DavorCukeric
Back home

The journal

Writing

Longer pieces on the questions I keep circling — sovereign AI, adopting it with care, and what it can mean for a person. Written to be useful, not to sell anything.

Eloryn in the federal government: governing AI that acts, not just advises

Ottawa wrote sensible rules for automated decisions years ago. Agentic AI — systems that take actions, not just recommend them — is about to test whether those rules hold in practice.

June 20267 minElorynFederal governmentAI governance

Not just Ottawa: AI governance at the provincial level

Health, education, social services, licensing — most government decisions that touch a Canadian's daily life are provincial. The privacy and human-rights rules there are real, specific, and already in force.

June 20266 minElorynProvincial governmentPrivacy

AI in healthcare needs more than a model — it needs a record

A model that suggests a diagnosis is impressive. A system you can trust near a patient is a different thing entirely — and the difference is almost all governance.

June 20266 minElorynHealthcarePatient safety

What model risk means for AI in finance

Banks and insurers have governed models they don't fully understand for decades. AI doesn't need a new philosophy of oversight so much as an old one, applied honestly and at speed.

June 20267 minElorynFinancial servicesModel risk

Governance is the on-ramp, not the brake

Outside the heavily regulated industries, no one is forcing the issue — which is exactly why getting AI governance right is a competitive advantage rather than a compliance chore.

June 20266 minElorynPrivate sectorAdoption

Who governs the agents?

We spent two years teaching AI to talk. The harder, more interesting question is what happens when we let it act.

June 20266 minAI governanceAgentsHuman-in-the-loop

Building steady in an uncertain world

The world feels unusually unsettled right now — economically, politically. I don’t think the answer is to freeze. I think it’s to build things that hold.

June 20265 minPerspectiveUncertaintyBuilding

What sovereign AI actually means — and why a country like Canada should care

Sovereignty isn't about owning every layer of the stack. It's about deciding, on your own terms, which parts have to stay yours.

June 20266 minSovereign AICanadaPolicy

Adopting AI well: a quieter set of principles

Most of the value in adopting AI isn't in the model. It's in the boring, human decisions about where to keep a person in the loop.

June 20265 minResponsible AIAdoptionHuman-in-the-loop

AI as a patient tool for personal growth

The most interesting thing AI can be for a person isn't a shortcut. It's a tireless, judgement-free tutor that meets you exactly where you are.

June 20264 minPersonal growthLearningHuman + AI