Manifesto

Intelligence should make work more human, not less.

IA Works exists to help organisations become more intelligent by aligning people, knowledge, technology, decisions and purpose.

Seven governed beliefs

One continuous doctrine.

  1. Collective over computational

    Intelligence is not just computation, and it is not contained in a model. It belongs to communities rather than isolated code, living in the relationships, culture and shared knowledge of people working together, and emerging between people and information, experience and evidence, technology and judgement, institutions and the communities they serve. Artificial intelligence is one participant in this wider system of human and organisational intelligence.

  2. The task is to design better work

    The task before us is not merely to automate existing work. The task is to design better work. True organisational intelligence cannot be engineered through code alone. It requires designing better ways for people and technology to collaborate. Removing steps is easy; deciding which work deserves to exist, who should hold it, and what must remain human is the harder and more valuable question.

  3. We design for human capability

    Technology should enlarge what people can understand, decide and create. We commit to building environments where new AI technologies amplify human capability, deepen collaboration and make the act of creating things together better, removing avoidable burden without removing meaningful agency; making expertise more available without pretending expertise no longer matters; and helping people see complexity, trace consequences, test assumptions and act with greater confidence.

  4. Organisations are communities of knowledge

    An organisation is more than a hierarchy, process map or portfolio of systems. It is a living accumulation of experience, language, relationships, memory, judgement, commitments and shared purpose. Good organisational design strengthens these qualities. Poor design fragments them, hides them, or makes them dependent on a few individuals.

  5. Capability should remain where the work is done

    We reject dependency as a measure of successful consulting. A successful engagement leaves stronger people, clearer methods, reusable tools, better evidence, functioning systems, accountable ownership, and the confidence to continue without us.

  6. Place intelligence close to consequence

    People closest to the work often hold knowledge that central systems cannot capture. Intelligence should sit as close as safely possible to those who understand the context, within shared standards, clear decision rights, visible provenance, appropriate review, and protection against harm. The aim is neither centralised intelligence nor isolated autonomy. It is aligned intelligence.

  7. Keep technology understandable and contestable

    Intelligent systems should not acquire authority merely because they are complex. People should be able to know what a system did, what information it used, what it inferred, what remains uncertain, who is accountable, how it can be corrected, and when it should not be trusted. Provenance, observability and reversibility are design requirements.

Five forms of value

Responsibility extends beyond one measure.

Economic
Is the work sustainable, productive and worth its cost?
Operational
Does it improve quality, reliability and burden?
Human
Does it strengthen agency, capability and the quality of work?
Institutional
Does it preserve knowledge, accountability and resilience?
Community
Does it improve trust, access, fairness and wider outcomes?

Synthesis

Better outcomes require better ways of working.

The way an organisation works shapes what its people and technology can achieve. We design roles, decisions, systems and safeguards together so that quicker or more capable tools lead to better work.

Return to the approach