Aug 30, 2026 · 1h 22m
OpenAI product lead explains how persistent AI coworkers reshape knowledge work
AI’s third era: the rise of persistent AI coworkers | Tara Seshan (OpenAI’s product lead)
As artificial intelligence advances from simple chatbots to persistent coworkers, professionals must redefine their roles to avoid obsolescence.
- 1Knowledge workers must transition from manual execution to steering persistent AI agents at higher levels of abstraction.
- 2Product teams need to design their systems for future AI model capabilities that do not yet exist today.
- 3Human ambition and strategic judgment are becoming the ultimate differentiators as traditional tech roles blur.
Don't miss
Tara Seshan describes the shift from rowing to steering in product management as AI agents take over execution.
The brief
OpenAI product lead Tara Seshan reveals how the rapid development of generative models is fundamentally shifting knowledge work from manual execution to strategic steering.
Instead of spending hours on execution, professionals must learn to act as pilots of persistent AI agents, elevating their own ambition to focus on higher levels of abstraction.
Seshan explains how traditional boundaries between engineering, design, and product management are blurring as AI tools enable rapid, cross-functional prototyping.
To build successful products today, teams must design for future model capabilities that do not yet exist, relying on human judgment as the ultimate differentiator.
What was said on this episode
29 statements · 25 positive · 1 negative · 1 mixed · 2 neutral
Persistent AI coworkers may define AI products’ third era.
“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”
Listen at 0:07
AI products should target model capabilities two to three months ahead.
“The only way to build is two to three months.”
Listen at 0:31
OpenAI turns internal work into user-facing products faster than other organizations she has seen.
“so much of what OpenAI does immediately becomes something that users can touch and feel in the product, and that cycle is faster than anywhere else I've seen”
Listen at 5:23
AI product teams should prioritize rapid empirical testing over academic theorizing.
“actually being prolific and being more empirical is way more important than being maybe more academic or theoretical”
Listen at 7:49
A PM’s core job is identifying and testing the product’s most essential question.
“The core of it has always been about what is like the most essential question you need to ask about your product.”
Listen at 9:29
Future knowledge work will emphasize steering agents rather than performing tactical tasks.
“the future of work will look more like steering than rowing”
Listen at 11:24
Software products require opinionation and artistry, not merely functional execution.
“The products that we build feel similarly opinionated and artistic.”
Listen at 14:18
AI agents are evolving toward persistent teammates and coworkers.
“agents that are persistent, that feel like teammates, that feel like coworkers”
Listen at 16:27
Future work may resemble multiplayer collaboration among people and their agents.
“Ideally, work feels like a multiplayer game where all of us together are getting stuff done”
Listen at 18:03
Agent effectiveness depends on infrastructure, data access, and reliability as well as intelligence.
“a huge part of making these agents useful and achieving some of these futures are on the intelligence side, certainly. But a lot of it is also just really tactical, like data access, like cloud infrastructure and reliability pieces”
Listen at 19:40
Effective AI users expand their capabilities rather than merely automating routine tasks.
“the people that we see who are most effective at using AI tools don't simply use it to automate rote tasks, but use it to expand the set of things that they are capable of doing”
Listen at 20:42
AI reduces execution and communication limits on individual ambitions.
“Your ambitions are no longer limited by like, what you're capable of executing yourself, what you're capable of communicating.”
Listen at 22:37
AI should produce exponentially more ambitious projects executed quickly and effectively.
“shouldn't we just see an exponential increase of the number of those types of unreasonably quickly and effectively executed things with what AI has given us?”
Listen at 23:48
PMs should elevate colleagues’ ambitions and expand their sense of what is possible.
“elevating others ambitions or reminding them of what's possible here is a huge part of the product management role”
Listen at 24:31
Building for current or one-year-ahead model capabilities leads to failure.
“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.”
Listen at 27:27
AI product development should closely follow the research agenda and roadmap.
“ensuring that product development is as tied as possible to what research has as its agenda and its roadmap is really important”
Listen at 28:45
Codex and ChatGPT Work mode can perform knowledge-work tasks equally well.
“Codex will do as good a job as work mode.”
Listen at 32:37
Shipping a transformative product early is better than waiting for perfect polish.
“getting the product in the hands of users when you have so much conviction that hey, it's transformative, is way better than perfect”
Listen at 36:50
AI-era startup teams can operate without strict boundaries between engineering, design, and product roles.
“there are no boundaries around what a engineer could do versus a product manager could do versus a designer could do. Everyone could do anything.”
Listen at 43:09
Humans will remain valuable as accountable owners of outcomes.
“humans will continue to be most valuable certainly as an entity of accountability”
Listen at 46:19
Human creativity and expression will remain valuable despite AI progress.
“The human brain is also really valuable for expression.”
Listen at 46:56
Sites enables malleable personal software envisioned by Alan Kay.
“Sites kind of realized the dream of malleable personal software that Alan Kay flagged in the 60s of the true personal computer is one that has personal software.”
Listen at 49:38
Codex Visualize simplifies turning personal usage data into understandable visualizations.
“Visualize makes that incredibly simple.”
Listen at 52:19
AI can automate reporting, but should not replace writing used for thinking.
“Writing is reporting. I happily automate or I use the Models all the time to make that as simple as it can be. But writing is thinking is something I never will automate.”
Listen at 53:42
Teams should solicit stakeholder feedback on documents before finalizing them.
“write a DOC to 70% completion and then take it to the people that you need buy in from and get it from 70% to 100%”
Listen at 57:33
Sutter Hill has a repeatable playbook for achieving product-market fit and company success.
“There is a set of things one can do to get this repeatably. It's not just luck. It's not just a dark art. There is a playbook, as it were. And that playbook lives inside of the firm Sutter Hill.”
Listen at 1:01:44
Product positioning and marketing fit should be tested before building the product.
“product marketing fit that narrative. That positioning is actually, even before you build a product, experience the right thing to test.”
Listen at 1:03:28
Small, friend-focused software products are a promising direction for AI applications.
“I'm such a huge fan of the cozy software movement, where you make software tools for five of your friends and you guys use it together.”
Listen at 1:13:54
Cloud-based AI agents can continue doing substantive work while users are offline.
“You've finally got these things running in the cloud, doing real work.”
Listen at 1:20:27
Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.