AI can generate code faster than your team can review it. Especially with the latest iteration of Claude, the code comes packed with lengthy comments, voluminous release notes, and logic in a bunch of different places, all of which needs to be reviewed before it goes to production. The human in the loop remains essential for quality. But with so much work coming at us, we can quickly become a bottleneck, too. The counterintuitive answer is to limit work in progress and focus

If your AI has read-only access to your email, Teams messages, and calendar, try starting every morning with these questions:

I keep my "to do" list in AI