Skip to content

PAN Lab levers

Lever

Pause AI on alarms

The organization sets alarms that halt the AI system's part in its decisions. When an alarm fires, the people making decisions stop receiving the system's output. The halt holds until a person reviews what happened and clears it. While the halt holds, failures cannot keep compounding.

What it is

The organization decides on its circuit breaker in advance, while nobody is under pressure. The organization names the signals that mean something has gone wrong, such as a rise in overturned decisions or a pattern of complaints. It commits to stopping the system when those signals appear. Deciding in advance means nobody has to argue for the halt while the pressure is on.

What it pushes on in the Lab

In the Lab, this lever closes the pathway by which people adopt the automated system's failed output, for as long as the halt holds. It also puts a ceiling on the failure regime, the network's tendency toward failures that compound rather than correct themselves.

removedFailures adopted by people or agents
cappedFailure regime

In the modes that offer aiming, you can aim this lever at particular parts and pathways of a network. Otherwise it applies to the whole network.

Its pattern in the Practice Library

The Practice Library describes the pattern behind this lever:Deployment circuit-breaker

The pressures it answers

These pressures list this lever among the levers that answer them:

A lever answers a pressure when it pushes the other way on something the pressure pushes on.

Where you can pull it

Networks in the Lab that offer this lever:75

Every network that offers it

The evidence behind its effects

The Lab cites no claim from the evidence registry for this lever's effects. The Lab's authors set the direction and size of each effect.

Pull this lever in the PAN Lab and watch which way it pushes the network.

Open the PAN Lab