The Ones Who Already Cleared It
86% never energize. We went from order to commissioning in 8 months. Here's the receipt.
AI Of The Coast - Jiri Fiala
AI of the Coast gives family offices and private equity firms an early-warning edge on AI infrastructure—so they can price risk better, see value shifts sooner, and move capital before consensus does.
Live Last read · last published · next check
Latest posts
Who Gets the Queue Repriced First
Capacity got repriced once. It's about to happen again, and not to everyone.
The One Line Every AI Infrastructure Checklist Is Missing
The checklist has to start before land, not after.
The 86% That Never Energize
The failure rate is already priced into the survivors.
You're Underwriting the Wrong Risk
The number diligence decks skip predicts failure best.
The Queue Is the Product
Compute capacity is worthless until it clears interconnection.
The $2,000 Superintelligence, and the Crack It’s Pointing At
A reasoning engine that runs frontier intelligence on a CPU for two grand a month would end the GPU arms race overnight...if it existed.
Prediction 13 Revisited: I Said Healthcare Would Pull Serious AI Investment and AI-Plus-Blockchain Would Find Real Trust Use Cases. Healthcare Roared. Crypto's AI Angle Got Quieter and Realer.
Healthcare delivered loudly; the blockchain-trust angle delivered quietly, reframed around AI provenance.
The "AI Bubble" Callers: The Billionaires Talk, I Build.
If AI is a bubble, it's the first bubble in history where the skeptics hold the opinions and the believers hold the keyboards.
AI is Quietly Gutting the Junior Job Market.
The automation arrived on schedule. The human cost is the story I have to tell honestly.
Prediction 11 Revisited: I Said AI Video Would Power VOD and Corporate Comms While Humans Kept Strategy. Veo and Sora Became Enterprise Plumbing — and Sora's Consumer App Died Proving My Caveat.
By 2026, AI video is enterprise marketing infrastructure — and the spectacular failure of Sora's consumer app validated the 'humans keep strategy' caveat I built in.
Prediction 10: I Said Runaway Valuations Could Trigger a Correction, Oligopolies Would Form, and Cost-Efficiency Would Be the Survival Trait.
Eighteen months later, circular-financing warnings are mainstream, the oligopoly is entrenched, and the correction watch is on.
Equity for Megawatts
The Washington sovereign barter shaping the future of artificial intelligence.
Prediction 9 Revisited: I Said Energy Would Become AI's Binding Constraint and Compute Would Decentralize Toward Power. It Did. I Bet My Company On It.
Waymo's 2026 expansion map is my January 2025 prediction rendered in geofences.
Prediction 8 Revisited: I Said the Transformer Era Would Hit Architectural Fatigue and New Architectures Would Rise. The Fatigue Came. The Transformer Didn't Die.
The convergence happened. The architectural revolt is real but still insurgent, not victorious.
Prediction 7 Revisited: I Said Driverless Cars Would Live in Sunny 'Safe Zones' and Humans Would Keep Driving Everywhere Else. Look at Waymo's Map.
Waymo's 2026 expansion map is my January 2025 prediction rendered in geofences.
Prediction 6 Revisited: I Said the US Goes Security-First, China Shifts to Original Research, Europe Risks Falling Behind. Two Out of Three, and China Did It Bigger.
In January 2025 I predicted diverging national AI strategies: the US leveraging AI for national security, Here's what happened.
Prediction 5 Revisited: I Said AI Augments Workers, Mass Unemployment Is Overstated. I Stand By It. The Crack Is Entry-Level.
In January 2025 I predicted augmentation over replacement — accountants becoming analysts, prompt engineering as a hot skill, ‘Grandma AI’ passing expert wisdom to systems. Broadly correct. But a faul
Prediction 4 Revisited: I Said Humanoids Would Be Form-Factor Progress and Niche Labor, Not Conscious Robots. Figure Now Bills BMW $25 an Hour.
In January 2025 I said conscious, self-aware humanoids stay science fiction; real progress would be form factors and narrow industrial deployment, with robots augmenting human work, not replacing it.
Prediction 3 Revisited: Last Year I Said No Super-Agent, Just Orchestrated Swarms. The Entire AI Industry Just Agreed With Me.
In January 2025 I bet against the single all-knowing super-agent and for orchestrated swarms of specialized sub-agents under a central coordinator. By 2026, ‘multi-agent orchestration’ became the indu
Prediction 2 Revisited: I Said the US Would Copy Europe on AI Rules and Asia Would Sprint. The Map Redrew Itself
In January 2025 I predicted a regulatory patchwork: the US lagging and taking cues from Europe, Europe’s own bureaucracy strangling its on-the-ground rollout, and Asia fast-tracking through sandboxes.
I Made 13 AI Predictions 17 Months Ago. Here Are the Receipts.
Seventeen months ago I made 13 predictions about AI. Here’s every receipt — what I nailed, what I got backwards, and what happens next.