Author
Antoine Buteau

Antoine Buteau

3 min read Free

Building AI Products Series #9: AI Product Metrics, Economics, and Support Burden

AI product metrics cannot stop at adoption. A feature can be used often and still be bad. Users may try it because it is new, because it is forced into the workflow, or because they are hoping it will improve. Usage does not prove value, trust, quality, or economics. AI
4 min read Free

Building AI Products Series #8: Data Flywheels Without Magical Thinking

"The product will get better with more data" is one of the most abused sentences in AI strategy. Sometimes it is true. Often it is wishful thinking with a diagram. A data flywheel only works when the product creates useful data, has the right to use it, can
3 min read Free

Building AI Products Series #7: Model Choice Is Product Strategy

Model choice is often treated as an engineering detail. It is not. Choosing between a frontier API, a smaller model, an open model, a fine-tune, retrieval, rules, agents, or human operations changes the product's cost, latency, privacy posture, reliability, roadmap, support model, and competitive durability. That is
3 min read Free

Building AI Products Series #6: Evals Are Product Requirements

Evals are often treated as technical hygiene. That is too small. For AI products, evals are product requirements. They define what good means, what failure means, what can ship, what must be reviewed, and what needs to improve before users are exposed to a change. If a team cannot evaluate
3 min read Free

Building AI Products Series #5: UX for Confidence, Correction, and Trust

Trust in AI products is not created by saying "trust us." It is created by repeated interactions where the product is useful, honest, controllable, and recoverable. Users learn what the system is good at. They learn where it is weak. They learn whether it respects their judgment. They
3 min read Free

Building AI Products Series #4: Designing Around Uncertainty and Failure

Traditional software can fail, but it usually fails in familiar ways: a bug, an outage, a missing permission, a validation error. AI products fail differently. They can be fluent and wrong. They can be partially right. They can be right yesterday and worse tomorrow. They can answer beyond their evidence.
4 min read Free

Building AI Products Series #3: Choosing Problems Worth Solving With AI

Not every product problem should become an AI feature. Some problems need a better workflow. Some need clearer information architecture. Some need rules. Some need a human approval step. Some need fewer features, not a model. AI is useful when model behavior creates leverage that a deterministic product cannot easily
4 min read Free

Building AI Products Series #2: Model Capability Is Not Product Value

A model can summarize, classify, draft, reason, search, translate, plan, and generate code. That does not mean your product has value. Model capability is raw potential. Product value appears only when that capability maps to a job users care about, inside a workflow they will actually use, with quality they
You've successfully subscribed to Antoine Buteau
Great! Next, complete checkout to get full access to all premium content.
Welcome back! You've successfully signed in.
Unable to sign you in. Please try again.
Success! Your account is fully activated, you now have access to all content.
Error! Stripe checkout failed.
Success! Your billing info is updated.
Error! Billing info update failed.
Press ESC to close.