This is code for an altCHI 2023 submission. It takes information about an event from a Google calendar, requests a list of objects needed to attend that event from OpenAI's GPT-3, uses Google's Stable DreamFusion to create a 3D model of an object from the list, and prepares that object for 3D printing.

The code has many moving parts and alternative ways to run it.

On a suitably powerful computer, it can be run locally from objectify.py .

On a Mac with Apple Silicon, we instead used [a CoLab notebook](https://colab.research.google.com/drive/1X5OOVEVfu93yNGy2akUXNOPRAwMdDfsr) to perform much of the intensive computation, then ran objectify_cron.py locally with a cron job (every 15 minutes).

Various configurations are also required.

You need to get an API key from OpenAI and an account on HuggingFace. Put the appropriate tokens in keys.py .

You may also need to purchase credits on Google CoLab to run the intensive computation sections.

You will likely want to set your own calendar to be pulled from; change this in objectify.py or the notebook.

If you don't have an SMTP server that you can send mail to yourself from, you can run a dummy locally with:

python -m smtpd -c DebuggingServer -n localhost:1025