MIT IQuHack 2023 Hackathon
Landing page: https://ionq.com/
To recap, for this challenge, you must use a quantum computer to generate something new. Some ideas:
- Make music with a quantum computer (https://arxiv.org/pdf/2110.12408.pdf)
- Render graphics with a quantum ray tracer (https://arxiv.org/pdf/2204.12797.pdf)
- Use procedural generation to make a new world (https://arxiv.org/abs/2007.11510)
- Make a QGAN (https://arxiv.org/abs/2012.03924) Once you've debugged your code with regular simulation, please try our noisy simulators before graduating to hardware (if you have time). Hardware noise can have unexpected effects! We will judge your entry based on both (1) how quantum it is and (2) how cool it is.
We are using a number preserving quantum walk search algorithm running on IonQ (other backends can be included) to output results of a graph traversal algorithm which represents the scattered journey of our main character, Mr. Quanta, and how he tries to figure out what happened to him during his incoherent rambling through QuantaLand. This output will be fed back into a transformers AI model and then to a stable diffusion AI model to generate a exciting dynamic storyline with associated graphics. For a more in depth discussion of the applied circuit, see the following file.
Mr. Quanta awoke on the rooftop bar, feeling a bit disoriented. He had no idea how he had gotten there, but he had a vague recollection of a concert venue. He remembered the music, the lights, and the energy of the crowd. He had been having a great time, but he couldn't remember how he had ended up on the rooftop bar.
You'll need to provide your own API keys as env variables at the beginning on main.py
- provider = IonQProvider("tQgNZln2nI3JSOg7hZhRXjSJHYfgrS2S")
- openai.api_key = "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"
You need to install those python packages:
- qiskit
- qiskit-ionq
- flask
- flask-cors
- openai
To run "Quantum Haze" you need to run the following command:
python main.py
- Victor Drouin-Touchette | [email protected]
- Chris Lamb | [email protected]
- Philippe Cochin | [email protected]
- Gavin Rockwood | [email protected]
- Rob Davis | [email protected]