AI Summary
5 min readIn 2023, everything changed. BlackRock's Tony Kim describes the shift from the internet era to the AI era as a transition akin to going from BCE to AD. The base unit of compute went from a $10,000 server to a million-dollar one. The market cap distribution flipped: before AI, software and services dominated; today, chips and hardware command roughly $30 trillion in market cap, compared to $10 trillion in software and $22 trillion in the Mag 7. Kim’s central argument is that we are living through a complete rebuild of the data center, the compute stack, and the value chain—and most people haven't fully absorbed what that means.
The Primacy of Compute and the Data Center Redesign
The old cloud model was built on cheap, commodity CPUs and hard drives, with value concentrated in high-margin SaaS applications. AI inverted that. Now, the compute layer is massively expensive—a single server can cost millions—and the business model revolves around selling tokens rather than software subscriptions. This shift forces a complete rethink of the data center itself.
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
Never miss an episode of Sourcery
Get every new episode summarized in your inbox — free, ~5 minutes to read.
No spam. Unsubscribe anytime.
What you'll learn
- 1 (00:00) **Opening: The Shift to a Compute-Centric World** - Tony Kim introduces the core macro thesis: the AI era has flipped the tech stack from software-centric to compute-centric, with massive market cap implications.
- 2 (03:52) **The Three Macro Themes of the AI Era** - Kim outlines the key through-lines for the AI infrastructure buildout: compute primacy, data center redesign, and the physics of power density.
- 3 (08:09) **The Co-Design Revolution: Chips and Models** - The new path forward is the tight integration of silicon design with model parameters, a concept Kim will discuss with Broadcom.
- 4 (10:41) **The "RAMpocalypse" and the Primacy of Memory** - The AI era is creating a massive shortage of memory (RAM), driven by models that increasingly need to store and recall information.
- 5 (14:22) **The Complete Rebuild: From Cloud to AI Factories** - The old cloud (reselling CPUs with hard drives) is being replaced by a new model: AI factories that sell tokens.
- 6 (21:10) **Playing Offense: A New Investment Philosophy** - The complete rebuild requires a fundamental rethink of investment strategy, aligning capital allocation with the new reality of intelligence as the core product.
- 7 (27:37) **Capital Allocation: The Now vs. The Tomorrow** - Kim explains his framework for balancing short-term (3-year) investments with long-dated, asymmetric bets on the frontier.
+ Full timestamped outline available in the app
Show Notes
Tony Kim is Managing Director and Head of the Global Technology Team within Fundamental Equities at BlackRock. Recorded at the RAISE Summit in Paris, where Tony spoke on 4 panels, next-gen accelerators with d-Matrix, quantum computing with PsiQuantum, optics in data center design with Lumentum, and XPU and AI chip co-design with Broadcom.
Tony breaks down the shift from a software-centric world to a compute-centric one, and why roughly $1 trillion of CapEx this year, and $10 trillion over the next 5 years, is being spent to move data centimeters and millimeters. He maps the market cap transformation, roughly $10 trillion in software, services, and internet, $20 trillion in Mag 7, and $30 trillion in non-Mag 7 chips and hardware.
We cover the RAMpocalypse and why memory intensity is skyrocketing as AI models start to mirror the human brain, the 3 to 4 year duration mismatch between fab buildouts and today's demand, and the move from copper to light inside the data center.
Tony also lays out his capital allocation framework, 90% in the 3-year AI vortex and the rest on frontier bets like quantum, orbital data centers, SMRs, and 800-volt power architectures, all converging on 2030. Plus, the coming wave of 30 to 40 Chinese robotics IPOs, mixing Chinese robot bodies with Western brains, his contrarian case for social robots addressing loneliness and aging populations, token flow as the new enterprise framework, and the mentors who shaped his career.
We cover:
› Why AI is forcing a complete rebuild of data centers
› The shift from software to compute
› Why memory becomes the next critical bottleneck
› The future of AI chip co-design
› How investors should think about AI infrastructure
› Quantum, orbital data centers and the road to 2030
› Why robotics could become AI's next trillion-dollar market
› China's robotics advantage
› The enterprise AI stack and "token flow"
› How BlackRock evaluates long-term technology investments
Tony Kim: https://www.linkedin.com/in/tony-kim-3150053/
Molly O’Shea: https://x.com/MollySOShea
Sourcery: https://x.com/sourceryy
𝐄𝐏𝐈𝐒𝐎𝐃𝐄 𝐋𝐈𝐍𝐊
YouTube : https://youtu.be/zXB-LI7skL8
𝐒𝐏𝐎𝐍𝐒𝐎𝐑𝐒
• Brex—The modern finance platform, combining the world’s smartest corporate card with integrated expense management, banking, bill pay, & travel. https://brex.com/sourcery
• MongoDB–Millions of developers a
More from this podcast
Sourcery →