How can a team share knowledge with AI without anyone being measured?
Facts as of October 17, 2026
In detail
Anyone who works with AI, with agents and scheduled tasks, transfers knowledge that has accumulated over a whole working life into the machine in a short time. If it sits in a shared store, a team vault, others can access it.
For the organization, that is a big gain: if someone is out, for example due to illness, temporary cover is easier, because much of that person's knowledge is neatly stored.
But the same knowledge could also be analyzed – for example with a prompt asking how to “optimize” the work performance of individual people. That is what measuring means here. That's why the first rule is: whoever shares something must not be measured on the basis of it. If this isn't ruled out through agreements in which the employee representatives are involved, nobody is willing to put anything in.
Working like this requires a high level of trust between team members. Anyone who leads and is part of the team must therefore be explicitly bound not to measure anyone.
In practice
Thomas J. Weiss is building such a team, starting with four people, now about ten. As of October 2026, the rule against measuring is still an intention; the rulebook is being developed with the works council, and a first meeting has taken place. He only wants to expand further once it is in place. The rules will appear here as soon as they are worded, at least in part.
The first rule is, whoever shares something in the team vault may not be measured on the basis of it.
(Original: „Die erste Regel ist, wer etwas im Team-Vault teilt, darf nicht anhand dessen vermessen werden.“)
Sources
- Interview „Team mit Claude“ with Thomas J. Weiss, 16./17.10.2026
Cite: Thomas J. Weiss: “How can a team share knowledge with AI without anyone being measured?” replicapong.de, as of October 17, 2026. https://replicapong.de/en/question/share-without-being-measured/ – CC BY 4.0