Possible

Useful robots won’t be programmed one task at a time; they’ll need to adapt to unfamiliar objects, environments, and robot bodies. Physical Intelligence cofounder Chelsea Finn (named last month to the TIME100 AI list for 2026) joins Reid Hoffman and Aria Finger to explain how general-purpose robot models learn to act in the messy physical world. She explores why lower-quality training data can strengthen a model, what months of failed laundry-folding attempts revealed, and how the same system can work across different robot platforms. A robot mistaking an oven for a drawer becomes a lesson in the strange line between useful generalization and obvious error. They also examine why robot demos can mislead, how hardware reliability can stall progress, and what it will take to move AI off the screen and into the world.

Congratulations to Chelsea, recently named one of TIME’s 100 most influential people in AI:

https://time.com/collection/time100-ai/2026/chelsea-finn/

AI strategist Matthew Tiemann is building toward a world where AI agents work around the clock—and humans step in for the decisions that matter. Across his work at energy services company Foundry-Logic and his personal projects, he uses Claude Code, Codex, and Fable to automate knowledge work and accelerate software development. Those projects range from an AI-assisted iOS game to a personal dashboard informed by his goals, to-do list, and Oura Ring sleep data. Drawing on experience across data analytics, supply chain, robotics, logistics, and energy, Matthew joins Reid Hoffman and Parth Patil to explore proactive “heartbeat” agents and shared context. They also discuss adversarial review, 12-hour human check-ins, and AI’s limits in subjective judgment—revealing why human vision and alignment remain essential.

Jonathan Brazeau joins Reid Hoffman and Parth Patil to explain how AI animation and filmmaking are making the one-person studio possible—letting small teams create films that once required 100 people.

From Dungeons & Dragons world-building and an RPG art book illustrated with Stable Diffusion, to a fantasy-western short film made with Midjourney, Luma AI, and Suno, Jonathan breaks down his workflow: character design, storyboarding, mocap, rotoscoping, music, and human curation.

The conversation explores the “cogs and sparks” framework, what a lean AI-native studio can realistically accomplish, how taste separates art from AI slop, and why technology still can’t provide the essential creative spark.

Read Questmaxxing on Substack: https://reidhoffman.substack.com/p/questmaxxing

Cadi Zhang joins Reid Hoffman and Parth Patil to explore how generative AI is changing game development, world models, and robotics. Drawing on her work across Unity games, VR teleoperation, AI accounting, and robotics product operations, Cadi explains what virtual worlds can teach embodied intelligence—and why physical robots still lack the tactile, real-world data that can’t be scraped from the internet.

She shares how she uses GPT, Blender, PixelLab, and Aseprite to prototype games faster, including a duck-themed imposter game and PokéTax, her Pokémon-inspired tax-filing game. AI can rapidly generate code, gameplay mechanics, and 3D assets, she says, but it still can’t judge whether a game feels fun, maintain a consistent art style, or model the precise force needed to fold a sheet of paper.

Reid, Parth, and Cadi discuss simulation-to-reality gaps, the limits of current world models, robot safety, humanoid versus task-specific form factors, and why human taste remains the decisive creative skill.

Sougwen Chung has spent more than a decade drawing alongside robots—not to automate their art, but to discover what changes when a machine becomes a creative partner. They join Reid to trace the evolution of D.O.U.G., a series of robotic collaborators that learns from Sougwen’s gestures, including a neural network trained on two decades of their drawings. Sougwen explains why unexpected machine marks became errors to “poeticize,” what becoming machine-readable taught them about remaining human, and why they call their practice “operational art” rather than “AI art.” They also explore a new system guided by brainwaves and an ancient calligraphic pause—and ask whether the intelligent systems we build need more than speed and capability. Maybe they also need silence, restraint, and the wisdom to wait.

Credits:

A big thanks to Natalie Stone, Tessa Nijdam, Arina Ermakova , the Scilicet team, Fellowship, Reece Straw, Sam Osborn, Nick Capezzera, Maximilian Andereya, Peter Butterworth, Haruka Wang and the team at Mosaic Ventures for hosting the conversation.

Additional Credits:

MIMICRY (Drawing Operations Unit: Generation_1), 2015. Footage: Sam Osborn and Nick Capezzera.

MEMORY (Drawing Operations Unit: Generation_2), 2017. Footage: NTT InterCommunication Center Tokyo.

COLLECTIVITY (Drawing Operations Unit: Generation_3), 2018. Footage: New INC & Nokia Bell Labs.

SPECTRALITY (Drawing Operations Unit: Generation_4), 2020. Footage: Maximilian Andereya.

ASSEMBLY (Drawing Operations Unit: Generation_5), 2022. Footage: Peter Butterworth.

GENESIS Process Film, 2023. Courtesy of the artist and Studio SCILICET.