open-source-ai speech-models healthcare-ai ai-ethics

Mozilla: The state of open source AI

stateofopensource.ai

July 17, 2026

11 min read

🔥🔥🔥🔥🔥

67/100

Summary

A Māori broadcaster in New Zealand trains speech models for te reo, ensuring data remains with its community. PwC has fine-tuned an open model for financial language, serving hundreds of clients on its own hardware without a per-token fee. Researchers in Lausanne developed an open medical model in collaboration with the Red Cross, focusing on humanitarian applications.

Key Takeaways

Open models are leading in adoption, with 79% of developers using them compared to 71% for closed models, though only 51% of open-model teams reach production versus 63% for closed models.
A Māori broadcaster in New Zealand is training speech models for te reo under a license that retains data ownership with the community.
Researchers in Lausanne have developed an open medical model in collaboration with the Red Cross, which is set for clinical trials in Switzerland and Tanzania.
The majority of production tokens now route through open weights, indicating a significant shift towards open-source AI in practical applications.

Read original article

View story →

Coverage & discussions

Hacker News · stateofopensource.ai

Mozilla: The state of open source AI ↗

Community discussions

Hacker News discussion ↗

Historical count: 5 · 5 saved in sample · Historical sample

Analysis ready

Community Sentiment

Mixed

Positives

The rapid growth of open models handling tokens is a game changer, almost 5x increase in just four months — that’s a clear sign of shifting tides in AI development.
Commenters are bullish on open models potentially dethroning giants like Anthropic and OpenAI, especially as hyperscalers can run these without licensing fees.
The community's enthusiasm for open models demonstrates a desire for democratization, as many believe they will keep the system honest against monopolistic practices.

Concerns

Skeptics argue that the costs of training open models remain astronomical, making them impractical for widespread use compared to frontier models owned by big players.
Concerns about the lack of a competitive moat for companies like OpenAI and Anthropic are rampant, with many feeling their high investments may not yield sustainable advantages.
Some commenters criticize the article's tone and writing style, suggesting it undermines the credibility of its pro-open model stance, which could alienate potential supporters.