AI Commons

1. Introduction

AI Commons was born in 2017 from the collective discussion and efforts of a group of individuals and organizations working towards promoting AI for Good and bringing the benefits of AI to everyone and using the technology towards social and economic improvement. The reflections, dialogs, and gatherings resulted in the identification and formulation of an open and collaborative efforts to maximize the social benefits of AI. These efforts have led to the formation of an international nonprofit organization – the AI Commons.
This organization has gathered the best minds in academia, industry, and though leaders in AI to support the creation of a knowledge hub in AI that can be accessible by anyone, that can help inform governance, policy making, and investments around deployment of AI solutions, and be a catalyst for supporting diversity and inclusivity in how AI is deployed for sustainable development goals.

2. Context

AI is currently receiving strong interest from a range of players that benefit from access to the necessary computing power, data, funding, and talent. The benefits that AI can have on many domains is gradually understood and its economic and societal impact a center of attention at all levels of industry and government. Similarly, the use of AI for social good and Sustainable Development Goals (SDGs) has received considerable attention . And while there is a global effort to accelerate capacity building to use AI (education, access to platforms, lowering cost barriers, safety, governance), to truly decentralize and distribute its benefit, AI systems deployed across a range of communities should be designed with people whose perspectives reflect that diversity, and to build AI solutions that address actual needs.

The need for local deployment of AI solutions is key to how AI development is diverse and global. The long term economic and societal value of solving urgent local or global problems may not be always visible to the emergent ecosystem driving progress in AI. It is nonetheless critical to support the creation and distribution of new AI applications by those nearest to the most urgent problems. The decentralization of problem solving is an essential aspect in achieving the promise of AI.

Realizing the promise of an AI for all requires engaging a much broader diversity of actors and to focus on a wider range of solutions and tasks that respond to diverse global needs. It is becoming clear that systemic barriers to entry must be fixed and AI capabilities and benefits be made available to the largest possible community of developers, entrepreneurs, users, and organizations, which will then identify and enable far broader applications in response to actual needs.

3. AI Commons

Vision
To allow anyone, anywhere, to benefit from the possibilities that AI can provide.

Mission and Focus
The organization is focused on delivering on its mission via three directions:
1- Identify and promote community of intent in problem solving with AI
2- Promote shared resources and Frameworks
3- Engage and participate synergies and collaborations for impact

1-Communities of Intent – learning and solving problems together
Connecting problem owners and problem solvers is at the core of creating communities that work together towards sustainable solutions and share in a trusted and safe way their mutual knowledge. Our initial and first effort is to collectively work towards making the knowledge and approaches to problem solving with AI benefit anyone -and especially those who need it the most.
We do this by (1) creating an information hub on beneficial use of AI, and (2) by coordinating efforts to create an open collaboration environment so that problem stakeholders and problem solvers can work jointly towards solving real problems and challenges that matters to them.

2-Promote shared resources and frameworks
A commons become a reality when shared resources are made available and rules of access defined, along with governance to manage those rules. At minimum, having access to tools and expertise as well as computing and data resources in an open environment will allow for a democratized opportunity to allow participation by anyone.
We do this by (1) creating a repository and benchmark of beneficial solutions (ImpactNet), (2) a framework to recommend availability criteria and accessibility of data commons (Data Commons initiative), and (3) partnership to provide cloud and compute resources to enable joint collaboration between problem owners and solvers.

3-Synergies for scale and impact
The AI Commons is a collective effort to help scale AI for Good and for all. Its focus is to collaborate and leverage existing research, standards, frameworks, and other efforts made by various institutions globally, to help its mission. By working together with other organizations on aspects of AI solutions feasibility and deployment (including ethics, privacy, safety, diversity, transparency, protection of IP rights) the knowledge hub can be made available as a centrally accessible reference framework usable for the community on aspects of standardization, governance and policy making.
Finally, to help scale AI solutions, a standardized framework for problem scoping and readiness for scaling will facilitate funding and support by both investment groups and governments locally.

Summary of AI Commons Mission Initiatives

Impact
There are many interested groups which can be contributors to the AI Commons: for instance, NGOs and local government agencies (which may have problems and/or data), or machine learning researchers (which have expertise in the science of AI), startups, non-profits or local governments (which may deploy solutions), and philanthropy (which may be interested in funding specific projects).

Let’s imagine the possibilities if we connected these various groups through a knowledge platform which is open to all? Beyond the simple knowledge of how we can work together to use AI for social good, is the opportunity to allow anyone to have some level of agency and being a part of the problem solving chain. Creating a repository of “how to” that can be used universally is - as demonstrated by many open initiatives, e.g. Wikipedia- will be at the foundation of AI Commons and useful to all actors, from entrepreneurs to academics, and from investors to policy makers and governments.

The initiative will be beneficial to the whole ecosystem working on the advancement of AI, as well as organizations and industry groups dedicated to capacity building and impact investing. It also intends to enable the wider ecosystem to identify and enable new participative models with economic incentives, and to support/increase the flow of funding for sustained development and gradual ownership of problem solving closer to the source in a sustainable way.

Focus on Collaboration
AI Commons is a contribution to help AI be used for the common good and its benefits made available to all. We aim to collaborate with other AI related initiatives to implement their research, guidelines, and best practices. These organizations include MILA, CHAI, IEEE, the Partnership on AI, industry groups, research institutes, and all organizations working towards safe, inclusive, and diversity focused AI developments.

 
4. Governance and Operations

Organization
AI Commons is a nonprofit organization and is creating a multi-stakeholder advisory governance model. It has already strong support from the ecosystem of AI practitioners, entrepreneurs, academia, NGOs, AI industry players and various organizations/ individuals.
The “AI Commons Association” is a nonprofit association currently incorporated in France, with partner offices in Canada, and USA.

Who is behind the AI Commons?
Initiators
Yoshua Bengio, Full Professor, head of the Montreal Institute for Learning Algorithms (MILA)
Stuart Russell, Professor of Electrical Engineering and Computer Sciences, UC-Berkeley
Francesca Rossi, IBM Lead on AI ethics, full professor of computer science at the University of Padova
Amir Banifatemi, GM, XPRIZE Foundation, Chairman Anima, and Managing Partner K5 Ventures,

Founding Members
Konstantinos Karachalios, Managing Director, IEEE Standards Association
Vilas Dhar, McGovern Foundation, US
Zoubin Ghahramani, Professor of Information Engineering, University of Cambridge; Chief Scientist, Uber
Rev. Paolo Benanti, Pontifical Academy for Life, Vatican City
Manuela Veloso, Head of AI Research, JP Morgan, and Professor Robotics Carnegie Mellon University
Illad Nourbakhsh, Professor of Robotics, Carnegie Mellon University
Chaesub Lee, Director, ITU Telecommunication Standardization Bureau
Stefan Germann, CEO, Foundation Botnar, Switzerland
Alexandre Cadain: CEO, Anima, France
Alain Bensoussan, Founder and Partner, Lexing Network, France
Sean McGregor, Syntiant, Machine Learning researcher, US
Trent McConaghy, Founder, Ocean Protocol, Germany
Xavier Herve, Director, District 3, Canada
Myriam Côté, Ph.D., Eng., Director AI for Humanity, Mila

Founding Partners

Wadhwani AI, Radiant Earth, University College London, ACM, Bill and Melinda Gates Foundation, Foundation Botnar, Element AI, University of California Irvine, Carnegie Mellon University, the Partnership to advance responsible technology, Sahara Ventures, and the Swissnex network.

Current Status
The organization has been working through 2018 to establish a charter for its mission. A series of workshops were organized: (1) during the World Government Summit, Dubai, February 2019; (2) in Montreal with Mila, March 2019; and (3) during the AI For Good Global Summit, Geneva, May 2019.
The organization is looking to establish operating offices in Montreal and Paris. It has so far been funded by its founding members time and resources, and is looking for initial external support and grants to establish a secretariat and hub to help with fundraising and establishing operations.
The Montreal hub is important because of the proximity of Mila and the already strong relationship with University of Montreal, Concordia University, Montreal International The Montreal hub is expected to have a 5-7 people team. The Paris hub is important as it brings the European support of ENS, The Future Society, with a strong support from the French - Canadian axis on AI for Humanity. The Paris hub is expected to have a 5 people team. A New York hub is being discussed as IEEE Standards is a strong partner in the organization and can help with areas of ethical design and focus on governance. The New York hub is expected to have a 2-3 people team.
In addition, it is expected to work with a network of :”living and learning labs”. These independent labs are organizations that are already engaged in AI for good efforts. They can be incubators or local labs or working within NGOs. We have already 3 identified living labs: Sahara Ventures in Dar-es-Salam, Wadhwani AI in Mumbai, and District 3 in Montreal. In addition, the Swissnex network is supporting AI commons by offering resources and logistics support to host local teams that are working towards AI for good solutions.

Current Operating Members
Alexandre Cadain: CEO, Anima, France
Myriam Côté, Ph.D., Eng., Director AI for Humanity, Mila
Sydney Swaine-Simon, AI Fellow, District 3, University of Concordia
Gilles Fayad, Deus aiX Machina, Canada
Julien Cornebise, Director of Research and Head of London Office, Element AI, UK
Mathilde Forslund, Foundation Botnar, USA
Maria Axente, PWC, UK
Nicolas Miailhe, The future Society, Harvard Kennedy School, France
Cyrus Hodes, advisor to the Minister of Artificial Intelligence – UAE

Website