At Slack and Palantir, I ran standups religiously.

Not because I loved them. The coordination problem was just real. You had engineers across workstreams, and without a daily sync, people drifted. Someone would be blocked for half a day before anyone noticed. Two people would quietly start solving the same problem. And then you'd find out in the standup that three days of work needed to be thrown away. The standup existed for one reason: to make sure every single person was on the right thing, at the right time - every day.

That logic goes back to the factory floor (literally). Scrum borrowed from Toyota's production system, from all the kanban boards to the just-in-time allocation stuff. Assembly-line thinking, adapted for software. Even after "Scrum" became a dirty word, the rituals stuck. The underlying problem never went away: humans need synchronization points to stay productive.

Until now.

In an AI-native development world, you're orchestrating agents, not syncing humans in a room. Tickets become prompts, allocation is continuous instead of sprint-bounded, agents pull their own work and move on to the next thing, and they do it 24/7 without you having to check in on Monday morning to find out what happened last week! Could it be that we don't actually need the standup anymore?

Same job (make sure the right work gets done in the right order), but the cadence is faster and more continuous. You're running an orchestration loop instead of a sprint planning meeting. The standup doesn't disappear... it just stops being a meeting.

I've managed product engineering teams at places like Slack and Palantir. The hardest part was always throughput: making sure smart people weren't idle or pulling in the wrong direction. That problem doesn't go away in an AI-native world. But how you solve it looks completely different now.

At ThriveAI, this is what we're building around: what product management looks like when your team includes agents. The playbook is changing faster than most teams realize.

— Brian, ThriveAI