Slow Down Faster With AI
Introduction
Often, speed and quality are at odds with each other. It’s hard to make something good, and make it fast; it’s hard to make something fast, and make it good. And if you do manage to attain both of those things, chances are it won’t come cheap. AI is unlikely to resolve that conundrum, but if you’re prepared to sacrifice some amount of speed in favor of quality, AI can amplify the quality improvements you stand to gain. AI can improve the quality of outcomes outright, like it has done for early skin cancer detection; more interestingly, it can also make learning opportunities more apparent and bolster team collaboration — when it’s taken slow.
Outright Increase
The most direct means of using AI to improve quality is using the AI to get a second opinion, like an editor or verification system. Dermatologists using an AI to detect cancerous skin lesions found that using the machine increased the likelihood of early detection - quite possibly saving lives (1). They used the machine as a “decision support” tool, sharing the cognitive work with an AI collaborator. This elongated the process by making the two decision phases (prediction, judgement) more distinct from one another (2), but it increased the likelihood of early detection. When lives are on the line, quality is of utmost importance, and the time it takes to get this second opinion is a worthwhile tradeoff.
