Thrive's Four-Category Map
lindo@mac_matz
The idea
I believe four forces define this era of AI: the emerging platforms building it, the scarce inputs powering it, the data-rich businesses being rewired from inside by it, and the human-rooted assets gaining value precisely because intelligence gets cheap. I built four equal-weighted layers holding only the clearest leader in each category, avoiding single names I can't access directly where public proxies make more sense, and kept 10% in cash. I'm holding for years, and I'd rethink a layer if a leader lost its edge, not if a quarter looks noisy.
What will be bought
- The take bets that chips, memory and power become the limiting factors as AI scales, and whoever controls these scarce inputs captures the most value. · Taiwan Semiconductor Manufacturing CompanyIt makes nearly all of the most advanced AI chips, so it controls capacity when leading-edge chip production is still the bottleneck for AI growth., Micron TechnologyIt makes fast memory at scale alongside only two other companies, so it controls a scarce input that AI data centers must have in huge amounts., GE VernovaIt builds the turbines and grid equipment that power data centers, and power becomes the real limit on how fast AI infrastructure can grow.23%
- The take holds that brands, live experiences and community grow more valuable as intelligence and content become abundant, making scarcity itself the asset. · Ferrari N.V.It sells identity and scarcity through long waitlists and rising prices, which the take expects becomes more valuable as intelligence gets cheaper., Madison Square Garden SportsIt owns trophy teams whose value comes from fans and culture in finite supply, which the take expects grows as in-person community becomes scarcer., Live Nation EntertainmentIt runs concerts and ticketing, the kind of live experience that software cannot replace and that people will keep paying more to attend in person.23%
- The take includes companies building the new platforms seen as most important, owned directly or through their public shareholders. · Alphabet Inc.It owns Isomorphic Labs and DeepMind, putting it among the few companies building frontier AI and the platforms others will build on., Microsoft CorporationIt is the closest public way to own frontier AI through its stake in and partnership with OpenAI, making it the platform play on that partnership., Space Exploration Technologies Corp.It builds Starlink and launch capability as platforms that others can build on, making it one of the foundational infrastructure plays in the take.23%
- The take bets that companies with proprietary data and workflows get stronger with AI, reversing how the market priced them this year. · Thomson Reuters CorporationIt owns Westlaw and other legal and tax data that professionals renew every day, so AI applied to that data makes existing relationships stronger., Intuit Inc.It owns TurboTax and QuickBooks where millions already work, so it can turn that daily-use data into AI features customers will pay to keep., Verisk AnalyticsIt owns decades of insurance claims and risk data that insurers cannot easily get elsewhere, so AI running on that data makes it more valuable to its customers.23%
- Cash10%