Start With One Recurring Report
Connect the report read-only. Then run it on a clock. The first useful AI move for a finance team is one recurring report wired to the real source.
13 min read →Blog
Essays and updates. First principles, not playbooks.
Make Your Work Easier to Review
If you cannot review it fast, you cannot delegate it safely. The first AI skill is review design: writing the standard before you generate anything.
10 min read →Compute Is An Asset Class
NVIDIA wants to mobilize $500 billion for AI factories, and GPU-hours are getting a public price. What that means for finance, and why the hard part is still what last year's chip is worth.
12 min read →The Price of Electricity Belongs in Your AI Forecast
Electricity and grid capacity can flow through data-center costs and compute markets into enterprise AI budgets. A practical forecast workflow for CFOs planning agentic spend.
13 min read →The Question Economy
AI makes polished output a weaker signal of understanding. The Question to Decision Loop shows schools and companies how to make knowledge, judgment, and accountability visible.
12 min read →Kimi K3 Could Reset the Price of Frontier Intelligence
Kimi K3 is still hosted, but its promised open-weight release could turn model portability into enterprise purchasing power and reset what frontier intelligence costs.
15 min read →Stop Asking AI Questions. Give It Work to Finish.
The useful unit of AI is no longer the answer. It is the finished artifact. Why context, memory, and harness design turn agents into reusable digital labor.
14 min read →Every Agent Conversation Comes Back to Security
After Fable 5, enterprise AI is consolidating around auditable open models, sovereign stacks, and security as the adoption gate. Three paths forward for operators building agent workflows that survive policy shifts.
16 min read →One Chat at a Time Is Already Behind
Serial chat caps your compounding rate. The new bottleneck is concurrency: how many agent loops you run before lunch, not how clever your next prompt is.
10 min read →Stop Managing Complex Work in Chat Threads
Agentic coding tools are useful for more than creating product. They can act as context engines and workbenches for complex initiatives that need context, organization, communication, and a clean output package.
9 min read →The Real Risk of Fable 5
Claude Fable 5 shipped as the most capable model ever released to the public, then a government directive took it offline within days. The real risk was never the model. It is who holds the off switch on frontier intelligence, and why open source should be America's strategy, not its loophole.
11 min read →Stop Counting Tasks. Start Closing Loops.
Task automation makes work cheaper, but closed loops make the company smarter by turning every cycle of work into memory, correction, and compounding advantage.
8 min read →Your Job Is Not Going Away. The Boring Parts Are.
AI is most useful when it removes the repeatable tax around work so people can spend more of their careers on judgment, creativity, relationships, and owning outcomes.
8 min read →Scale Your Top 1%
Your best operators should not be bottlenecks. Skills make expert judgment available to every employee, 24/7, before you add agents, tools, and APIs.
13 min read →Enterprise Problems Are Coming for Everyone
AI may expand software TAM by making solo operators and tiny teams capable enough to need enterprise-grade tools, while enterprise products become easier to sell downmarket through agents, better interfaces, and lower adoption friction.
11 min read →Agentic Performance Management Starts at the Data Layer
AI agents only become useful in finance when they can work directly on governed data. Agentic performance management starts at the fabric layer, where data, definitions, permissions, and business context are close enough for agents to help form the first read of performance.
10 min read →Build Your Digital Twin Before Someone Else Does
A digital twin is not a voice clone. It is a way to scale your philosophy and judgment so cheap output does not drown you. Second brain was memory. The next layer is how you choose. Build it on purpose, or platforms will infer a cheaper you.
6 min read →Intelligence Got Cheap. Here's What Just Got Expensive.
When cognitive work gets cheap, scarcity migrates. Production, selection, and standards become the new staircase. Five skills that reprice: taste, framing, customer proximity, standards, and loop ownership.
10 min read →The AI-Native Company
The AI-native company is not defined by AI features. It is defined by redesigning work around abundant intelligence, lower coordination costs, and human judgment reserved for where it matters most.
10 min read →The Wrong AI Metric Is Winning
Token budgets feel like control, but they are a poor proxy for business value. Real AI ROI comes from measuring Verified Work Units and Time-to-Verified Outcome.
5 min read →The Zero-Latency Business
Human latency is the drag on compounding. The next companies win by architecting work so progress rarely waits on a person, and by trading dashboards for sovereign systems.
5 min read →The Subsidy of the Century
AI looks cheap because it is subsidized. A compute war is mispricing intelligence, and the naive builder can exploit the gap before the toll arrives.
3 min read →The Case for the Beginner's Mind
Why staying naive is the ultimate strategic advantage in business and how the greats operated from first principles.
1 min read →