Trustworthy data-centric geotechnics
Kok‐Kwang Phoon
Abstract
There are two major technology related changes that are widely viewed as seismic. One is the emergence of artificial intelligence (AI) that is close to passing the Turing test. The second is the introduction of the first AI Act in the European Union (EU). The former portends business opportunities that are too large to ignore in all industries. The latter mandates businesses to manage risks appropriately depending on the impact of AI systems on individuals and society. Machine learning (ML) and AI are at most deployed as proofs of concept in geotechnics. This perspectives article advocates a greater appetite for data-driven innovations in practice to take advantage of major breakthroughs occurring elsewhere. Besides the obvious benefits to practice and to the stakeholders that we serve, embracing data-centric geotechnics is the only way for geotechnical engineers to remain in the driver seat to deploy ML/AI responsibly and safely. This agenda to introduce trustworthiness explicitly in research and practice is called “trustworthy data-centric geotechnics”. Pockets of preliminary research that relate to trustworthiness are highlighted.