problem

1:05:34 Matt's Startup Rant: Where's the Founder Parent Community?

1:07:23 Pensions, Yachts and Families: The Anti-Hustle Founder Club

1:08:28 Yann LeCun as Scott's Future CTO?

This Episode Covers:

- Tesla FSD Hurry mode and Formula One-speed solo travel hacks

- Why showing up unannounced at VC offices beats AI-generated pitch decks

- The AI consultant charlatan problem plaguing business conferences

- How Australia's tech lag creates a unique accent-based moat against AI

- ChatGPT's Pulse feature and why it's worth the Pro subscription

- The explosion of AI slop and how to cut through the noise

- Why most AI subscriptions at $300/month aren't worth it for non-coders

- The missing community for founder parents who refuse hustle culture

- Jeff Bezos's three-decisions-a-day philosophy for mature founders

- Why Y Combinator's rage bait culture is failing responsible adults

KEY INSIGHTS:

- Physical presence at VC offices creates authenticity AI can't replicate. Video proof of being there beats any AI-generated pitch

- ChatGPT's Pulse feature represents the future of proactive AI assistants that work in the background on your behalf

- The Bogan Australian accent is an accidental moat. Not enough training data for AI models to fake it convincingly

- Startup culture needs a counterbalance to 20-year-olds sleeping at the office. Successful founders can build while being present parents

- AI subscriptions are consolidating. Most founders only need one premium tier, not three $300/month products

- The AI consultant wave is filled with people who don't understand the technology they're selling

- "Pensions, Yachts and Families" beats hustle porn. Sustainable founder lifestyles build better long-term companies

About Built 2 Scale:

Weekly analysis of artificial intelligence business strategy, robotics adoption, and technology market dynamics. Hosted by Scott Wilcox (Australia) and Matt Perrott (Austin), providing actionable insights for founders, investors, and technology executives navigating the AI transformation.

Subscribe for AI business intelligence: /@built2scale

Available on Spotify | Apple Podcasts