Napkin Math for Training Data Value

Napkin math, back-of-the-envelope estimates, and ballpark figures – this interactive page explores order-of-magnitude estimates for important "data value" questions. How will the proceeds and benefits of AI be distributed?

Scenarios

Explore the various scenarios below. Modify input values and choose fill options to see how outcomes change.

Loading inputs...

{{ inputs[inputKey]?.nice_name || formatLabel(inputKey) }} ({{ inputs[inputKey].display_units }})

Calculation Function:

{{ calc.calculate.toString() }}

Raw Value: {{ calc.result.rawValue }}

Unit Details: {{ calc.unitDetails }}

About this page

This website is in an early beta state! Head over to our GitHub page for details and to contribute. There are many open debates about policy and norms around the use of data for AI systems – and often, back-of-the-envelope estimates like these are the starting point.

This interactive tool allows you to adjust assumptions and play with the numbers. The underlying math is simple arithmetic, yet it provides a framework to reason about order-of-magnitude estimates.

To participate in the discussion regarding reasonable default values, check out the GitHub page or leave a comment in our public folder.

Inspector

{{ selectedInputDetails.nice_name || formatLabel(selectedInputKey) }}

({{ selectedInputDetails.display_units }})

Raw Value: {{ selectedInputDetails.value }}
Default Value: {{ selectedInputDetails.default_value }}
Description: {{ selectedInputDetails.value_description }}
Variable Type: {{ selectedInputDetails.variable_type }}
Assumption: {{ selectedInputDetails.key_assumption }}
Source Notes: {{ selectedInputDetails.source_notes }}
Units: {{ selectedInputDetails.units }}
Source URL: {{ selectedInputDetails.source_url }}

Change Log:

{{ entry.time }}: {{ entry.value }}

Click on an input (from the left panel or a scenario card) to inspect its details here.