Tag: current ai

Make room for yet another partnership on AI. Current AI, a “public interest” initiative focused on fostering and steering development of artificial intelligence in societally beneficial directions, was announced at the French AI Action summit on Monday. It’s kicking off with an initial $400 million in pledges from backers and a plan to pull in $2.5 billion more over the next five years.

Such figures might are small beer when it comes to AI investment, with the French president fresh from trumpeting a private support package worth around $112 billion (which itself pales beside U.S. investments of $500 billion aiming to accelerate the tech). But the partnership is not focused on compute, so its backers believe such relatively modest sums will still be able to produce an impact in key areas where AI could make a critical difference to advancing the public interest in areas like healthcare and supporting climate goals.

The initial details are high level. Under the top-line focus on “the enabling environment for public interest AI,” the initiative has a number of stated aims — including pushing to widen access to “high quality” public and private datasets for AI training; support for open source infrastructure and tooling to boost AI transparency and security; and support for developing systems to measure AI’s social and environmental impact. 

Its founder, Martin Tisné, said the goal is to create a financial vehicle “to provide a North Star for public financing of critical efforts,” such as bringing AI to bear on combating cancers or coming up with treatments for long COVID.

“I think what’s happening is you’ve got a data bottleneck coming in artificial intelligence, because we’re running out of road with data on the web, effectively … and here, what we need is to really unlock innovations in how to make data accessible and available,” he told TechCrunch.

On open source, he said the goal is to support ecosystem building by directing investment with the aim of ensuring that open source tools “are as seamlessly usable as a proprietary tools.”

When it comes to AI accountability, the partnership hopes to “unify the field” — working toward buy-in on standards for auditing AI systems that are accountable on merit of having the “deep involvement by different populations and communities who are focused on the problems that we want [AI to help with].”

“For understandable reasons, there’s a lot of focus around the huge [AI] investments. That’s different,” he also told us. “Our focus here is around the public interest. Our focus is on smaller models. We’re not optimizing for AGI [artificial general intelligence] … We’re looking at smaller models where you need really high-value specific datasets.”

“For example, on Parkinson’s disease — there’s been an amazing standardization of datasets, put forward by the Michael J. Fox Foundation — like we’re looking at really specific stuff to make a difference in people’s lives.”

Europe and the global south chip in

The initiative is being backed by a mix of public and private bodies. Governments in France, Germany, Chile, Kenya, Morocco, and Nigeria are among the nine countries listed as partners at launch. (Note: the U.S. is not a participant, neither are any governments across Asia, so Current AI’s effort is being driven by policymakers in Europe and the Global South. The other listed countries are Finland, Slovenia, and Switzerland.)

That said, also listed in the PR as “core partners” are U.S. tech giants Google and Salesforce.

On the private sector side, Tisné said the partnership is keen to work with industry research labs doing cutting-edge work (e.g., in life sciences), with technology companies that have a distinct positioning versus the mainstream, such as a result of how they and/or their customers are using open source, and with other large companies that are users and buyers of open source products; and with startups that are pushing the envelop on openness.

Other core partners for Current AI that have been named at launch are the French government (which is in the spotlight on AI governance this week, hosting the AI Action Summit in Paris), along with several philanthropic backers, namely The Ford Foundation; John D. and Catherine T. MacArthur Foundation; and AI Collaborative — the latter being an Omidyar Group-backed AI governance policy lobby organization that sits within billionaire Pierre Omidyar and his wife Pam’s network.

Tisné, the founder of Current AI, is also AI Collaborative’s CEO. Discussing the gap Current AI’s backers are seeking to fill, he argued there’s a gap for a public-private funding vehicle that can build momentum behind efforts to drive AI development along public interest lines.

“It’s not a lack of public interest projects in AI, it’s a massive fragmentation in the field and how we can just work on a much bigger scale,” he told TechCrunch during a call, saying the initiative aims to drive “public and private financing at scale of critical public interest AI projects.”

“Some of which already exist,” he continued, saying it’s therefore “a question of really bringing them together, but focusing — focusing on effort so that we can help develop the next AlphaFold“, a reference to Google DeepMind’s pioneering AI system for accurately predicting the structures of proteins inside the human body.

“The private sector is rightly focused on private interests and works at a huge scale — and is doing that with, you know, compute investments in the order of the tens of billions per quarter — so we were trying to figure out what it is that we could do to really make a difference,” Tisné also told us. “AlphaFold was developed on the base of a public data set, including, but not limited to the Protein Data Bank. So a big focus of ours is going to be on data from that perspective.”

Public interest AI ecosystem support

Efforts to widen access to health data could for example focus on supporting development of privacy-preserving technologies to enable more patients to share their data for AI research, he suggested.

“There isn’t another partnership set up which is really designed to bring the whole field together and to bring together public financing at scale,” he claimed.

Boiling the effort down further Tisné said Current AI’s work will span three tracks. Firstly, it will provide financial support to the sector in the form of direct financial contributions. It will also seek to play an incubating role — aiming to, for example, support research work to nurture AI innovations. Thirdly, it will work on “aligning funding so that different funders can work jointly based on shared goals and objectives.” So here its backers hope to bring together diverse public interest AI support efforts and amplify impact.

The partnership will deploy around half its donated funds in the form of grant awards. The other half is being pegged for aligned funding efforts — “around openness, around data and accountability”, which Tisné said will include hitting “really specific goals and objectives” per program (that are as yet to be defined).

“This isn’t a this isn’t a policy or a regulation play. It’s really a building play,” he added.

Current AI’s PR includes a wider list of “supporters” and “champions” for the initiative — citing an open letter of support from a span industry figures including the likes of Arthur Mensch, co-founder and CEO of the French large language model (LLM) maker Mistral; serial entrepreneurs and investors Brent Hoberman and Reid Hoffman; Clement Delangue, CEO of AI firm Hugging Face; and Fidji Simo, an OpenAI board member and also CEO and chair of Instacart, among others.

In the letter they write: “To achieve the best out of AI, society must be in charge.”

“In practice, this means ensuring that high-value datasets are accessible in privacy preserving and safe ways, incentivising the development of smaller, open AI models that cater to people’s needs and are more environmentally friendly, and scaling up open-source AI to improve transparency, safety and accessibility for all,” they go on, adding: “The future of AI should belong to all of us.”

Current AI expects to be unveiling fresh supporters and backers in the coming months, with Tisné saying they’re particularly keen to work with the Gates Foundation given its focus on key verticals like healthcare.

If you’re curious about their choice of name (Current AI), he said they were shooting for something that grounds in the here and now — i.e. current-gen AI, not sci-fi stuff that might be coming down the pipe — as well as wanting to play on ideas of electrical current and even diversity (think the current of a river and all the life than can teem inside it).

“It’s important to have more diversity in the AI field,” he stressed. “We spent a lot of time in AI worrying about a really distant future and what might happen… So this is really an effort to focus on the opportunities and the harms today.”