Mission and values

AR Labs exists to create products and solutions that use machine intelligence to support what we call the higher values of humanity: connection and collaboration, education and progression, a better quality of life for everyone, and the sustainability of life on Earth for this generation and the ones after it. Alongside them we hold a set of working commitments about how AI should be built: human-centred, transparent, ethically guarded, and accessible to individuals.

This page states each value and says what it means in our practice.

The mission

We began with a voice-controlled system as our first foundation: conversational AI that understands the person it works with, communicates naturally, anticipates needs, takes the initiative on useful tasks and thinks alongside them to support their productivity and creativity. Our mission sets that work inside the larger purpose of the four values below, and our long-term aims follow the same lines: education, personal productivity, reducing the damaging effects of technology on society, and ethical deployment.

flowchart LR
  M["Machine intelligence"] --> C["Connection and collaboration"]
  M --> E["Education and progression"]
  M --> Q["Quality of life for all"]
  M --> S["Sustainability of life on Earth"]

Connection and collaboration

We think human and AI collaboration achieves what neither achieves alone, and that AI should augment people rather than replace them. In practice this value shows up in two places. The systems we design keep the person in charge of decisions; our AI agents (language models run in a loop with tools, so they can act, not only reply) follow an explicit rule, act or ask, about which steps they may take themselves and which belong to a person. And our design for an agentic organisation treats collaboration between agents as something to be specified: durable threads of work, explicit handoffs, and a record of who decided what.

Education and progression

Education appears in our mission, in our long-term aims, and in our view of where AI should be used carefully. We see AI as a way to make learning more personal and more accessible. We also hold that learning is one of the things that should not be sped up past the point where a person understands what they are doing.

In practice we teach what we build. We run a second brain workshop for coaches, freelancers and small-business owners, publish step-by-step guides, and document our engineering methods openly on this site. We apply the same value to ourselves: our own learning is built into our work, as the Loop, the cycle by which AR Labs works, describes. The essay series on the acceleration paradox argues the public side of the same point: if AI replaces junior developers, the route by which people become senior ones disappears.

Quality of life for all

Our hope for work, held as an outlook rather than a forecast, is that technology takes on menial tasks and is pleasant to use; the future of work with AI agents sets out that outlook. "For all" is meant literally: we want this capability held by individuals and small teams, not concentrated in large organisations, and we think intelligence should be accessible to everyone.

Quality of life also means attention. We treat information overload as a real cost to people, and design against it: cognitive chunking presents a long AI answer one piece at a time, and the VOX interface puts structure in a picture rather than in another paragraph. The small-business guides on this site aim at the same thing from the other side: removing clerical work while leaving judgement with the people who run the business.

Sustainability of life on Earth

The mission ends on the widest value: maximising the sustainability of life on the planet for current and future generations. Our long-term plans connect this to intelligent automation that makes better use of resources.

We apply a smaller form of the same value to everything we build, taken from our founding framework, cognitive ecology: an action is judged by whether its long-term consequences are acceptable for the person and for the wider system, not only by whether it reaches its immediate goal.

Working commitments

Alongside the four values, we hold commitments about how we build:

• Human-centred design with ethical guardrails. Agents work inside stated limits on what they may decide, and a person approves consequential steps; in our coding pipeline, a person, not an agent, merges agent-written code into the main branch.
• Transparent AI. Systems should show how they think and work. Our agents record their decisions and evidence, and our interfaces aim to make an agent's progress visible rather than hidden.
• Ethical deployment. We name ethical AI that benefits humanity as a long-term aim, and the essay series on our blog rejects uses of AI it considers dishonest, such as copying another company's product to take its clients, or advertising the replacement of staff as an achievement.
• Accessibility. Capability for individuals and small teams, proven first at the scale of one person.

What we believe about AI sets out the reasoning behind these commitments.