By the year 2100, will there be a generally accepted mathematical theory of consciousness that predicts that any of the following contemporary as of 2023 GPT models exhibit consciousness at inference time:

https://huggingface.co/mistralai/Mistral-7B-v0.1
https://huggingface.co/meta-llama/Llama-2-7b
https://huggingface.co/meta-llama/Llama-2-13b

https://huggingface.co/meta-llama/Llama-2-70b
https://huggingface.co/tiiuae/falcon-40b
https://huggingface.co/tiiuae/falcon-180B

I picked these models because they or either already free software or can be archived until the point in time their copyright runs out, so they are unlikely be lost to history.

Here consciousness is defined as whatever the "consciousness research" community considers to be consciousness at the time general scientific consensus is reached on a mathematical theory of consciousness. This does not necessarily have to include valence. Inference-time means when running on a electronic digital computer as a program.

Resolves to YES if there is a generally accepted mathematical theory of consciousness at any point in time before 2100 and any of the above mentioned models are predicted to be conscious by this theory. Even a tiny amount of predicted consciousness counts so be wary of panpsychism risk.

Resolves to No if there is not a generally accepted mathematical theory of consciousness by 2100, or if there is a generally accepted mathematical theory of consciousness and it predicts that all of these Models will have zero consciousness at inference time.

I will not vote, and I will try to be a good sport about resolving this market fairly.