Here’s another noteworthy partnership in the realm of AI. At the French AI Action summit on Monday, a “public interest” initiative named Current AI was introduced. This initiative aims to encourage and direct the development of artificial intelligence in ways that are advantageous to society. It is beginning with $400 million in initial commitments, with aspirations to accumulate an additional $2.5 billion over the next five years.
While these investment figures might seem limited compared to the larger AI funding landscape—especially considering the recent announcement of a $112 billion private support package by the French president, which is dwarfed by U.S. investments nearing $500 billion to expedite technological advancements—Current AI’s backers maintain that their approach, which does not concentrate on computing power, allows these smaller sums to create meaningful impacts in crucial sectors like healthcare and climate change.
The initial framework of the initiative is broadly defined. Focusing on creating “the enabling environment for public interest AI,” Current AI has several outlined objectives. These include expanding access to “high quality” public and private datasets necessary for AI training, bolstering open-source infrastructure and tools to enhance AI transparency and security, and supporting the development of systems intended to gauge AI’s social and environmental effects.
Founder Martin Tisné expressed that the initiative aims to establish a financial mechanism “to guide public funding for essential projects,” such as utilizing AI in cancer treatment and addressing long COVID challenges.
“We are approaching a data bottleneck in artificial intelligence as we are exhausting available datasets from the web… Hence, it is crucial to unlock innovative methods for making data more accessible,” he articulated to TechCrunch.
Regarding open source, Tisné mentioned that the goal is to foster ecosystem growth by directing investments towards making open-source tools “as user-friendly as proprietary ones.”
On the topic of AI accountability, the partnership intends to “unify the field” — aiming for consensus on standards for auditing AI systems that are accountable through comprehensive engagement from diverse populations and communities focused on pressing issues that AI can help address.
“There’s considerable emphasis on huge [AI] investments, which is a separate concern,” he added. “Our emphasis here is on public interest. We are concentrated on smaller models and don’t prioritize AGI [artificial general intelligence]… instead, we are focusing on smaller models that require highly valuable specific datasets.”
“Take Parkinson’s disease as an example—there has been remarkable standardization of datasets by the Michael J. Fox Foundation—we are concentrating on specific areas that can significantly improve people’s lives.”
Contributions from Europe and the Global South
The initiative boasts backing from a variety of public and private organizations. Among the nine countries participating at launch are France, Germany, Chile, Kenya, Morocco, and Nigeria. (Note: The U.S. is not involved, nor are any Asian governments, indicating that Current AI’s efforts are primarily driven by policymakers in Europe and the Global South. The additional participating nations include Finland, Slovenia, and Switzerland.)
Moreover, the press release cites U.S. tech giants Google and Salesforce as “core partners.”
On the private sector side, Tisné indicated that the partnership is eager to collaborate with industry research labs pioneering advancements (particularly in life sciences), tech companies that differentiate themselves from mainstream counterparts through their use of open-source technology, larger companies that consume and procure open-source products, and startups advancing openness.
Current AI’s launch also identifies several core partners, including the French government (which is currently in the spotlight regarding AI governance during the AI Action Summit in Paris) as well as renowned philanthropic organizations like The Ford Foundation, the John D. and Catherine T. MacArthur Foundation, and AI Collaborative—an Omidyar Group-supported organization focused on AI governance policy.
The CEO of AI Collaborative, Tisné, who is also the founder of Current AI, underscored the necessity of a public-private funding mechanism to generate momentum for AI’s positive societal impact.
“It’s not a shortage of public interest projects in AI; rather, there’s an overwhelming fragmentation within the sector that we can leverage to scale up our efforts,” he discussed with TechCrunch, emphasizing the initiative’s objective to mobilize “public and private financing at a significant level for critical public interest AI projects.”
“Some of these projects are already in existence,” he continued, explaining that it becomes a task of unifying these efforts while concentrating on developing the next AlphaFold—a nod to Google DeepMind’s revolutionary AI system for accurately predicting protein structures.
“The private sector is understandably focused on its own interests, wielding substantial scale through significant computing investments that tally in the tens of billions each quarter. We seek to identify how we can truly make an impactful difference,” Tisné affirmed, noting: “AlphaFold utilized a public dataset, including resources from the Protein Data Bank. Our primary focus will concentrate on data-centric initiatives moving forward.”
Supporting the Public Interest AI Ecosystem
For instance, initiatives aimed at expanding access to health data could support the development of privacy-preserving technologies enabling more patients to share their information for AI-related studies.
“Currently, there is no other partnership designed to cohesively unite the entire sector and attract substantial public financing,” Tisné asserted.
Breaking down the initiative further, Tisné explained that Current AI’s efforts will encompass three key tracks: First, providing financial support through direct contributions to the sector. Secondly, fostering an incubation role to support research initiatives that spur AI innovations. Finally, aligning funding efforts for various funders to collaborate based on shared objectives and goals. The aim here is to consolidate diverse public interest AI initiatives and amplify overall impact.
The partnership intends to allocate approximately half of its donated funds towards grant distributions. The remaining funds will be reserved for aligned funding efforts that focus on “openness, data, and accountability,” featuring precise goals and objectives for each program yet to be defined.
“This isn’t about policy or regulation; it’s fundamentally about building,” he concluded.
Current AI’s announcement has garnered a wide array of “supporters” and “champions,” including industry figures in an open letter, notably Arthur Mensch, co-Founder & CEO of French LLM developer Mistral; serial entrepreneur Brent Hoberman; Reid Hoffman; Clement Delangue, CEO of Hugging Face; and Fidji Simo, an OpenAI board member and the CEO of Instacart, among others.
In their letter, they stated: “To maximize the potential of AI, it is crucial that society takes charge.”
“In practice, this entails ensuring that high-value datasets are accessible in a privacy-conscious and secure manner, promoting the development of smaller, open AI models that cater to human needs while being more environmentally sustainable, and scaling up open-source AI to enhance transparency, safety, and accessibility for all,” they added. “The future of AI should be a shared asset for everyone.”
Current AI anticipates introducing additional supporters and backers in the upcoming months, with Tisné expressing a keen interest in partnering with the Gates Foundation, especially considering its focus on critical sectors like healthcare.
Regarding the choice of the name “Current AI,” Tisné mentioned they aimed for a designation that reflects present-day AI developments—distancing themselves from speculative sci-fi concepts—while also drawing inspiration from the notions of electrical current and diversity, much like the ecosystem teeming in a flowing river.
“It’s essential to advocate for greater diversity within the AI landscape,” he emphasized, adding: “There has been an excessive focus on distant futures and potential scenarios in AI… This initiative seeks to concentrate on today’s opportunities and challenges.”
Compiled by Techarena.au.
Fanpage: TechArena.au
Watch more about AI – Artificial Intelligence
