By PYMNTS | October 27, 2026
For software engineers and managers, the way to determine good artificial intelligence (AI)-powered engineering practices is through long-term outcomes, Terk Terkowitz, vice president and distinguished engineer, Worldwide Amazon Stores, wrote in a memo shared with software engineers across Amazon and posted online Tuesday (Oct. 27).
Terkowitz contrasted that metric with input metrics such as token usage or adoption of AI tools, and with throughput metrics such as code changes and reviews.
“We are all aware of how deeply flawed these metrics are, primarily because they are disconnected from what matters,” Terkowitz wrote in the memo. “We know that using these metrics could, if used as a target instead of a measure, drive negative outcomes, increasing costs (tokenmaxxing is at odds with frugality), and increasing complexity (measuring LOC [lines of code] is at odds with invent and simplify).”
Amazon has seen that teams that have embraced AI have increased productivity and delivery, while those that have not adopted the technology have fallen behind, Terkowitz said.
We’d love to be your preferred source for news.
Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!
However, good leaders understand their team’s work rather than just surface-level metrics, he added.
“If your work does not benefit from current AI tooling, that is a finding, not a failing,” Terkowitz wrote in the memo. “Document why, share it with your community of peers, bring it to your manager and work to find solutions. An engineer who can explain where AI does not yet help their work is demonstrating the judgement we are asking for.”
Leaders responsible for making sure teams or organizations are using AI to the best effect must understand the technologies and must be closer to their products and businesses, Terkowitz said.
“Partnering with you as strong, curious, customer-obsessed owners, we will invent the future of software development together,” Terkowitz wrote.
It was reported in May that Amazon employees had been using the company’s internal AI tool to delegate unnecessary tasks to AI agent to inflate their token consumption on internal leaderboards, because Amazon had set targets for AI use and tracked usage through the leaderboards.
“Managers are looking at it,” one Amazon employee told the Financial Times in the May report. “When they track usage, it creates perverse incentives, and some people are very competitive about it.”
It was also reported in May that some of Silicon Valley’s biggest spenders were rethinking how they used AI after being struck by token shock as spending increased.