Simpler Models with 18th-Century Mathematics Surpass Deep Learning

Researchers from the University of Jyväskylä developed a simplified AI approach, utilizing 18th-century mathematics, that challenges the dominance of deep learning techniques. The team discovered that simpler and shallow models can achieve comparable or superior results to deep learning, providing a greener, more transparent, and ethical AI alternative, with potential applications in various fields including nanotechnology, digital learning, and medical technology.