When answers are abundant, the ability to distinguish what matters from what merely sounds plausible becomes more valuable.
Writing & research
After Intelligence
AI is rapidly making sophisticated reasoning, generation, and software execution cheaper and more abundant. The more interesting question is what becomes scarce next: discernment, judgment, agency, trust, authority, and the ability to ask the right questions.
Current series
Privacy · authority · autonomous commerce
Privacy, Power & the Machine Economy
An eight-part series on what happens when autonomous agents negotiate, prove eligibility, exercise delegated authority, and settle value across organizational boundaries. The central question is not only how to make machine commerce private, but how to prevent privacy infrastructure from quietly becoming infrastructure for control.
01
Would You Publish Your Bank Statements in Real Time?Why radical ledger transparency collides with serious enterprise use.
Read essay →
02
The Machine Economy Has an Open-Mic ProblemThe four privacy boundaries of autonomous commerce.
Read essay →
03
When Transactions Become Competitive IntelligenceHow adversaries can learn strategy and policy from observable agent behavior.
Read essay →
04
Your AI Agent Needs a Cryptographic Power of AttorneyBounded, private, consumable delegated authority for autonomous systems.
Read essay →
05
Compliance Without SurveillanceSelective proof, authenticated requests, and keeping compliance out of custody.
Read essay →
06
Private Settlement for the Machine EconomyShared verification without shared visibility.
Read essay →
07
The Freedom Safeguards: Privacy Without PermissionArchitectural constraints intended to keep confidentiality from becoming administrative control.
Read essay →
08
Privacy Has a Dark Side. So Does Compliance.The dual-use problem: impunity on one side, cryptographic permissioning on the other.
Read essay →
Why this exists
The interesting problems begin after the model works.
When capable models become widely available, the bottleneck moves. Generating an answer becomes easier. Deciding which problem deserves attention becomes harder. Producing code becomes cheaper. Deciding what should be built, who should authorize it, how it should behave, and what happens when it fails becomes more important.
After Intelligence is where I explore those second-order questions: the systems, institutions, markets, security boundaries, governance models, and human judgments that become more important as machine capability grows.
Better systems amplify the quality of the question. The framing, assumptions, and objective increasingly determine the outcome.
As AI moves from answering to acting, authority, accountability, security, and institutional design move to the foreground.
Future themes
After Intelligence is intentionally broader than AI commentary. Beyond the current series, here is where I expect to go next.
Agent Security & Containment
What changes when software agents hold credentials, call tools, access infrastructure, and act with increasing independence.
Software After Code Scarcity
How architecture, verification, product design, and engineering practice change as implementation becomes dramatically cheaper.
Institutions After Intelligence
How organizations, markets, professions, education, and governance adapt when high-level cognitive capability becomes widely available.
The premise is not that thinking stops mattering.
It is that as machine intelligence becomes abundant, discernment, judgment, agency, and the ability to ask the right questions become more consequential.
It is that as machine intelligence becomes abundant, discernment, judgment, agency, and the ability to ask the right questions become more consequential.