Responsible Stewardship of AI | 2 of 15 Image by Sharosh Rajeseekher on Unsplash Most technological harm does not begin with someone deciding to cause harm. It begins with a reasonable purpose. A company introduces an AI system to make a process faster. At first, the system serves the purpose for which it was chosen. Then speed becomes the measure of success. What can be counted crowds out what cannot. Employees adapt to the metric. Managers trust the dashboard more than the