Welcome to Issue 161 of The CTO Show Brief.
The Physical Layer Becomes the Only Layer
This week confirmed what has been building for two years. The primary competition in AI and infrastructure is no longer for models or software. It is for energy, optics, sovereign alignment, and physical compute capacity. Five deals and one government action made the structural shift undeniable.
SIGNALS
The Circular AI Economy Is Now Structurally Closed
OpenAI closed $110B at an $840B valuation, led by Amazon ($50B), Nvidia ($30B), and SoftBank ($30B). The architecture of the round is the signal. Nvidia supplies the chips. OpenAI uses Nvidia capital to buy those chips. Amazon hosts the infrastructure that OpenAI pays for. This is a self-reinforcing loop, not a funding round.
The deeper shift: AWS gets stateful runtime environments for persistent agents while Azure retains stateless API traffic. OpenAI is deliberately fragmenting its cloud dependencies to prevent any single provider from holding leverage over its AGI roadmap. Founders building agent middleware that bridges stateless and stateful environments are sitting on a structural gap.
Who gains leverage: Infrastructure orchestration startups. Vertically specialized AI players with proprietary data. Anyone simplifying multi-cloud agent persistence.
Headcount Is No Longer a Proxy for Scale
Block cut 40% of its workforce, roughly 4,000 roles, citing agentic AI as the operational substitute. Klarna, Amazon, Salesforce, and HP have posted similar moves. The market rewarded Block with a share price increase on the announcement day.
This is not cost-cutting dressed up as transformation. It is the first wave of companies genuinely reorganizing around AI orchestration rather than human management layers. The valuation implication is direct: revenue-per-employee is replacing headcount growth as a proxy for scale. Startups running lean from day one have a structural advantage over legacy orgs now dismantling their own floors.
Secondary risk: AI-generated code shipping at scale without adequate security vetting. As 30% of enterprise code comes from AI tools, the attack surface is expanding faster than the defenses. Cybersecurity vendors focused on AI-generated code auditing are entering an uncongested market.
Who gains leverage: Agent-first GTM vendors. Automated code security auditors. Founders benchmarking on output per person, not team size.
AI Safety Is Now a Federal Compliance Category
The US government designated Anthropic a supply-chain risk to national security after Anthropic's CEO declined to waive policies prohibiting autonomous weapons use. This is the first time the US has applied a label typically reserved for Huawei or ZTE to a domestic firm for ethical non-compliance with defense contract terms.
Within hours, OpenAI secured a Pentagon classified network deal under a negotiated but permissive safety framework. xAI reached a separate classified deployment agreement. The competitive dynamic is now explicit: firms willing to align with the Hegseth Doctrine get access to the largest single infrastructure buyer on earth. Firms that hold safety lines get locked out of federal capital.
The second-order implication for the region: Gulf sovereigns now face a binary. Do their national AI strategies follow the permissive OpenAI model or build sovereign alternatives like Falcon that operate outside US defense-use constraints entirely?
Who gains leverage: Dual-use AI vendors. Sovereign cloud architects. Startups building AI governance frameworks that navigate this new compliance environment.
Energy Is Now the Compute Moat
A BlackRock-led consortium acquired AES Corporation for $33.4B, taking private the world's largest corporate clean energy supplier. AES has 11.8GW of signed power agreements with major tech firms. The deal is 100% equity-funded. Qatar Investment Authority is a co-underwriter.
The logic: public markets cannot provide the patient capital needed to build gigawatt-scale AI power supply. Private infrastructure funds can. The right to reliable power is now priced above the right to compute. Any AI data center project without a secured power purchase agreement is effectively stranded.
QIA's presence is the MEA signal. States that control energy are using that leverage to position as co-owners of the AI infrastructure stack, not just investors in it.
Who gains leverage: Microgrid and SMR startups. PPA-secured data center operators. Infrastructure funds with energy expertise.
Physical Interconnects Are the New Chip Race
Nvidia invested $2B in Lumentum to secure domestic US supply of silicon photonics, the technology that moves data between AI chips using light instead of electricity. Jensen Huang framed it as building the physical foundation of gigawatt-scale AI factories. The focus on US-based manufacturing is not incidental. It is a direct response to the national security logic now shaping the entire hardware stack.
The bottleneck in AI scaling has moved from compute to interconnect. Copper cannot carry the bandwidth required at gigawatt scale efficiently. Whoever controls the optical layer controls the energy budget of every major AI training run. This is a hardware IP race that most software-focused investors are not tracking yet.
Who gains leverage: Advanced semiconductor packaging startups. Thermal management and optical integration specialists. Infrastructure vendors building workload orchestration for optical networks.
MTN Re-Integrates Its Towers. Asset-Light Is Finished in Emerging Markets.
MTN announced a $6.2B acquisition of the remaining 75% of IHS Towers, taking private 29,000 African tower sites. The deal reverses a decade-long strategy of selling towers to independent operators. Currency devaluation, particularly the naira shock, made dollar-denominated tower leases a structural liability.
Each of those towers is now a candidate for edge AI compute. MTN is not just buying back its own infrastructure. It is positioning as a continent-wide distributed data center operator. For African founders building low-latency or offline-first applications, the access dynamics are about to change significantly.
Who gains leverage: Edge AI startups in Africa. Infrastructure operators building on top of telco-owned assets. Investors in vertically integrated emerging market digital infrastructure.
The question shaping the next 18 months: which founders are building at the physical layer from the start, and which are still assuming someone else owns it?
🎙️Episodes Recap:
In this episode of The CTO Show with Mehmet, Mehmet Gonullu sits down with Betsy Atkins , serial entrepreneur, board member of global companies, and advisor to leading organizations including Google Cloud. The conversation explores how the role of the board is evolving in an era defined by AI, agentic systems, and accelerating technological change. Betsy shares deep insights on how boards must move from passive oversight to active orchestration, the risks introduced by agentic AI, and what leaders must do today to stay relevant. From “corporate cholesterol” slowing down organizations to the emergence of AI governance frameworks, this episode offers a candid look at the challenges and opportunities shaping the next generation of leadership.
In this episode of The CTO Show with Mehmet, Mehmet sits down with Dr. Bill Andrews, a renowned molecular geneticist and pioneer in telomere research. Dr. Andrews has spent decades studying the mechanisms of aging and is known for leading discoveries around telomerase, the enzyme connected to cellular aging. The conversation explores the biological root causes of aging, the role of telomeres in limiting human lifespan, and the scientific pursuit of extending human healthspan. Dr. Andrews also shares insights into the challenges of biotech innovation, the economics of pharmaceutical research, and why breakthroughs in longevity science often struggle to reach the public. This episode bridges biology, technology, and the future of human health, offering a deep dive into one of the most fascinating scientific frontiers of our time.
📖 From Nowhere to Next
Every week I share startup lessons and stories through The CTO Show Brief. But if you want to go deeper, my book From Nowhere to Next brings together the experiences and insights that shaped my own journey.
Thanks for reading — and for being part of this growing, global-minded network.
— Mehmet
