Andrew Ng

If you have come across his name through a lecture clip or a talk on machine learning, the reason is fairly simple: Andrew Ng's work sits at the point where online education at scale meets artificial intelligence put to practical use. Born in London in 1976 and working in the United States, he is a computer scientist and researcher whose reputation rests less on a single breakthrough than on a way of operating — treating scale as something to be designed for from the beginning, rather than something that happens to a project if it succeeds.

The early pattern was translation: taking what was technically possible and making it work in practice. Small improvements accumulated into a recognisable method, tested in laboratories and in markets alike, and shaped by the constraints that mentors, collaborators and critics imposed on his judgement. In the middle stretch of his career he led teams that turned prototypes into platforms, making decisions about architecture, organisational culture and where to put capital. Failures were written up as experiments with findings attached rather than quietly buried, and his public output — memos, interviews, keynotes — largely functioned as teaching about method. Questions of regulation and ethics were handled as part of the design problem, not as an afterthought.

Later on the emphasis shifted towards mentorship and building institutions that outlast individual projects. What survives in the archives of his talks and design notes is the rigour underneath what can look, from the outside, like instinct. Others have taken the approach into new fields, and his name now works as a shorthand for a particular standard of execution in large-scale online learning and applied AI — a legacy carried in organisations and working habits as much as in any product.