What Is Generative AI

March 11, 2023

Generative AI

It creates data using algorithms, discriminative AI identifies differences. Examples: ChatGPT, DALL-E. Limitations include "too many fingers" in artificial art.



Recent advances in Generative AI, like ChatGPT, DALL-E, and Stable Diffusion, have gained attention for creating human-like text and images.


How Does It Work?

Generative AI processes data to determine likely associations, responding to prompts within that probability; ChatGPT and DALL-E are advanced examples.


Training AI models

It is trained through techniques like transformers and generative adversarial networks to create realistic output.


Working Of AI models

Transformers understand semantic relationships, while GANs have algorithms compete to create the best output.


Is It Sensituve

Yes it is Sensitive because it uses complex mathematics and coding to process vast amounts of data and create human-like text and images.


Testing the limits

ChatGPT can answer riddles correctly based on its training set, but it doesn't understand them logically.


Testing the limits

It generates output based on predictions of what should follow a question. If you ask it if two pounds of feathers are heavier than a pound of lead, it will confidently tell you they weigh the same, based on its training.


Thank You For Reading...

Next: 9 Jobs Most 'Exposed' to Artificial Intelligence