ABSTRACT
Machine learning (ML) enables a computerized system to use data in regard to the human accomplishment of a task – the data input utilized by the human and the data output provided through the human response to this input – to learn to accomplish this task on its own with comparable accuracy and much greater speed. In regard to many human faculties and skills, such as the ability to recognize images and faces or translate text, humans themselves are not aware of the various cognitive steps involved and therefore the possibility of using this awareness to write a program containing these steps or their analogues for the computerized system does not exist. But training on input and output data, as discussed above, does help computerized systems to learn to complete these tasks.
With robots taking over many tasks from humans through ML, there are direct negative implications for the employment of humans but at the same time there are indirect positive impacts because greater output of existing products and production of new products are facilitated by the lowering of costs of production. Consumers benefit from greater product variety as well as price declines resulting from declines in the cost of production.
