Title: AI DATA SYNDROME: COULD SYNTHETIC DATA LEAD TO HALLUCINATIVE AI ANALYTICS OUTCOMES?
Author:
Prof Dr m s s el namaki
Abstract:
Data lies at the heart of our artificial intelligence revolution. Massive volumes of data hold the key to the analytical processes that induce artificial intelligence outcomes. Data is drawn from a wide variety of sources. Yet data are neither homogeneous, generic nor mailable! Data are amorphous whole and carry with it a myriad of problems. This could and does induce “data malfunctions” or data syndrome and the derived issue of data analytics outcome reliability. A hallucinative outcome could emerge. The following article addresses this issue.
The article is qualitative in approach. It identified several triggers of data malignance including, as a prime trigger, synthetic data. Hallucinative outcomes of synthetic data application are projected as an outcome of resort to synthetic data in, among others, LLM software.
The article tries to place this observed data malignance within a synthetic data framework.
The conclusion suggests a set of hypotheses that could be the subject of future research.
Keywords: Data syndrome, Synthetic data, data analytics, Hallucinative analytics outcomes.
PDF Download