A package for generating stories about future concepts that are physically implementable.
Project description
futurephysics
futurephysics is a Python package that generates stories about future concepts that are not to far fetched, i.e. concepts should be possible to be implemented. It uses OpenAI's GPT-4, DALL-E models and Wikipedia to generate the story and corresponding images.
Background
It's great to have all these creative tools. One efficient way to use them is to create glorious visions of humanity's future, which will probably be powered by technology. I am a big fan of the Project Hieroglyph, an initiative of the University of Arizona. This initiative can be best explained by quoting from Wikipedia: "an initiative to create science fiction in order to spur innovation in science and technology." The goal of this repository and the easy-to-use Python package is to replicate this project but with generative AI and Wikipedia. I've also seen a few tweets where people tried to create visionary concepts with AI in Twitter threads:
- Skill Factory
- 3d printing trees should be easier than meat?
- Vienna expands into Antarctic GPU farms
- InventBot
Installation
You can install futurephysics from PyPI:
pip install futurephysics
Usage
Here's a basic example of how to use futurephysics:
from futurephysics import story
#Replace 'your-openai-api-key' with your actual OpenAI API key
openai_api_key = 'your-openai-api-key'
#Generate a story
html = story(openai_api_key)
The html
variable contains the story formatted as HTML.
You can also provide your own list of three categories with which the story gets drafted:
from futurephysics import story
#Replace 'your-openai-api-key' with your actual OpenAI API key
openai_api_key = 'your-openai-api-key'
#Generate a story
html = story(openai_api_key, categories=["Computational archaeology", "Cosmogony", "Plants"])
The story is generated based on these categories.
Since the Dalle3 API is used, the returned image URLs expire after a while. You can provide details to your Google bucket for hosting images (the bucket needs to be public for this):
from futurephysics import story
#Replace 'your-openai-api-key' with your actual OpenAI API key
openai_api_key = 'your-openai-api-key'
#Generate a story
html = story(openai_api_key, google_id="your-google-id", google_bucket="your-google-id")
Contributing
If you run into any issues, the most helpful thing you can do is open an issue on GitHub. Thank you so much!
Roadmap
- Improve image selection from Wikipedia
- Improve image generation via DALL-E
- Improve prompting
License
This project is licensed under the MIT License.
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for futurephysics-0.5-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | df9201b9177670b8916c44acc8161638fb0aec6fbf5c7d1b79a07182b636bf4c |
|
MD5 | 7386fce38ee71f3eab1feac570ee152f |
|
BLAKE2b-256 | 2ff8b7e49c33f2c3fec29b7aa3a0882a2b44af036f751c3ea889e5a289a7921f |