You are right that there are no algorithms which capture semantics, and therefore none which can be constructed not to introduce bias. In that sense it is a "d

You are right that there are no algorithms which capture semantics, and therefore none which can be constructed not to introduce bias. In that sense it is a "data solution", but it is not a data problem. > de-biasing the data The issue isnt statistical bias. The issue is semantic: occurrence doesnt capture meaning. Negative terms and racial terms can co-occur more frequently in some text (eg., a biography of MLK) without the remedy being to "rebalance them with positive terms".