In this paper, the effects of the rapid advancement of generative artificial intelligence (Gen AI) in higher education (HE) are discussed. A mixed exploratory research approach was employed to understand these impacts, combining analysis of current research trends and students’ perceptions of the effects of Gen AI tools in academia. Through bibliometric analysis and systematic literature review, 64 publications (indexed in the SCOPUS and Web of Science databases) were examined, highlighting Gen AI’s disruptive effect on the pedagogical aspects of HE. The impacts identified by the literature were compared with the perceptions held by computer science students of two different HE institutions (HEIs) on the topic. An exploratory study was developed based on the application of a questionnaire to a group of 112 students. The results suggest that while Gen AI can enhance academic work and learning feedback, it requires appropriate pedagogical support to foster critical, ethical, and digital literacy competencies. Students demonstrate awareness of both the risks and benefits associated with Gen AI in academic settings. The research concludes that failing to recognize and effectively use Gen AI in HE impedes educational progress and the adequate preparation of citizens and workers to think and act in an AI-mediated world.

Traditional estimation methods face challenges in adverse conditions in systems such as Multiple Input Multiple Output (MIMO) with Orthogonal Frequency Division Multiplexing (OFDM). To overcome those challenges, Deep Learning (DL) approaches have been proposed as an interesting alternative, thanks to their ability to capture channel features without much complexity. This paper presents a hybrid approach that combines DL with traditional estimation methods such as Least Squares (LS) and Minimum Mean Square Error (MMSE), which we designate as DL-Enhanced. Our main innovation is a phase-preserving mechanism that maintains critical phase information frequently degraded in purely data-driven approaches. We evaluate the proposed technique considering MIMO-OFDM systems considering 3GPP Clustered Delay Line Model C (CDL-C) channels. Simulation results demonstrate that our method outperforms conventional techniques at high-SNR levels, thanks to neural network-based feature extraction and adaptive processing.