Rolling out AI agents? Move fast and furious — but with extreme caution

Getting started and capturing the essence and value of AI can be exhausting. It also poses a vexing question: is the way forward to move fast and furious, jumping on every model that comes online, or to proceed with a plan and extreme caution? Testimonials from two enterprise IT leaders at a Section conference suggest a mix of both approaches is necessary.

First, keep humans central. "Stop being the human in the loop. The human is the loop," said Scott Likens of PwC. Lasherelle Morgan of NBCUniversal urged teams to start with the end user and work backward — "start with the pain point" and focus on easily repeatable processes and clean data.

PwC also runs AI experiments in one-day or five-day cycles to learn fast; Likens warned that a narrow cost focus is "the wrong way to look at it" and pointed to resistance in the "frozen middle" of organizations. Be ready to blow up a bad process.

ai agents, enterprise ai, ai experiments, human loop, pwc, nbcuniversal, end user, pain point, clean data, repeatable processes