Ai survey Assignment

Generative Adversarial Networks

GANs (generative adversarial networks) are still in their infancy, yet they have the potential to

raise severe ethical issues. The purpose of this technology is to create an illusionary reality by

employing algorithmic structures to generate two neural networks that are "pitted" against one

another ("adversarially," and using game theory methodologies).

A "generator network" is a program that converts a vector into an image or audio matrix. This

information is then sent into a "discriminator network," which learns to differentiate between

computer produced and genuine content. The two networks work together to train, and while the

generator network improves its tactics for deceiving the discriminator network, the discriminator

network improves its techniques for identifying artificially created information.

Using game theory dynamics to create content that looks and sounds like a recording of reality,

but is actually fake information or fake news, is the end result. While generative adversarial

networks offer a lot of promise for art and political satire, they may also be used to make false

news and ads. A single deceptive video may seriously harm a person's reputation.