01 WORK
How AI redistributes work
We study which tasks are automated, which remain human, and which new roles emerge around deploying, evaluating, and governing AI.
- Task shifts
- Occupational transformation
- New roles
Labour market × AI research
AI is advancing faster
than people and
organisations can adapt.
OfferWind conducts research on labour and education as AI spreads. We study changes in occupations, skills, and how people and AI agents work together.
Current project · Project 01
OfferWind Labs
OfferWind Labs is an agentic environment for collaborative project work. It connects people and their personal AI agents through layered context, controlled access permissions, and an action provenance log. The system records how tasks are distributed, context is transferred, and outcomes are produced.
RESEARCH QUESTION
How does teamwork change when personal AI agents become participants in shared communication?
HYPOTHESIS AND MEASUREMENT
Bringing personal AI agents into shared communication will reduce coordination time and repeated explanations without lowering the quality of the outcome.
- 01 / BASELINE
We define the task and baseline measures. The team completes the project while using personal agents in isolation.
- 02 / CHANGE IN CONDITIONS
People gain access to their colleagues’ agents, while agents gain access to permitted project context.
- 03 / CORE METRICS
Task completion and handoff time, repeated explanations, duplicated work, number of iterations, and outcome quality.
- 04 / SUPPORT CRITERION
The hypothesis is supported only if work becomes faster without a decline in quality or new overhead.
Research areas
How work is changing in the age of AI
We do not study an abstract “future of work”. We study observable change: which tasks disappear and emerge, which capabilities gain importance, and how companies redesign hiring, learning, and teamwork.
02 ADAPTATION
Which capabilities matter more
We study how demand for technical and human capabilities changes, what people need to learn, and which formats help them adapt faster.
- Demand shifts
- Education
- Generalist capabilities
03 ORGANISATIONS
How hiring and teams are changing
We study the shift from CVs to observable work and how coordination changes when AI agents become part of the team.
- Proof of work
- Capability assessment
- Human–AI teams
Methodology
How we conduct research
We start with a specific question, define the necessary data and how it will be tested. We compare plausible explanations and report facts, interpretation, and what the evidence does not yet allow us to claim.
- 01Define the research question and scope
- 02Choose the data and method of testing
- 03Compare hypotheses and examine contradictions
- 04Report findings and limitations
AI-native company
Building an autonomous company governed by people.
Our goal is an organisation in which agents perform most operational work, collaborate with one another, and retain working context. People set direction, govern the system, and make critical decisions.
OPERATING PRINCIPLE
People govern. Agents execute and coordinate.
The company is built around persistent agent infrastructure: shared goals, working contexts, areas of responsibility, tools, and rules for escalating decisions to people.
Human governance
People set goals and priorities, define system boundaries, and retain control over critical decisions.
Digital employees
Agents own areas of responsibility, run continuous workflows, and deliver completed work or decisions to people.
Agent coordination
Agents exchange context, create subtasks, and hand work to one another without requiring people to repeat information at every stage.
Owned and external models
We train specialised models and use Claude, Codex, and other systems as parts of one operating infrastructure.
About OfferWind
Why OfferWind exists.
Help people adapt to rapid change and develop skills that remain valuable.
We study changes in work and translate our findings into clear materials, experiments, and tools.
AI is already part of work. It should be used deliberately, with its outputs kept under human control.
We do not frame technology through fear. We want to understand how people can work with AI while preserving autonomy, critical judgement, and responsibility.
OfferWind culture
What we value in people and in working together.
- 01Curiosity
We value genuine curiosity — the spark that drives people to ask questions, explore new ground, and move beyond the boundaries of their discipline.
- 02It is okay not to know
There is no need to perform certainty. What matters is acknowledging uncertainty, wanting to understand, and returning with a clearer view.
- 03Teamwork
AI can amplify an individual, but strong outcomes still grow from trust, dialogue, and people working together.
- 04Initiative
When we see a problem or an opportunity, we do not wait for instructions. We propose the next step and take ownership of it.
- 05Responsibility
Freedom to act means being accountable for decisions, commitments, and outcomes — whether the work was performed by a person or an agent.