Daniel Gebre, an Eritrean graduate of Mohamed bin Zayed University of Artificial Intelligence (MBZUAI), aims to bridge the digital divide for underserved communities through his groundbreaking AI project, iShrink. Growing up in Dekemhare, where internet access was scarce, Gebre’s experiences fueled his passion for technology. His thesis focuses on compressing large language models (LLMs) for offline use, enabling users in low-connectivity areas to access essential resources without waiting for internet access. iShrink has demonstrated significant size reductions in models like LLaMA and Falcon, making them more efficient for mobile devices. Now working as an Applied Scientist at Inception, he aspires to develop AI solutions tailored to industry needs while planning initiatives to enhance digital infrastructure in Eritrea. Gebre emphasizes mastering fundamentals and building professional networks as vital for success, aspiring to foster technological empowerment in his home country, where the gap in digital resources remains substantial.
Source link
