Overview · Policy proposal

The Amplified Economy

What could people achieve if AI extended the reach of their expertise?

AI can help us finish familiar tasks. It can also change how we organize the work: what we ask people to decide, what AI carries out, and what becomes possible as a result. [§5.1]

The Amplified Economy proposes a way to discover which approaches produce substantial, useful gains, and learn how those gains can work beyond a single person or workplace. [§6]

Explore the idea through real-world experience, illustrative scenarios and the policy itself. You can question any part of it.

Draft notice

A proposed framework, open for discussion. The examples explain the idea; they do not establish that the policy will succeed.

Discover what works. Find what limits it. Share what holds up.

Policy proposal

A promising AI workflow in one workplace raises several questions. Is the work actually better or faster after checking and correction? Does the gain survive beyond one person's task? Can another team reproduce it?

The proposal is to run bounded workplace experiments, compare different ways of dividing work between people and AI, and measure the results. Where gains are promising, investigate the constraints that prevent them from becoming useful organizational outcomes. Test what transfers, and share credible lessons, including failures. [§6] [§5.2]

Public support would follow the evidence. An unsuccessful approach could be stopped or redesigned. A promising result would earn further testing before broader claims were made. [§12]

DISCOVER promising ways of dividing work between people and AI

MEASURE their full, quality-adjusted effects

IDENTIFY the constraints that limit wider benefit

TEST whether the result transfers

SHARE what holds up, including failures

Why begin in Newfoundland and Labrador?

Policy proposal

The proposed first step is a workplace learning pilot in Newfoundland and Labrador. The province offers varied businesses, public services and communities within a scale where institutions could work together on shared experiments. [§10.1]

The pilot would test how to discover, measure and reproduce useful configurations. It would be a first test of a wider Canadian approach. It could not, by itself, establish an economy-wide benefit.

What are we looking for?

Policy definition

means a substantial, disproportionate increase in useful results relative to the resources required. A multiple increase in productivity can qualify even when people are doing familiar work. Previously impractical work or something newly viable can qualify too. [§3] [§8.2]

The comparison must account for and the other work required to achieve the result. Producing more material does not establish the gain by itself. [§8.3]

Choose a starting example

Based on a real-world practitioner account

A web developer changes where their effort goes

An experienced developer reports completing their own contribution much faster by putting more expertise into research, specification and direction of AI work. The wider project still has other constraints.

Illustrative scenario with assumed outcomes

A cabinetmaker builds a tool around the business

A cabinetmaker turns practical knowledge of jobs, dependencies and recurring mistakes into a custom management tool, freeing time across the company.

More examples

Repairing equipment: A repair technician investigates an unfamiliar fault
Practical experience directs a wider search, but a convincing diagnosis still needs a physical test.
Illustrative scenario — outcome not yet established
Preparing funding applications: A nonprofit increases its capacity to prepare sound applications
A checked foundation lets a small team prepare more suitable applications without multiplying its workload.
Illustrative scenario — assumed threefold preparation capacity
Coordinating municipal repairs: A municipal team finds where faster work stops helping
Requests reach the field queue sooner, but the crew cannot complete more jobs until another constraint changes.
Illustrative scenario — local gain, service outcome unresolved

Connected ideas

Augmentation, automation and amplification

Augmentation and automation describe how work is done. Amplification describes a substantial gain in what that arrangement achieves, taking quality and resources into account.

Domain expertise

Domain expertise is the practical knowledge that helps someone recognize what matters, set useful constraints and judge whether a result will work.

Bottleneck migration

When one stage speeds up, another stage may become the main limit on the whole process. The gain and the remaining constraint both matter.

Quality-adjusted productivity

Compare useful results with the resources needed to produce them, including checking and correction. Faster output counts only to the extent that its value survives those demands.

Reinvesting capacity

Time released by AI can be used for additional work, stronger testing, better quality or learning. Whether that creates value must be checked.

About this prototype

Status. Draft interactive brief · Examples and explanations are being reviewed.

Sources. Policy documents contains the complete text of three discussion drafts, frozen on 29 September 2026: the policy framework and its theory and measurement companions.

Conversation. Questions you type, with the current page context and relevant sections of the policy documents, are sent to an OpenAI model (GPT-6 Astra) through Lovable AI. The gateway lists this model as zero data retention on the endpoint used. This prototype keeps the conversation only in this browser tab's session and does not store it on a server.

Answers use the full policy framework, its theory and measurement companions, and labelled teaching examples. A source explains what the document says; it does not make every claim proven.