Artificial Intelligence, Government Employment and Productivity: Implications for Canadian Federal Government Employment and Costs From AI-augmented Services Implementation
Bibliographic record
Abstract
For any government, being responsible with the taxes of its citizens is of the upmost importance. Therefore, the services it provides must be of the highest quality and efficiency. In recent years, that expectation was unrealized by the average Canadian citizen. Federal services continue to be slow, bothersome, unsatisfactory, and inefficient. However, new technological investment could reverse this course and facilitate a more efficient and satisfactory service. Artificial Intelligence (AI) is a group of cognitive technologies that could provide such changes. As AI is increasing in ubiquity across all sectors of the economy and society at large, there is a substantial debate over how AI will affect them. While there is substantial optimism of the eventual productivity growth and new job creation, there is also pessimism that the technologies will replace far more jobs than they create. Some estimates project that nearly half of the current American jobs will be replaced with AI labour in the next 40 or so years. AI is proving itself to be a general purpose technology that will change the shape of how businesses are run and how the economy is structured. However, this is causing a substantial amount of concern associated with middle class income and employment growth. [vii] In the next 5 to 7-year period, Ai will continue to cause disruptions in employment decisions in all sectors of the economy—that includes government work. As such, the Canadian federal government needs to formulate a plan regarding if and how much it should invest into AI. This study explores the effects of AI on the Canadian government based on comparative studies of AI on US government employment. It finds that in a medium-term timeframe, the federal government could accrue an annual benefit of up to $4.1 billion annually from forgone salary costs from a sample of 23 departments and agencies. These savings are from up to 97 million annual labour-hours that could be replaced through a high level of investment into AI. Based on a comparative analysis with developments in the United States, a number of policy recommendations are advanced. These include the following. • Start investing into AI technologies today. The benefits from machine learning and the development processes are compounding. The earlier the government starts, the sooner they can produce increasing efficiencies. • Provide AI literacy training for the current workforce. This prepares and equips government workers to adapt to complementary tasks and prevents morale disruptions. • Reduce the workforce. Government jobs that become predominantly redundant should be eliminated with prudent investment of the savings in salary costs. • Consult citizens. Citizen input is critical for AI implementation to foster political support for a long process that might not produce visible results quickly. • Avoid AI decision-making. For ethical reasons, humans should be the primary decision makers to prevent data biases.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.005 | 0.023 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.006 | 0.010 |
| Science and technology studies | 0.019 | 0.004 |
| Scholarly communication | 0.013 | 0.004 |
| Open science | 0.005 | 0.004 |
| Research integrity | 0.005 | 0.005 |
| Insufficient payload (model declined to judge) | 0.025 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".