What if AI could serve Peace?
Peace & AI explores how citizens, organisations and communities can understand, shape and use AI as a tool for agency, participation and positive peace.
"Every system we build carries a choice about who it serves. We want AI to carry ours toward peace."
AI: OUR CHOICE
AI can become a driver of conflict, or a tool for positive peace.
The negative effects of AI are already visible everywhere: from the massive energy and water demands of data centres to disinformation, social polarisation, job displacement and the concentration of power in the hands of a few companies. AI does not operate outside society; it amplifies the systems, inequalities and conflicts already within it.
THE SYSTEM WE HAVE
- Energy and water-hungry data centres
- Disinformation and social polarisation
- Job displacement
- Power concentrated in a few companies
→ conflict amplified
THE SYSTEM WE NEED
- Agency for citizens and communities
- Participation in how AI is built and used
- Ownership staying close to home
- Power distributed, not concentrated
→ positive peace
OUR APPROACH
What we do
To ensure AI strengthens human dignity, responsibility and positive peace, rather than domination, manipulation and violence.
1. PREVENT HARM
Building tools to prevent harm and violence.
2. BUILD PEACE CAPABILITY
Building peace capability through relational AI tools for mediation and perspective-taking, increasing reflection and accountability.
3. SHAPE THE CONDITIONS
Shaping the conditions for positive peace through tools for democratic governance and common-good economics.
War & autonomous violence
Dialogue & mediation
Democratic governance
Disinformation & surveillance
Perspective-taking & de-escalation
Relational AI design
Manipulation & power concentration
Responsibility & reflection
Common-good & regenerative economics
Our cases
We exist to prove out the use cases.
At Peace & AI, we exist to prove out use cases that help individuals and communities organize and relate to each other. Two projects set the bounds of that spectrum: from the individual to the community.
ONE PERSON
The Aligned Assistant
In the very near future everyone will have a fully capable AI assistant that will be involved in every aspect of organizing their daily life.
PROBLEM
If these assistants are controlled by commercial interests, true alignment with an individual will not be possible. There will always be a temptation to monetize the person's data, upsell them, advertise to them, surveil them, etc. in an effort to maximize shareholder returns. Individuals whose AIs may be opaquely acting on behalf of commercial interests will not be able to properly advocate for themselves.
SOLUTION
The Aligned Assistant, fully capable, and offered by an organization that is not beholden to any interests other than the individuals it serves. It will be crafted with every piece of technology selected with alignment in mind. In this way an individual can have confidence that their AI assistant is properly representing them and not some outside interest.
A WHOLE COMMUNITY
The Community Action Platform
PROBLEM
Individuals powered with AI chat bots are producing huge amounts of content that never gets turned into action. Everything from legal filings to business plans sit in growing backlogs, unable to be processed by pre-AI organizations and governments.
SOLUTION
An AI-native commons that can receive, curate, and process this content into community-benefitting action. On the individual's side, it will become easier than ever to file everything from claims, to complaints, to requests. On the side of an organization (be it public or private) they can have a way to process, triage, and make action plans based upon the growing influx of AI-generated input.
"We're not trying to replace communities' judgement with a model's; we're trying to give people back the time and tools to use their own."
The team has all bases covered.
Between them, the founders cover the whole distance this work runs: from the design of AI systems and the experience of building and selling companies, through to years of professional conflict mediation in war zones, and inside government, NGOs and private enterprise. Enough technical depth to build the thing, enough practised peace work to know what it should do, and enough institutional range to be credible in a ministry, a boardroom and a community hall.
Technology
Sam Hopkins
Serial technology entrepreneur with deep experience in AI and data. He has built AI products used by brands such as Skyscanner, Oculus, Under Armour, and several more. Founding record 3/3 with one VC-backed company successfully exiting, a non-profit that has run for 9 years, and a bootstrapped AI consultancy which is still growing after 8 years. On top of being a hands-on engineer, he loves designing robust institutions that can pursue their mission above all else.
Peace
Martin Michaelis
Mediator, lawyer and political scientist working on conflict, power and systemic transformation. For more than two decades, Martin has worked across peacebuilding, institutions and organisations, examining how conflict is shaped by the systems around us, and how those systems can be changed.
UNDERSTANDING AI
Building better understanding of AI
We publish and advocate in order to increase understanding of the following.
→ How AI works
What AI can and cannot do.
→ How AI shapes society
How it influences information, decisions, work and human relationships.
→ Who shapes AI
Who owns the technology, controls data and defines its goals.
→ Where risks and opportunities emerge
Bias, manipulation and concentration of power, but also new possibilities for participation, dialogue and collective intelligence.
CONTACT
Talk to the founders.
We are talking to grant-makers and impact investors, to people who have built or governed something like this before, and to towns. If any of that is you, the useful next step is a short call.
Peace & AI: AI as peace infrastructure. A nonprofit in formation.