Mozilla’s AI-Focused Venture aims to build trustworthy tech
Mozilla, the non-profit that developed Firefox, celebrates its 25th anniversary with the founding of artificial intelligence (AI), a company called Mozilla.ai.
Mark Surman (executive president of Mozilla.ai) stated that the company’s goal is to develop open-source, “trustworthy” AI tech.
Over five years of working on trustworthy AI, he felt “constantly a mix of excitement and anxiety.” TechCrunch states, “The last few months of fast-fire big tech AI announcements have been no different.” There are a lot of exciting techs that are now available. These tools have sparked the imaginations of artists and founders. Anxiety sets in when you realize that almost everyone is not looking at the guardrails.
AI’s Real-World Impacts
Surman was referring to the current avalanche of AI models, which, while impressive in their abilities, have worrying practical implications.
Text-to-image AI systems such as Stable Diffusion can be misused to create pornographic, nonconsensual deep fakes and ultra-graphic depictions of violence. OpenAI’s ChatGPT was initially able to develop malware and find vulnerabilities in open-source code. It could also design phishing websites based on popular ones.
Developers of the models have promised to do all they can to stop exploitation. Mozilla believes that more can be done.
“We have been working on trustworthy AI for the public interest research side for approximately five years. We hoped other industry players would be able to step up and help us build more reliable tech. Surman stated that they haven’t.
Mission Statement for the Firm
Mozilla.ai is a wholly owned division of the Mozilla Foundation. It was seeded with $30 million.
According to the report, Mozilla.ai is a company that aims at helping build an independent, reliable, and open-source AI stack. It combines the characteristics of a research organization with a community.
Mozilla.ai’s initial goal is to gather a team of 25 developers, researchers and product managers to develop trustworthy recommendation systems and large language models similar to OpenAI’s HTML4. Mozilla.ai’s bigger goal is to create a network of like-minded institutions and businesses, such as Mozilla Ventures-backed companies.
Mozilla.ai will be spending the next few months developing tools that allow users to look at the sources that underlie the answers provided by AI chatbots. All this is done to continue following the current AI research trends.
The company uses Mozilla’s prior research to create systems that optimize suggestions for individual and community values. This will give consumers more control over content recommendation AI, like those that power YouTube, Twitter, TikTok, and Twitter.