You can start collecting and preserving all remaining data about him including your memories and his DNA samples, so future superintelligent AI will be to create a good model of him.
I am very afraid that the best even a superintelligent AI can come up with would be uncanny puppet versions of the people we cherished or rather completely new people with just some similiarities, more akin to clone siblings than to the original individuals. What I actually want is what was left over of his connectome, and that is, for all I know, gone forever. Unless some AI can extract it from thermodynamical noise—which does not seem all that likely to me.
We don’t know for sure what is personal identity and how much data is actually is needed for it, but the data is decaying while we are hesitating to collect it.
That might be right, but it’s also very possible that the connectome will one day prove easier to reconstruct using functional data (like a set of thoughts he would agree with, to fine tune a brain model from a distribution of possible brains given his genetic information). Brain models based on the anatomical data left from present day vitrification technics are probably excellent for C Elegans, but not necessarily as good for larger brains.
Notice I might be biased from thinking to much about this for my own Alzheimer mother. Sorry for loss.
I don’t, for complicated ethical reasons, but if I would my present choice would be:
a frozen DNA sample (or the corresponding genomic sequence)
a detailed biography from interviewing her and anyone still available who can testify how she was perceived at any time point of her développement
any available data from participating in various fMRI, EEG, MEG, NIRS & behavioral experiments in academia (it doesn’t matter it’s academia, but in practice it’s the only place where you can get paid to participate instead of paying 500/hour—just make sure they accept to share their own data to the interested participants)
In the long run a detailed ECoG is probably the way to go, but I agree with Musk that it will likely need some surgeon robot to decrease costs and risks.
You can start collecting and preserving all remaining data about him including your memories and his DNA samples, so future superintelligent AI will be to create a good model of him.
I am very afraid that the best even a superintelligent AI can come up with would be uncanny puppet versions of the people we cherished or rather completely new people with just some similiarities, more akin to clone siblings than to the original individuals. What I actually want is what was left over of his connectome, and that is, for all I know, gone forever. Unless some AI can extract it from thermodynamical noise—which does not seem all that likely to me.
We don’t know for sure what is personal identity and how much data is actually is needed for it, but the data is decaying while we are hesitating to collect it.
That might be right, but it’s also very possible that the connectome will one day prove easier to reconstruct using functional data (like a set of thoughts he would agree with, to fine tune a brain model from a distribution of possible brains given his genetic information). Brain models based on the anatomical data left from present day vitrification technics are probably excellent for C Elegans, but not necessarily as good for larger brains.
Notice I might be biased from thinking to much about this for my own Alzheimer mother. Sorry for loss.
Are you taking any steps to preserve your mother’s data? Can you explain how?
I don’t, for complicated ethical reasons, but if I would my present choice would be:
a frozen DNA sample (or the corresponding genomic sequence)
a detailed biography from interviewing her and anyone still available who can testify how she was perceived at any time point of her développement
any available data from participating in various fMRI, EEG, MEG, NIRS & behavioral experiments in academia (it doesn’t matter it’s academia, but in practice it’s the only place where you can get paid to participate instead of paying 500/hour—just make sure they accept to share their own data to the interested participants)
In the long run a detailed ECoG is probably the way to go, but I agree with Musk that it will likely need some surgeon robot to decrease costs and risks.
If you can recreate even 1% of his consciousness with this kind of data I would be surprised.