Should AI companies build in the open?
One guest hopes more companies take an open source route with AI. Another believes in open source but says it is hard to turn into a profitable business.
One view
Yes, more often: AI does not need to be closed source all the time.
Rizel Scarlett, Block EP 10 · AI, Open Source & Developer Safety | Block’s Rizel ScarlettAnd I also hope that it encourages other companies to take an open source route. I don't think AI needs to be closed source all the time.
The other
It is hard: open source products struggle to become profitable, and enterprises want mature solutions.
Aurimas Griciūnas, SwirlAI EP 38 · From Demo to Defensibility: How to Build an AI Business that Lasts | Aurimas GriciūnasWhen it comes to open source, I'm a strong believer in open source, but it is really also hard to make an open source product a profitable business.
EP 10 was published January 29, 2025; EP 38 was published August 27, 2025. Scarlett encourages companies to open their AI tools while acknowledging financial reasons to keep them closed. Griciūnas believes in open source but emphasizes profitability and enterprise requirements. Their comments address different constraints; neither rejects open source.
Beyond the podcast
Scarlett says she understands keeping AI closed for economic or financial reasons (20:32), and she encourages big and mid-size companies that rely on open source tools to get into open source (23:58). Griciūnas calls himself a strong believer in open source but says "enterprises do not work well with open source" and need a mature solution (7:08). He describes security, stability, on-prem deployment, support and enterprise features as requirements for an infrastructure company (7:45).
Their vantage points differ. Scarlett was speaking about Block releasing its Goose agent as open source, and Griciūnas about what it takes to start an AI infrastructure company.
Open vs. proprietary AI models covers the related choice of which models to build on.
Disclosure: Episode 38 was sponsored by Galileo, where Conor then worked. Episode 10 was recorded while he worked at Galileo; he mentions the company at 27:26.
More debates
- Will AI replace software engineers? One guest says the agent now writes the code. Another says writing code was never the job.
- Is the agentic AI boom a bubble? Spicy take versus the long view on whether agents are hype or here to stay.
- Where do AI agent frameworks add value? Two founders emphasize different parts of the problem: operating the agent stack and improving the context it uses. These positions can coexist.
- Do you need to fine-tune a model? Habib says Writer's models spare customers fine-tuning. Srinivasan says most open-model use cases she sees involve fine-tuning.