Mozilla launches new startup focused on ‘trusted’ AI

On the eve of its 25th anniversary, Mozilla, the nonprofit behind the Firefox browser, is launching an AI-focused startup.

Call Mozilla.aiThe newly formed company’s mission isn’t just building any AI — its mission is to build “trustworthy” and open-source AI, according to Mark Surman, Mozilla’s executive chairman and chief executive officer. top

“Working with trusted AI for almost 5 years, I always feel both excited and nervous,” he told TechCrunch in an email interview. “A month or two ago, the quick big tech AI announcements made no difference. Really exciting new technology is emerging — new tools that have immediately spurred artists, founders…everyone to do new things. The anxiety comes when you realize almost no one is looking at the railing.”

Surman is referring to the explosion of AI models in recent months that, while impressive in terms of their capabilities, have disturbing real-world effects. Upon release, OpenAI’s text-generated ChatGPT could be promote to write malware, identify exploits in open source, and create phishing sites that look like heavily trafficked sites. Meanwhile, text-to-image AI like Steady Diffusion were co-opted to create pornography, without consensus deepfakes and extremely graphic depictions of violence.

The creators of these models say they are taking steps to limit abuse. But Mozilla felt it wasn’t enough.

“We’ve been working on trusted AI in public interest research for about five years, hoping other industry players with more AI expertise will step up to build the technology,” Surman said. more reliable. “They do not have. So we decided in the middle of last year that we needed to do it ourselves — and find like-minded partners to do it with us. Then we started looking for someone with the right mix of academic and industrial AI experience to lead it.”

Funded by a $30 million initial investment from the Mozilla Foundation, Mozilla’s parent organization, is a wholly-owned subsidiary of the Mozilla Foundation — much like the Mozilla Corporation (organization). responsible for Firefox development) and Mozilla Ventures (Mozilla Foundation Ventures). Its CEO is Moez Draief, who was previously chief scientist at Huawei’s Noah’s Ark AI lab and global chief scientist at consulting firm Capgemini.

Harvard’s Karim Lakhani, Credo’s Navrina Singh and Surman will be initial members of Lakhani is the president and co-founder of the Digital, Data, and Design Institute at Harvard, while Singh is a member of the US Department of Commerce’s National AI Advisory Committee, which advises the president. on a wide range of AI ethical issues.

Surman describes as part research company, part community — a startup dedicated to helping create a trusted, independent, open-source AI stack. Initially,’s priority will be to build a team of about 25 engineers, scientists, and product managers to work on “reliable” recommendation systems and large language models. along the lines of OpenAI. GPT-4. But the company’s larger ambition is to establish a network of consortium companies and research groups — including startups and academic institutions backed by Mozilla Ventures — that share the company’s vision. company.

“We think there is a credible commercial market for AI — and it needs to grow if we are to transform the way the industry builds AI into the apps, products, and services that we all need to know. use every day,” says Surman. “ — loosely collaborating with many companies, researchers and allied governments — [has] opportunity to jointly create an open source AI stack ‘trust first’. If we succeed, the industry mainstream will pull from this stack as part of their usual toolkit, just as they have done with the Linux and Apache stacks for the past two decades.” won’t work alone — not quite. Several nonprofits are on a mission to democratize AI tools, including the recently established Institute of EleutherAI, sponsored by backing companies, including Canva and Hugging Face. There is also the Allen Institute for Artificial Intelligence, founded by the late Microsoft co-founder Paul Allen, and the Alan Turing Institute. Smaller promising efforts include AI startup Cohere’s Connect for AI and Timnit Gebru’s Distributed AI Research, a global decentralized research organization.

In a word, is not a non-profit organization. Although it is bound to certain ethical principles (namely Mozilla Manifesto).

Draief sees this as more of a plus than a disadvantage, arguing that it gives flexibility that nonprofits lack. In his view, there are cautionary tales like OpenAI, which was founded as a non-profit in 2015 but was later forced to switch to “profit limit” structure to fund its ongoing research.

“The big question is, how many newer, smaller, more trusted AI startups will be able to remain independent?” Draief told TechCrunch via email. “It is clear that the big players — especially the cloud platforms from Amazon, Google, and Microsoft — are rushing to consolidate the AI ​​space. This is where all the money is made. And it will be difficult for small companies not to get caught up in this consolidation.”

Pursuing current trends in AI research — and not coincidentally, better funded research fields — will spend the next few months developing tools, such as enabling users interrogate the sources behind the answers the AI ​​chatbot gives. give them. The company will also look to create systems that give users more control over content recommendation AI (i.e. the algorithms that drive YouTube, Twitter, and TikTok feeds), such as Recommended optimization systems for personal or community values ​​— based on Mozilla’s platform. existing research.

Draief doesn’t pretend that transforming the AI ​​stack in a meaningful way will be a quick process. While he pledges that will ship the code “within this year,” he talks about years.

But measurable success will require more than time.

If history is any indication, voluntary frameworks and disposable tools won’t change much, if at all.’s challenge will be to convince the industry that its credible AI vision is right — and accept it.

“Reliable AI features like these don’t seem like they’re worth adding — but we still mostly see them in the lab,” says Draief. “ will work with researchers to turn their work into working code and make it usable in conjunction with more traditional AI tools.”

Source by [author_name]


News7h: Update the world's latest breaking news online of the day, breaking news, politics, society today, international mainstream news .Updated news 24/7: Entertainment, the World everyday world. Hot news, images, video clips that are updated quickly and reliably

Related Articles

Back to top button