$DIS · IP holders like Disney should train proprietary models on open-source foundations to control IP and enable fan creativity

episode briefing

extracted intelligence

All-In Podcast

Open Source Wins, AGI Is Here, and Scorsese’s AI Toolkit with CEOs of Cerebras & Black Forest Labs

now viewing▶ 59:00

IP holders like Disney should train proprietary models on open-source foundations to control IP and enable fan creativity

2026-07-10 · 5 company · 8 thematic

sentiment

3 bull0 bear2 neu

speakers

andrew feldman

Leads Cerebras, a public company (CBRS) building wafer-scale AI inference chips; previously co-founded SeaMicro (acquired by AMD); describes $25B backlog and global data center buildout at unprecedented scale.

robin rombach

Co-invented latent diffusion (basis for Stable Diffusion, Flux); leads Black Forest Labs building open-source multimodal models for images, video, audio, and robot action prediction; company has 100+ employees across Germany and San Francisco.

quote

“We do work with certain IP holders to develop models together with them... some of them based on our open source models some of them based on like our more powerful proprietary models but I think that is like a very like attractive value”

—robin rombach

now playing · $DIS

$CBRS···bullish· high· posandrew feldman

Cerebras sees $25B backlog as AI data center buildout exceeds 50 years of global power growth

Demand for inference compute is insatiable with booked orders from OpenAI, Anthropic, Google, Microsoft, AWS; Cerebras' wafer-scale architecture delivers 15x faster inference for…

$BLACK-FOREST-LABSbullish· medium· posrobin rombach

Black Forest Labs raises significant capital, hits 100 employees building multimodal world models for robotics

The company behind Flux and latent diffusion is converging image, video, audio, and action prediction into unified multimodal models that can serve as 'brains' for robots, with ea…

$OPENAIbullish· highandrew feldman

OpenAI's reasoning models (o1, o3) validate inference compute scaling thesis

OpenAI's release of reasoning models that 'think' for extended periods confirms the shift to inference-time compute scaling; these models consume massive internal tokens making ul…

$NVDA···neutral· mediumandrew feldman

Hyperscalers building custom silicon to reduce Nvidia dependency, not to compete on performance

Major cloud providers (Amazon, Google, Microsoft) are developing proprietary chips primarily to control their supply chain destiny and avoid single-vendor dependency learned from…

$DIS···neutral· mediumrobin rombach

IP holders like Disney should train proprietary models on open-source foundations to control IP and enable fan creativity

Major IP owners can license open-source base models (like Flux) and fine-tune on their proprietary libraries to maintain control while enabling new revenue streams from fan-genera…