Abstract: A method of improving performance of machine learning models by leveraging crowdsourced artificial intelligence includes sending first data to a plurality of data source compute nodes and receiving indications of stakes and estimates based on the first data from the plurality of data source compute nodes. Each data source compute node is ranked based on the received indications of stakes, to generate a plurality of ranked data source compute nodes. An accuracy of each received estimate is calculated by comparing the received estimates to second data. Until a predefined resource is depleted, and in order of rank, if the accuracy of the estimate associated with a ranked data source compute node exceeds a predefined threshold, the predefined resource is decremented and a token augmentation can be assigned to the ranked data source compute node.

Abstract: A method of improving performance of machine learning models by leveraging crowdsourced artificial intelligence includes sending first data to a plurality of data source compute nodes and receiving indications of stakes and estimates based on the first data from the plurality of data source compute nodes. Each data source compute node is ranked based on the received indications of stakes, to generate a plurality of ranked data source compute nodes. An accuracy of each received estimate is calculated by comparing the received estimates to second data. Until a predefined resource is depleted, and in order of rank, if the accuracy of the estimate associated with a ranked data source compute node exceeds a predefined threshold, the predefined resource is decremented and a token augmentation can be assigned to the ranked data source compute node.