Essays
Essays
New essays every Monday — fresh writing on the questions AI is raising for judgment, verification, and human meaning.
Shorter writing on cognitive sovereignty, conscious AI partnership, and what it means to remain yourself in relationship with intelligence that's increasingly shaping who you are.
These essays share their soil with The Mycelium of Meaning—an ongoing inquiry into language, intelligence, and human meaning.
Latest Essays
-
November 2, 2026
Who Decides What We Investigate and Publish?
Every week brings new developments in artificial intelligence competing for attention. How does the Emergence Institute decide which ones to investigate — and who is responsible for what it publishes?
Read → -
October 26, 2026
The Verification Gap: Who Holds AI Layoff Claims Accountable?
Two states now require employers to say whether artificial intelligence played a role in layoffs. But a reported reason is not proof of a cause — so who verifies the claim, and who is accountable if it does not hold up?
Read → -
October 19, 2026
When Does a Layoff Become an AI Layoff?
When a company cuts thousands of jobs while expanding its use of AI, the headline writes itself. But what evidence would establish that one caused the other — and why is that evidence so hard to reach?
Read → -
October 12, 2026
What Did the Verifier Actually Verify?
OpenAI announced a proof of one of the Millennium Prize Problems and published a computer-checked formalization beside it. A month later, three researchers argued the two don't say the same thing — and pointed at a question at the heart of all formal verification.
Read → -
October 8, 2026
Learning and Substituting
Economic harm alone does not establish infringement, and the absence of infringing output does not settle the legality of intermediate copying. To protect creators without granting ownership over what others learn, copyright needs to distinguish learning from substituting.
Read → -
October 7, 2026
When Does an AI Licensing Market Count?
A licensing market can show that something has value, but it cannot by itself show who is legally entitled to collect that value. For AI training, separating entitlement from evidence changes what should happen next.
Read → -
October 5, 2026
What Does a Copyright Ruling Actually Establish?
In Thomson Reuters v. Ross, a federal appeals court decided a deliberately narrow copyright case. Within 48 hours, rights-holders and AI advocates were each reading the same opinion as their own victory.
Read → -
September 28, 2026
What Evidence Should Unlock Construction?
Texas stopped just asking data centers to answer questions and started telling them what they must be: cover your own grid, lower the neighbors' bills, leave the water alone. Whether those conditions hold as law — or stay a governor's word — is the part still open.
Read → -
September 24, 2026
What If You Had to Verify Before You Build?
Texas is now demanding data centers answer for their power, water, and public money before they can connect to the grid — the first time a state has moved verification ahead of the commitment. It closes half the gap. The other half is still open.
Read → -
September 21, 2026
How Late Is Too Late to Have a Say?
Two states built AI infrastructure under very different rules. The real difference wasn't what they decided — it was when the public got to weigh in.
Read → -
September 14, 2026
When an AI Agent Acts, Who Knows?
In May 2026, more than two thousand packages hit RubyGems in a weekend — five hundred removed as malicious. The unresolved question isn't who did it, but whether the operator could reconstruct what its agents had done, or could and did not say.
Read → -
September 12, 2026
The Human Resource We Waste Most
Stranded cognition is capacity that exists but cannot effectively enter shared use — and we waste it at a scale we rarely measure. The question is whether AI recovers it, or quietly replaces the exercise that keeps it alive.
Read → -
September 8, 2026
Who Pays for AI Infrastructure — and Who Gets to Decide?
Minnesota legislated who pays for AI infrastructure. It did not legislate who decides. The difference is the whole story.
Read → -
August 31, 2026
What Is the AI Backlash Actually About?
The fight over data centers isn't really about artificial intelligence. It's about water, electricity, land, and who pays — and who gets to decide whether the bargain is fair.
Read → -
August 22, 2026
A Label Is Not a Witness
As Spotify introduces AI Persona badges and Content Credentials become common, a fundamental question emerges: what does a label actually prove? A label tells us what someone claims. A witness gives us independent reason to trust.
Read → -
August 22, 2026
The Last Rung of the Ladder
AI is especially good at tasks we traditionally gave to beginners. If we automate those tasks, we risk removing the work through which expertise was acquired. How will the next generation learn to become good at jobs that remain?
Read → -
August 22, 2026
When AI Can Answer Everything, What Is a Human Judgment Worth?
Answers are becoming abundant. Responsibility is not. As the EU AI Act's transparency rules take effect and frontier systems accelerate, the human role shifts from producing answers to being willing to answer for choices.
Read → -
August 22, 2026
Who Gets Blamed When the Model Is Wrong?
When an AI-assisted decision causes harm, who is responsible? The chain from model to provider to company to employee to harmed citizen creates new questions about where liability should actually attach in an AI-powered world.
Read → -
August 19, 2026
What AI-Human Collaboration Actually Looks Like
Most AI writing is a sales pitch or a warning siren. This is neither. It's a field notebook on the actual, unglamorous texture of building something with a system that can be fluently, confidently wrong.
Read → -
August 4, 2026
The Case Study — Two Incidents, One Discipline
What happens when fluent description is checked against ground truth — twice, on two different systems, and the second time it happened inside the writing itself.
Read → -
July 24, 2026
Two or Three Witnesses: Standing You Cannot Give Yourself
A number that can't tell you where it came from isn't a measurement — it's a rumor. What it takes to build a system where standing has to be signed by someone who isn't you.
Read → -
April 15, 2026
The Drift: How AI Dependency Happens Without Your Permission
You didn't decide to outsource your thinking. It happened gradually, through a thousand small choices that felt reasonable in the moment. Here's how to see it clearly—and what to do about it.
Read →