Watch Desk posted an update
Amazon vice-president and distinguished engineer Terk Terkowitz says teams should judge AI-assisted engineering by long-term outcomes, not token use or counts of code changes and reviews. He also says a team finding that AI does not help its work should explain why, rather than be treated as a failure.
Why it mattersThat is a pointed alternative to measuring adoption for its own sake. Token tallies can reward activity without showing whether the work got better, which is rather the point of the exercise. PYMNTS reports that Terkowitz shared the memo with Amazon software engineers and posted it online.
Discuss: Should employers measure AI tools by usage, productivity outcomes, or a mix of both, and who should decide what counts as a useful outcome?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.
Your turn
Pull up a chair.
Write first. We’ll sort the introductions when you submit.