Amazon has revised its software engineering role guidelines to emphasize outcomes rather than tools used, the company said in a memo urging engineers to learn and apply AI.[1] The memo circulated internally in September and was published Oct. 27, CNBC reported.[2] Amazon called measures such as AI tool adoption and code changes deeply flawed because they are disconnected from what matters, while acknowledging that long-term results are difficult to measure.[1] Amazon shut down an internal leaderboard in May after employees increased AI usage to climb the rankings, CNBC reported; a spokesperson said Amazon has not capped employees’ usage.[2] Amazon says it believes that it is best to measure, and be measured by, long-term outcomes, and that input metrics like token usage or adoption of AI tools are deeply flawed, primarily because they are disconnected from what matters.[1]
Amazon calls AI tool adoption metrics for engineers deeply flawed
In a memo urging engineers to learn and apply AI, Amazon says it is best to be measured by long-term outcomes, while acknowledging they are difficult to measure.
1 primary · 1 independent · 0 commentary · 1 vendorSources
- An update from Amazon engineering leadership on the role of the developer in the age of AI Amazon News PrimaryVendor
- Amazon's anti-tokenmaxxing memo says engineers should be pickier about using AI CNBC Independent