The "Centaur Model": AI+humans

[tevgeniou]

“what types of tasks are better for humans, computers, centaurs”.

Example (where i come from): investment related decisions (e.g. Stock picking vs portfolio construction). there must be a way to do a careful study on this (even though narrow one at first)…

[tevgeniou]

https://techcrunch.com/2016/11/01/how-combined-human-and-computer-intelligence-will-redefine-jobs/

https://medium.com/conversations-with-tyler/garry-kasparov-tyler-cowen-chess-iq-ai-putin-3bf28baf4dba

https://www.huffingtonpost.com/mike-cassidy/centaur-chess-shows-power_b_6383606.html

[tevgeniou]

Book: https://www.amazon.fr/Deep-Thinking-Machine-Intelligence-Creativity/dp/1473653509/ref=sr_1_fkmr0_1?ie=UTF8&qid=1510656281&sr=8-1-fkmr0&keywords=Deep+Thinking%3A+Where+Machine+Intellig

[tevgeniou]

Visual arts and music: Collaboration will replace the linear nature of artistic creation that we think of today. Two different algorithmic versions of a music program could give a human enough content to combine the two and generate an entirely new genre. Or, like Google’s Deep Dream, humans can input seeds of information for machines to generate artistic products.

Film and television: There are enough test cases for us to truly understand what a well-framed scene looks like. Teaching a machine how to essentially direct means filmmakers can set up scenes in VR and focus more on storytelling and creative connections than the minute details of production.

Architecture and product design: Function over form has dominated each of these fields. However, leaving a machine to design based on function over form might have us living in buildings that are just white boxes. However, IoT sensors can teach machines how we interact with our environments to learn exactly what people need in terms of function, leaving humans to balance function with form — spending more time on the art and less on the details.