In partnership with
Hello {{first_name|Motivated and Miffed Community}},
If 2024 was “models,” and 2025 was “agents,” 2026 is turning into “infrastructure + legitimacy.” Not just what the models can do—but where they run, who approves them, and who pays the energy bill.
✅ TL;DR
🏗️⚖️ The compute boom is colliding with permitting, climate scrutiny, and local politics.
🌍🤝 OpenAI is pitching a country-by-country adoption playbook (and nudging governments toward more data centers).
⚡💰 Investors are rotating from “AI winners” to the companies that sell the megawatts and build the pipes.
Today’s Sponsor
Introducing the first AI-native CRM
Connect your email, and you’ll instantly get a CRM with enriched customer insights and a platform that grows with your business.
With AI at the core, Attio lets you:
Prospect and route leads with research agents
Get real-time insights during customer calls
Build powerful automations for your complex workflows
Join industry leaders like Granola, Taskrabbit, Flatfile and more.
🧠 AI News
1) The UK’s AI data center push just hit a legal wall
The UK government admitted its approval for a major AI data center in Buckinghamshire should be overturned after failing to properly account for environmental impacts.
This is the physical layer of AI finally acting like… the physical layer. Data centers aren’t “software.” They’re land use, grid load, water draw, traffic, and years of construction. When approvals get quashed, it’s not a minor delay—it’s a reminder that compute has neighbors.
And this is where the narrative is shifting: it’s not only “can we build it?” but “can we justify it?” Water is becoming a headline constraint alongside power—Texas, for example, is already modeling data-center water demand as a meaningful slice of statewide usage by 2030.
Why it matters: The next bottleneck isn’t GPUs—it’s permission. Zoning, environmental review, and public trust are becoming real competitive variables.
2) OpenAI is going “country mode” at Davos
OpenAI is expanding “OpenAI for Countries,” pitching governments on AI adoption in education, health, and disaster preparedness—while also encouraging more data-center buildout.
This is a savvy reframing: instead of “here’s a chatbot,” it’s “here’s a national capability.” The reported examples are telling—ChatGPT Edu embedded into Estonia’s secondary schools, and infrastructure partnerships in places like Norway and the UAE, with OpenAI as an early anchor customer.
The deeper play is distribution and default settings. If you’re the platform that helps a government modernize services, you become the standard—and standards are sticky in a way consumer hype isn’t.
Why it matters: AI leadership is starting to look like industrial policy: capacity, procurement, partnerships, and “who gets there first” relationships.
3)The market is betting on power and pipes, not just the model makers
BlackRock says investors are leaning toward energy providers and infrastructure plays over mega-cap tech as their preferred way to ride the AI wave into 2026.
The numbers in their survey are blunt: only about a fifth picked big U.S. tech as the most compelling AI opportunity; more than half favored power providers, and a large chunk picked infrastructure.
That’s not anti-AI. It’s a recognition that the “AI trade” is broadening from apps to the rails: generation, transmission, cooling, construction crews, and anyone who can reliably deliver uptime at scale.
Why it matters: The AI economy is starting to price like a buildout. When the gold rush matures, the durable winners often sell shovels—and electricity.
📱 From the Feed
How would you rate this newsletter?
👋 That’s All
AI is leaving the lab and negotiating with reality: courts, grids, water tables, and governments. The next decade won’t just be about smarter systems—it’ll be about who can run them credibly.
Stay MOTIVATED,
Gio