Discussion about this post

User's avatar
Andy McKenzie's avatar

Interesting post! I totally agree that death is an (important) engineering challenge.

Just to lay my cards on the table, I personally/currently expect that something like branching identity is true - that personal identity can be theoretically maintained via emulation on some sort of substrate, potentially needing to be a neuromorphic or even biological one, but only if it is done a certain way that maintains an unbroken branch of subjective experience. For example, the reasoning for this is described here: https://link.springer.com/article/10.1007/s11023-014-9352-8 and here: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2596460

It seems to me that an implication of your theory is that patients who emerge from long comas may have lost personal identity, even if all of their memories are intact, because their brains would have undergone a lot of changes while they had lost consciousness. So maybe we should be trying even harder to prevent comas because they are effectively death? 

How about deep hypothermic cardiac arrest - do people die when they undergo this form of anesthesia? Or is consciousness not lost for long enough? If so, where is the threshold?

I also don't see how first-person verification would be more reliable during gradual replacement than instantaneous replacement. If someone comes out of a teletransporter and they confirm that they are still them, we just don't believe them? But we do believe them if it is done gradually? Or is the idea that we don't have a teletransporter yet, so we have to wait to develop one before we can test if it works or not, and then if someone does (or enough people do) verify that they are themselves, then we have learned that teletransportation is in fact not death, but because they might not do so, we should work on alternatives? I guess I just don't see how first-person verification during gradual replacement solves the problem of philosophical zombies that you posit. 

It's a complicated topic, so I could certainly be missing aspects of your viewpoints. Not trying to start a flamewar, just thought I would bring up these questions because I believe this is a very important topic.

Expand full comment
Randal Koene's avatar

Hey Daniel, great read and am actually very interested to keep tabs on this work.

I wasn't sure I understood a few things. The claim that causal connectedness is broken, howcome? If a machine reads the brain and constructs a brain based on those readings, how is that not a causally connected chain. Surely, the constructed brain didn't get there by random spawning?

As for the synconetics alternative approach, I'm curious about practical calculations. What is the rate of grafting / adaptation for something like a brain with 86 billion neurons. You'd want to be careful that the rate isn't such that the procedure exceeds lifespan.

Expand full comment
4 more comments...

No posts