logo
Menu
Is artistic creativity exclusive to humans?

Is artistic creativity exclusive to humans?

Generative AI faces a significant philosophical challenge: can machines genuinely be creative?

Published Mar 11, 2024
From a philosophical perspective, generative AI raises some interesting questions about the nature of creativity, the relationship between humans and machines, and the limits of artificial intelligence. For example, some argue that generative AI is merely a form of advanced pattern recognition, while others view it as a valid form of creativity that can rival human intelligence.
Murat Durmus (CEO @AISOMA_AG) Murat Durmus (CEO @AISOMA_AG)
The idea of building an application that takes care of the various aspects of making a movie came from the question: Can an AI create a movie just by being given the title and some plot lines?
Before explaining the application and how it could be migrated to AWS, let's try to answer this question.
As Margaret Boden has pointed out, if an AI were to become as creative as Bach or Einstein, it would appear to many to be creative only in appearance and not in reality.
Of course, our first impulse is to reject this idea on the grounds that artistic creation is something only humans are capable of, and that it is one of the qualities that distinguishes us from animals.
Some people argue that artistic creation requires an awareness of achievement, but as Stephen J. Gould put it: "If creation requires a visionary creator, how does blind evolution manage to build things as new and magnificent as ourselves?”
Recent videos created entirely by generative AI show us something of what is irrevocably to come....
Alexa, I want a Chaplin and Buster Keaton movie together, one hour and 20 minutes long, set in a traveling circus.

About the project

  1. The first thing we do is to enter in the interactive box the title of the movie and some details about the plot. In the following image we entered a movie called "Echoes of Tomorrow" which is a science fiction movie to watch on a Sunday while eating popcorn.
  2. In the window that says One-Line Synopsis the LLM model gives us a one-line synopsis of what the movie is about.
    capture_01
  3. The Simple scene script window (next image) does not show a sample scene, which can be exterior or interior, also if it takes place during the day or at night and some dialogues of the characters.
  4. The next window Possible cast list for the movie, suggests different actors for the characters and the roles of each of these characters.
    capture_02
  5. In the following image we see a window called Production cost calculation that allows us to know an estimate of the different production costs and also the final cost of the movie.
  6. The Inspiration for the movie poster window shows us a possible image for the movie poster in the style of John Alvin and Bob Peak.
    capture_03
  7. The Difficulties and obstacles to filming window (image below) offers us different points to take into account when filming the movie and the possible challenges we are going to face.
  8. And the Write down some of the requirements for the movie window, which is interactive, allows us to enter some requirements that must be fulfilled as in this case: the budget should not exceed 10.500.000 USD and the movie should be filmed in black and white.
    capture_04

 Migrate to AWS

aws_template
  • To migrate the application to AWS, I would use API Gateway for communication with Cognito authentication.
  • Different projects could be created and each project could have different stages, all this thanks to SageMaker.
  • A Lambda function would act as an interpreter between API Gateway and Sagemaker.
  • Sagemaker would make use of Amazon Bedrock for the creation of LLM models. Also in the future, Amazon Lex would be used to create a chatbot for a more fluid interaction.
  • The Sagemaker results would call a lambda function (perhaps this through Amazon EventBridge), to store the results in a DynamoDB table and some important data would be sent by mail through SNS.
  • The different data generated by SageMaker would also be stored in a S3 bucket.

Links