¿Prefiere leer en español? Leer en español →
Traducción hecha con IA: algunos términos pueden no traducirse bien y puede haber errores.
For researchers
What we don’t know yet.
If you study work, organisations or AI, these are the questions we’d most like help with, including from people who think the premise is wrong. We want to find overlap with existing work, not hide it.
Can a capability be verified?
Who checks a claimed yield, and how often, so “verified” means something?
What would settle it: Records of real capabilities, kept over months.
Where does description stop working?
Which kinds of work can’t be written as input, transformation and output?
What would settle it: Attempts to describe judgement-heavy work, and where they fail.
Split or bond?
When is one large capability better than a chain of small ones?
What would settle it: The same job done both ways, measured on cost, speed and failures.
Who should hold the data?
Does it matter whether the organisation, the capability, both or a steward holds the data work produces?
What would settle it: Pilots under different models; switching costs at exit.
What does a twin need to be trusted?
How complete must an organisation’s model of itself be before decisions rest on it?
What would settle it: Twins compared with what people later confirm or correct.
What measures each stage?
On the way from a twin to a digital corporation, what proves an organisation has moved on?
What would settle it: A measure per stage, tested across organisations.
How we plan to start testing: section 12 of the flagship paper and the lab.
Honest about where we stand: this is our own research, with no university partnership, ethics approval or field study yet. We won’t imply one until it exists. Any research with an organisation would need its own agreement on scope, consent, confidentiality, data protection, ethical review, attribution and publication. Nobody is asked to share confidential information through this page.