Does the Digital Intern use Artificial Intelligence or Machine Learning?
Artificial Intelligence is defined as “the theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.” The concept was developed by Alan Turing in 1950. There are many other definitions for AI however as the field of study is ever changing and expanding.
Machine Learning is a subset of artificial intelligence and is defined as “the study of algorithms and mathematical models that computer systems use to progressively improve their performance on a specific task.” With that definition however, a computer algorithm with predetermined weights, set by a priori knowledge, could be considered to have learned. Thus, a more recent definition has specifically addressed machine learning as a field of artificial intelligence that uses statistical techniques to give computer systems the ability to “learn” from data, without being explicitly programmed on how to interpret that data.
Often adaptive algorithms are synonymous with machine learning. However, not all adaptive algorithms need to have been learned by a machine that is analyzing data. If an algorithm takes into consideration specific parameters and one or more responses to at least one of those parameters is used to vary the output of the algorithm, the algorithm can be considered adaptive. As an example, if an algorithm considers prior responses to blood transfusions of multiple patients and predicts a transfusion amount will last for a presumed period of time for an individual patient, its adaptability may be very limited. However, if the algorithm uses patient specific prior responses and only those patient’s responses are used to predict a response to a transfusion, the algorithm becomes more adaptive. In this manner the algorithm does not give an overall result for all patients, it gives a result that is tailored for the individual patient. This has considerably more power to deliver on the idea of precision medicine.
If an algorithm is designed by a human rather than a machine, it is not truly considered machine learning. It is still artificial intelligence because there is intelligence built in to the system that can be communicated out of the system. Had the mathematical approach with its weights and general shape been chosen and optimized by a machine, then machine learning can be said to have been employed in the process. The Digital Intern is adaptive and yields patient specific results but it does not truly involve machine learning.