An experimental proposal for finding out more about the measurement problem in quantum mechanics
Set up a standard double-slit, which-way experiment.
Contain both the which-way information and the board on which the interference pattern is displayed in a way that can be completely destroyed / erased.
Set up a neural network that is capable of using image recognition to tell the difference between an interference pattern and the pattern created when each slit is observed.
The neural network outputs a 1 or a 0 depending on what it sees. 1 if there’s an interference pattern, 0 if no interference pattern
After the neural network has outputted it’s 1 or 0, destroy both the board with the interference pattern on it and all which-way data, including anything inside the neural network that could be salvaged. the only information remaining from the experiment should be the 1 or 0 output by the neural network.
Have a human being come and check the output of the neural network.
This experiment is designed to separate and remove human conscious awareness from all data related to the double slit experiment.
A human being will know that the experiment should produce an interference pattern — will this affect the results? Or must there be a potential for observation of the which-way data itself in order to collapse the wave function?