Amazon has a question for its engineering teams that are slow to adopt AI
Amazon has asked engineering leaders to focus on teams slowest to adopt AI and find out why, as part of an updated role guideline for software development engineers. The internal memo also calls AI usage metrics like token counts deeply flawed, says teams not adopting AI are falling behind, and tells its engineers it is fine if current AI tools do not help their work.
Amazon has called on its engineering teams to assess the pace at which their colleagues are adopting artificial intelligence (AI). In a recent memo, the company urged software engineers to focus on the 10th percentile of AI adoption, identifying teams that are not utilizing AI effectively, as well as those at the 90th percentile, which are heavily utilizing the technology.
The memo, penned by Amazon's VP and Distinguished Engineer Terk Terkowitz, signals that AI will play a central role in the company's software engineering practices.
Amazon has long maintained a hands-off approach when it comes to specific technologies in its software development engineer (SDE) guidelines. However, it recently recognized that AI is changing the field more rapidly and fundamentally than previous innovations like cloud computing and microservices. The company acknowledges that its existing metrics for measuring AI adoption are flawed, as they do not accurately reflect the true impact of AI on engineering teams.
Metrics such as token usage and code changes are disconnected from the core outcomes that matter. The company warns that an overemphasis on these metrics could lead to increased costs and complexity, as they may push engineers towards wasteful practices like "tokenmaxxing."
Despite the flaws in current AI usage metrics, Amazon maintains that teams embracing AI are delivering better results, while those who are slow to adopt are falling behind. The memo reassures engineers that it is acceptable if AI does not enhance their specific work, emphasizing that identifying where AI is not useful is a valuable exercise. Managers are also expected to demonstrate AI application, describing both successful and unsuccessful implementation scenarios.
Furthermore, Amazon is encouraging its engineers to engage more closely with customers, recognizing that those who maintain direct contact with products are reaping the most benefits from AI-driven development. The company is even encouraging engineers to take an AIQ test to assess their AI quotient.
Written by urgent.news from Times of India's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
