COOKIES: By using this website you agree that we can place Google Analytics Cookies on your device for performance monitoring. |
University of Cambridge > Talks.cam > Darwin College Science Seminars > Creative Intelligence in Generative Models and Why Consciousness Matters
Creative Intelligence in Generative Models and Why Consciousness MattersAdd to your list(s) Download to your calendar using vCal
If you have a question about this talk, please contact Dr Sandra Petrus-Reurer. Grab some lunch from the Darwin servery and enjoy an interesting science talk and discussion over lunch (talk starts 13.10, so make sure you're seated by then). Looking forward to seeing you there. With the development of new artificial neural networks like generative models, it is increasingly important to talk about whether these systems are truly creative. But in order to have a conversation about the creative capabilities of these systems, it is important to first identify, quantify and define their creative intelligence. It is also crucial to talk about the tasks generative systems like the DALL E and Stable Diffusion cannot perform as opposed to the tasks that we, as humans, can. Some tasks that are akin to humans in terms of creative intelligence are those wherein an artist draws from her subjective experiences of her inner life to generate a work of art. But when it comes to generative models, they do not have an inner life of experiences. The art these systems generate is a mere coalescing of information. Now the question persists, does having an inner life of subjective experiences really matter when it comes to the creation of artwork? In this talk, I attempt to explore the question of whether quantifying creative intelligence in generative models could lead to understanding how phenomenal contents of conscious experience are necessary in creating art that is sublime. This talk is part of the Darwin College Science Seminars series. This talk is included in these lists:
Note that ex-directory lists are not shown. |
Other listsGreece and its History Cambridge Neuroscience Interdisciplinary Seminars Faculty of EconomicsOther talksElliptic finite-band potentials of a non-self-adjoint Zakharov-Shabat operator. Data-driven structural assessment Bigger Picture Talks with Professor Chihaya Adachi: The past, present and future prospects of OLEDs Finish CD4+ T cell memory during type 1 immune response A spectrum of physics-informed machine learning approaches for problems in structural dynamics |