Try Pro free

MTS · 2 min · Nov 3, 2026

RuneScape gold inspired Tutor’s human-tutor approach

A story about Venezuelans earning money in RuneScape shifted Josh Gruenstein’s thinking about robot learning.

The episode’s best minutes

Full episode
  1. 3:33Robot deployment requires solving the whole stackPro pick 4 min
  2. 10:44New models make robotics simulation more scalable4 min
  3. 0:39RuneScape gold inspired Tutor’s human-tutor approach2 min
  4. One more moment in the full episode
Up next · In Good Company with Nicolai TangenAI agents could flatten business and speed up workPro pick 6:14
“what if we could connect the 8 billion human beings on earth with robots in the field and teach them?”
JGJosh Gruensteinat 1:52 · MTS
1:52

Takeaways

  1. The idea grew out of a story about Venezuelan RuneScape players.
  2. People already know how to perform many physical tasks.
  3. Tutor links field robots with human tutors and gathers training data.

About this segment

A story about Venezuelans earning money in RuneScape shifted Josh Gruenstein’s thinking about robot learning. Tutor’s idea was to connect human know-how directly to robots in the field.

JG

Josh Gruenstein

Josh Gruenstein is cofounder and CEO of Tutor Intelligence.

  • Theo Jaffee, Host · Co-founder & host, MTS

    One of the first members of a16z's new media team; earlier hosted the Emergent Ventures-backed Theo Jaffee Podcast.

  • Sophia Dew, Host · Host, MTS

From “Robots Will Proliferate Daily Life in 5 Years | Josh Gruenstein”, Nov 3, 2026. Watch the full episode

Transcript

Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.

  1. 0:43

    five and a half years. Tell me a little bit more about that origin story. Where did you guys start and where are you guys today? Yeah, I guess important context, have you ever played RuneScape? I know RuneScape. You know RuneScape. Okay. So back in 2019, there were this series of blackouts in Venezuela. And at that time, I was a grad student at MIT working on robot reinforcement learning. And I read this article talking about how in RuneScape, there was a huge population of people in Venezuela who were killing green dragons and exchanging green dragons for RuneScape gold and exchanging that RuneScape gold for United States currency. And you could avoid the rapidly depreciating ball bar that way. And this was like a kind of shocking insight because I had been spending the last five years or really 10 years of like my personal

  2. 1:38

    passion of working on robot learning to be like, how do we get robots to know all the things that humans know? And meanwhile, there's this whole internet of humans who already understand how to pick up a Red Bull and any other task. So we had this idea of maybe instead of trying to figure out like the algorithm through reinforcement learning from scratch of how do robots learn to interact with the physical world, what if we could connect the 8 billion human beings on earth with robots in the field and teach them?

  3. 2:03

    Wow. So to understand exactly what's happening, it's like from the video that I saw, your launch video, you have human tutors essentially mapping out to robots how they actually should function. And then the robots can learn from the humans, then update their own knowledge and just become a lot more effective? Correct. The way our business works is we're kind of three businesses in a trench coat. We are manufacturing robots, so humanoids and robots that look kind of less like humanoids.

  4. 2:29

    We then deploy those robots out into real-world businesses in the United States, businesses that help ship goods, manufacture goods, all sorts of things across the physical economy. And then we connect those robots in the field with our tutors. And that both makes the robots actually work and be able to do these tasks that robots can't do today. And it also generates this beautiful stream of data.

  5. 2:49

    And we can use that data to train models of the frontier with real-world physical data that other labs don't have. Tell me why robots and automation inside of factories and manufacturing is critical and important right now.