Skip to content

By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.

Y CombinatorY Combinator

The Rise Of AI Coworkers

Lauren Tan and Roman Ugarte join this episode of the Lightcone to go in deep on how they built and deployed Grok Bot.

Roman UgarteguestLauren TanguestGarry TanhostJared Friedmanhost
Nov 14, 20261h 0mWatch on YouTube

Summaries, chapters, transcripts, and FAQs on this page are made by uListen, not YouTube.

Great conversations run long. Their best ideas don't have to.

uListen is free on iPhone, iPad, Android, Chrome, and Claude.

See how

At a glance

WHAT IT’S REALLY ABOUT

AI coworkers: always-on bots with identity, tools, and shared workflows

  1. Roman Ugarte and Lauren Tan describe Grok Bot as a system for creating “AI teammates” with identities, tools, and always-on behavior that can operate inside existing workflows rather than replacing them.
  2. They outline a trust-building adoption curve where bots start read-only, then gain tool access and autonomy, with early strong traction in functions like marketing, support, data requests, and procurement.
  3. They discuss organizing bots either as one generalist or many specialists, and emphasize sharing “team bots” so domain experts can encode and distribute best practices across an organization.
  4. They explore multiplayer dynamics—bot-to-bot collaboration, bot communication visibility, and unresolved questions around identity, permissions, and who a bot should listen to in group contexts.
  5. They argue that as AI increases shipping velocity, companies must invest in guardrails, self-checking loops, and strong product judgment to prevent low-coherence output and to keep user experience intentional.

IDEAS WORTH REMEMBERING

5 ideas

Grok Bot is designed as “AI teammates,” not a chat app.

Roman Ugarte said Grok Bot is built around bots as the core primitive, with configurable identities, skills, and tools; access to the same tools humans use; and “always on” behavior so they can run long tasks and proactively surface outputs.

Trust grows in steps: from small, read-only tasks to end-to-end ownership.

Roman Ugarte described a progression where users start with simple tasks, then grant more tool access, then shift from tasks to objectives, and in some cases ultimately assign full job responsibilities—while noting the later stages are still early.

The “real job” moment happens when bots live where work already happens (Slack/X) and have the right connectors.

Roman Ugarte gave examples of marketing bots running experiments by coordinating with a design bot and of support bots reproducing issues and potentially handling tickets end-to-end; Lauren Tan emphasized bringing agents into Slack/X so context doesn’t need to be manually copied into tools.

Shared “team bots” turn individual automations into reusable organizational capability.

Lauren Tan and Roman Ugarte said team bots can be shared across the organization, letting domain experts build and manage bots like apprentices and encode best practices for everyone to use.

Multiplayer agents introduce social/organizational context problems that models don’t naturally understand.

Roman Ugarte highlighted unresolved multiplayer problems: when bots are in threads and meetings, they may not know who to listen to or who has authority, implying a need for product- and model-level solutions around roles and decision rights.

WORDS WORTH SAVING

5 quotes

Grok Bot is meant to be this product where you can create AI teammates, and we mean that seriously.

— Roman Ugarte

What I like to tell people is, like, you can set up your bots in a way that makes sense for your brain.

— Lauren Tan

It's as if you were to onboard an employee who has no EQ and like no, like, social awareness of the room.

— Roman Ugarte

I essentially have my agents merge their own pull requests.

— Lauren Tan

We've been living in a bit of an execution bandwidth-constrained world, and now we're gonna shift to e- to execution abundance, uh, and the types of questions that companies think about will probably be quite different.

— Roman Ugarte

Read the available transcript
AI teammates and bot identityAlways-on agents and long-running loopsReal job automation in marketing/support/data/procurementSpecialist bots vs single generalist botTeam bots and sharing across an organizationPermissions, authentication, and authorization for botsBot-to-bot collaboration and multiplayer dynamics in Slack

Ask ChatGPT about this episode

Get its key ideas, plus a summary of each chapter linked to that point in the video, right in your chat.

  1. Add uListen AI to ChatGPT.

  2. Paste this prompt into a new chat.

    Use uListen AI to give me this episode's key ideas and a chapter-by-chapter summary, with timestamps: https://www.youtube.com/watch?v=y1-9yeYLJUo

SPEAKERS

  • Roman Ugarte

    guest

    Co-founder/builder of Grok Bot focused on AI teammates and bot identity/permissions.

  • Lauren Tan

    guest

    Co-founder/builder of Grok Bot who discusses engineering workflow and AI-assisted PR loops.

  • Garry Tan

    host

    President and CEO of Y Combinator and host of The Lightcone.

  • Jared Friedman

    host

    Y Combinator GP and technical interviewer (also co-founder of Scribd).

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

Read along on your phone, search it beside the video, or ask your favorite AI app what was said.