The signal worth reading.
A working library of the research, reporting, and primary sources shaping our investment thesis — what AI changes, what it can't, and where the durable opportunity lives.
How AI is reshaping labor & output
Frontier-lab research and economist forecasts on what generative AI does to GDP, white-collar work, and the shape of demand over the coming decade.
Measuring the labor-market impacts of frontier AI
Anthropic's first-party study of how Claude is actually being used at work — which tasks are being augmented, which are being substituted, and what that says about which roles compress first.
Generative AI could raise global GDP by 7%
The widely-cited macro frame: a one-time productivity step-change worth ~$7T over a decade, concentrated in advanced economies — and unevenly across sectors. Useful as a reality check on top-down market sizing.
How will AI affect the U.S. labor market?
The companion labor read: estimated exposure share by occupation, and where automation is most likely to land first. We use the exposure map in reverse — to find the businesses on the safer side of the curve.
Generative AI will have greater impact on white-collar work
Pearson's task-level analysis arguing that knowledge-worker roles face disproportionate disruption versus skilled, hands-on, customer-facing roles. The asymmetry sits at the core of our hedge.
AI-resistant work & the skilled-trade reset.
The other side of the labor curve: hands, hard assets and on-site execution. Reporting on the structural shortage of skilled workers and the categories where shovels matter more than software.
Inside the AI-proof world of skilled trades
While desk jobs absorb the disruption, electricians, plumbers, HVAC techs and other trades are facing a generational shortage that isn't going to be modeled away. Wages and franchise economics keep climbing.
The AI data-center buildout is creating a skilled-trade gold rush
The same companies pouring capex into compute need humans to wire it, cool it, and pour the concrete around it. Trade-services platforms with the labor pool are pricing power they haven't had in decades.
Shipbuilding, manufacturing and the AI-proof Gen-Z opportunity
Reshoring, defense industrial policy and a retiring workforce are converging on a labor problem that AI cannot solve. Wages, training programs and political tailwinds all point the same direction.
Bridging the trade gap — can Gen Z save America's skilled workforce?
An industry-side view of the same shortage: apprenticeship, retention and the slow shift in cultural attitudes toward the trades. Useful color on operator dynamics inside any trade-services roll-up.
AI's physical footprint: power, land & metal.
The compute boom is also a real-estate, energy and infrastructure boom. The picks-and-shovels — and the constraints — show up here first.
Energy and AI — Energy demand from AI
The International Energy Agency's framework for how AI workloads translate into electricity demand at the grid level. The supply-side ceiling on the buildout is going to define winners and losers.
AI to drive 165% increase in data-center power demand by 2030
The forward curve on data-center electricity, with implications for power generation, transmission, cooling, and the real-asset businesses that sit downstream of all of it.
Where AI augments, but doesn't replace.
Categories where trust, judgment, regulation, or in-person presence keep humans in the loop — and where AI tools become accelerants for incumbent operators, not their replacement.
4 human financial-services activities AI can't do
A grounded look at advisory, fiduciary judgment, complex relationship work and trust-bound services — the places financial-services AI hits a hard ceiling. Translates directly to wealth, advisory and high-touch B2B.
What AI agents can — and can't — do for healthcare
Agentic AI as an operational accelerant for clinics, payers and providers. Where the productivity is real, where the regulatory and clinical gates remain, and what that means for healthcare-services operators.
The customer funnel is gone — and trust is the new moat
As AI flattens distribution and pricing transparency, brand, trust and recurring relationships are doing more of the work. The argument for businesses with real customer affinity, not just acquisition spend.
Longevity — the next big opportunity to drive revenue growth
The wake of the AI wave: a time-rich, productivity-rich consumer rotates spend toward health, wellness, hospitality and experience. The case for hospitality-adjacent assets riding the longevity tailwind.
Why the old playbook is breaking.
The state of private equity in late cycle: stuck portfolios, slow exits, and the new operating playbooks emerging on the other side. Why a different posture is needed for the next decade of vintages.
Why private equity is stuck with zombie companies it can't sell
The exit drought, the DPI problem, and the structural mismatch between five-year fund cycles and the operating timelines transformations actually require. Sets up the case for longer holds and embedded operating support.
Private equity in the doldrums — and out of favor
An LP-side critique of where the asset class has under-delivered, and which strategies are losing institutional appetite. Useful pressure-test on the current consensus.
The new AI roll-up investment strategy
How a new generation of investors is buying traditional services businesses and re-platforming them with AI as the operating edge — fragmented industries, real customers, and a software-grade margin lift on the other side.
We don't trade headlines. We read the wake — and back the operators building in it.