Deep learning artificial intelligence and the law of causation: application, challenges and solutions

摘要

Artificial Intelligence technology has rapidly advanced in this era. Many of the Al appliances nowadays have been infused with abilities to self-develop their knowledge and self-enhance their operational precision through machine learning. The emergence of deep learning technology, a sub-set of machine learning, which observes Al mimicking after the way a human brain functions, presents an even more revolutionary step towards perfecting Al's accuracy, precision and efficiency. Similar to a human brain, the cognitive process of the deep learning Al, through its artificial neurons, is not decipherable. This presents a fundamental legal / constitutional problem for most jurisdictions across the world, where any errors or mistakes made by the concerned Al algorithm, are attributable to none other than itself. This research identifies major issues in relation to the application of the law of causation in an Al's case, in effort to propose solutions in this dilemma of justice.