¿Prefiere leer en español? Leer en español →
Traducción hecha con IA: algunos términos pueden no traducirse bien y puede haber errores.
Predictions
What we expect, and when we’ll check.
Each prediction has a confidence level, what would prove us wrong, and a review date. When the date comes, we mark it right, wrong or not yet.
Organisations will start describing some work by output, not role
Job adverts and contracts that specify outputs and rates, not titles.
Confidence: Medium · What would prove us wrong: Role-based hiring stays the overwhelming norm in UK SMEs. · Review: Oct 2027
Registries of capabilities will appear alongside job boards
Public listings where people, teams, software and AI are searched by what they produce.
Confidence: Medium · What would prove us wrong: Listings stay limited to software agents, or to freelancers by job title. · Review: Oct 2027
Small organisations will start building twins of themselves
Models built from everyday systems (email, files, finance, CRM), not only enterprise ERP.
Confidence: Low to medium · What would prove us wrong: Organisational twins remain enterprise-only products. · Review: Oct 2028
A named person will stay accountable for automated work
Law and contracts keep requiring a human to answer for work done by software or AI.
Confidence: High · What would prove us wrong: UK rules allow work to be owned with no accountable person. · Review: Oct 2028
Some work will come back in-house after being automated
Organisations publicly reverse automation where quality fell.
Confidence: Medium · What would prove us wrong: Automation keeps spreading with no notable reversals. · Review: Oct 2027
Think we’re wrong? Respond to the flagship paper. The argument behind them is in the flagship paper.