Lenny's PodcastAI’s third era: the rise of persistent AI coworkers | Tara Seshan (OpenAI’s product lead)
At a glance
WHAT IT’S REALLY ABOUT
AI’s next leap is persistent coworkers, built through rapid experimentation
- Tara Seshan outlines three eras of AI products—chat, agents, and an emerging era of persistent AI coworkers that collaborate continuously and potentially in multiplayer human+agent workflows.
- She describes OpenAI’s culture as founder-like and low on top-down strategy, emphasizing rapid shipping, tight user feedback loops, and “done over perfect” to match the pace of model improvement.
- She argues AI product strategy cannot be built on static long-term plans; teams must run empirical, 2–3 month experimentation cycles aligned tightly with research roadmaps.
- She explains how product roles are blurring as AI expands individual capability, while human value remains in accountability, taste/creative expression, and relationship-driven collaboration.
- She shares concrete workflows and features—ChatGPT Work mode (Codex power with different UI), Sites for instant shareable tools, and /visualize for fast data storytelling—plus guidance to avoid ‘AI brain rot’ by preserving writing-as-thinking.
IDEAS WORTH REMEMBERING
5 ideasAI is moving toward persistent, teammate-like coworkers—not just chat or one-off agents.
Seshan argues AI products are evolving from conversational chat, to agent-based execution, to “persistent AI coworkers” that stay contextually engaged over time, sync at different cadences, and collaborate with humans (and eventually with other agents) like teammates.
In frontier AI, product strategy is a 2–3 month loop, not an annual roadmap.
Because model capabilities shift so quickly, she believes long-range planning is brittle: building for today’s models is outdated, but building for a year-out guess is also wrong. The practical strategy is tight iteration cycles anchored to likely near-term model improvements and continuous user feedback.
The PM craft concentrates around hypothesis definition and rapid testing—not documentation.
She describes a shift from heavy planning artifacts to fast experiments: define the sharpest hypothesis (“Eigen question”), ship something testable, observe user results, and iterate. The core PM job becomes hypothesis clarity + fast validation, while many traditional PM “trappings” matter less.
Knowledge work is shifting from doing to directing: from rowing to steering.
As agents do more execution (“rowing”), humans increasingly provide direction (“steering”) at higher abstraction levels—from code lines to goals to outcomes. This amplifies the importance of taste, intuition, and opinionated decisions about what future you want to create.
AI rewards people who expand scope, not just automate tasks—ambition becomes a differentiator.
Seshan frames ambition as a competitive advantage when tools make “easy” work trivial. The best users expand their personal capability frontier (prototype, design, model pricing, build tools) rather than only automating rote tasks—and PMs should actively raise the ambition bar for teams.
WORDS WORTH SAVING
5 quotesIf you think about the first era of AI products as chat, the second era of these products working with agents. That third era that might come soon is how do you work with a persistent coworker who is able to get things done with you?
— Tara Seshan
You fail if you build for where the models are now. You fail if you build for where you think the models will be in a year. Both outcomes are equally wrong. Only way to build is two to three months.
— Tara Seshan
Being prolific and empirical is way more important than being academic or theoretical. Rather than writing out some long reasoning doc, instead it's like, "How do I get to something I can try out and test with users as fast as possible?"
— Tara Seshan
Ideally, work feels like a multiplayer game where all of us together are getting stuff done, steering our agents as our agents continue to take care of more and more of those, like, rowing tactical tasks.
— Tara Seshan
I really strongly believe that I do two types of writing at work. One is writing as thinking, and the other is writing as reporting.
— Tara Seshan
High quality AI-generated summary created from speaker-labeled transcript.
