Ep 327 – Rebecca Port (Chief People Officer, Okta)

1. Stop thinking about AI replacing jobs. Start deconstructing the work.

The most useful way to think about AI isn’t, “Which jobs disappear?” It’s, “Which tasks inside each job should be done by a human, and which could be handled by a digital coworker?”

That means breaking roles down into their component parts, deciding where AI can add value, and then rebuilding the job around a new set of human + AI handoffs.

Some responsibilities may disappear. Others will become more complex. And entirely new roles will emerge because someone still has to manage the systems, quality, and workflows surrounding those agents.

“What we need to do is break down jobs into tasks and say, ‘Okay, there are parts of your job that could be done by an agent… and we can create a new job that didn’t previously exist.’”

The practical takeaway for People teams: don’t start with headcount reduction. Start with task mapping. Understand the work first, then redesign the role.

2. If an AI agent is doing real work, somebody needs to own it

Rebecca’s view is simple: when an agent is executing meaningful tasks or making decisions, it should be treated more like a coworker than a piece of software. That means someone needs to be accountable for its output, the quality of its work, and the handoffs between the agent and humans.

Not every bot that summarizes Slack messages needs a box on the org chart. But an agent that coaches managers, completes employee-facing work, or takes action across systems probably needs much clearer ownership.

“If you think of an agent as a coworker, then it has to be managed by somebody.”

That creates a new org design question for HR: Who manages the digital workforce? And what does good management look like when one of your direct reports is an AI agent?

3. The future People team may have a digital front door

At Okta, Rebecca is already thinking about a People function where an employee’s first interaction with HR may not be with a human.

Their internal agent, Dex, can answer questions and complete tasks across People, IT, travel, and expenses. Instead of spending as much time answering repetitive employee queries, People Operations can focus more on checking the quality of the agent’s answers and making sure the underlying information stays accurate.

That points toward a People team with a few distinct capabilities: a digital front door for routine interactions, people focused on employee and manager experience, strategic coaches and workforce architects, and technical talent who can actually build and connect the systems.

“Their job has radically shifted. The job still exists, it just looks different in this new world.”

For People leaders, the question isn’t only which HR tasks AI can automate. It’s what higher-value work becomes possible once those tasks move elsewhere.

4. AI proficiency may become the safest career development bet

Traditional career paths are getting harder to predict because nobody knows exactly what every role will look like a few years from now.

Rebecca’s answer isn’t to pretend we can build perfect career ladders for an uncertain future. It’s to make employees as AI proficient as possible today.

That means giving people access to the tools, creating opportunities to experiment, and helping them learn how to work with AI as a colleague. At Okta, leaders were even tasked with building their own agents as a way to learn by doing.

“Build something, name it, learn to work with it as a coworker.”

For People teams, AI development shouldn’t live only inside a training catalog. Give employees real problems to solve with the technology and let proficiency develop through repeated use.

5. Don’t abandon early-career talent just because AI can do junior work

One of the biggest long-term risks Becks sees is companies automating entry-level work without thinking about where future mid-level talent will come from.

She expects entry-level roles to change rather than disappear entirely. The work may become more complex earlier because AI can absorb some of the repetitive tasks that historically occupied junior employees.

That could even bring rotational graduate programs back into focus: fewer entry-level employees, but more deliberate investment in giving them broad experience and accelerating their development.

“If we don’t invest in early career talent, then… the supply and demand issue for mid-level talent is gonna hit us.”

The near-term efficiency play is obvious. The harder question for People leaders is whether today’s workforce design still creates tomorrow’s experienced talent.

‍