Generative AI : Between Limitation and Future Expectations

Abstract

Generative Artificial Intelligence (AI) has considered as a very advanced technology, producing texts like human, creating images, audio, and other contents. Its rapid development has generated both enthusiasm and concerns across many fields including academia, industry, and society. Although it has produced promising results, but many challenges have appeared due to public misconceptions about its capabilities, and usage. Many users mistakenly expect AI systems to make genuine understanding, reasoning, consciousness, and to provide reliable therapy or treatment based only on what the user-has provided from information. In reality, generative AI systems generate outputs from statistical patterns learned from large datasets rather than true comprehension or human-like understanding and decision-making. This misunderstanding has come to unrealistic expectations regarding the reliability and its capabilities for making decisions. This review examines generative AI from the perspectives of its current capabilities, limitations, and future prospects. It discusses major challenges associated with its usage, including AI hallucinations and lack of transparency. The review also explores the huge dependence on AI-generated outputs and its potential impact on human brain, creativity, critical thinking, and making the right decisions, highlighting the gap between public expectations and actual AI capabilities. Furthermore, it proposes a methodology for promoting responsible and aware use of generative AI, emphasizing awareness, critical evaluation, ethical considerations, and Green AI principles.