Research Is Not Engineering at a Slower Speed (voiceinthemachine.com)

A recent article highlights the crucial distinctions between research, research and development (R&D), and product engineering in the AI/ML landscape, advocating for a clearer understanding of these roles within organizations. Drawing on experiences from various tech environments, the author emphasizes that conflating these categories leads to mismanagement, underfunding of essential work, and the potential loss of talented individuals. Each type of work has its unique goals and metrics for success: product engineering focuses on delivering measurable outputs in short timeframes, R&D aims to push existing systems further with defined trajectories, while research operates on long timelines with high epistemic risks and uncertain outcomes. The piece introduces a framework to categorize innovation efforts into four quadrants based on their probability of success and impact: Pearls, Oysters, Bread and Butter, and White Elephants. Understanding these categories can help organizations allocate resources effectively and set appropriate expectations for innovation endeavors. The author urges companies to recognize and institutionalize research as a strategic asset rather than an isolated activity. By aligning definitions of success and fostering collaboration between researchers and engineers, organizations can better position themselves to innovate meaningfully and avoid the pitfalls of mismatched expectations.