How to carry out mind upload practically? Any ways to do it without losing identity?
Mind uploading is said to be a process of scanning a physical structure of the brain accurately enough to create an emulation of the mental state . In that sense, it is like what we already have with virtual machines, that can be suspended and restored to an exact state on another computer or network or virtual machine.
Identity preservation, however, is as important in many cases, because a mind copied, even if exactly the same, cannot occupy the same role in the world to follow through the same future state space. It is similar to how a copy of a virtual machine cannot be instantiated twice on the same memory addresses.
I'd like to reserve this question for searching of practical ideas of performing mind upload -- both identity-preserving and non-preserving kinds, with different applications for various intents and purposes, and not limiting the idea to just humans, but including "mind-uploading" of machine intelligences as well.
Survive by growing neurons out of skull! :)
A way to determine if a copy of a mind is test-indistinguishible from the actual mind copied.
Replace different neural circuits incrementally, and get one's mind uploaded to the digital form without losing one's identity.
Brain Cell Replacement to Beat Aging | Jean M. Hebert, Albert Einstein School of Medicine
The (identity preserving) way I would like to be mind uploaded goes as follows:
Assume we have read/write access to the brain. First my mental state is read. I am then anesthetized. Then my mental state gets booted up in a digital medium. In there I have access to a camera so I can see myself sleep. I then do some first steps in the digital space so i gather some more new memories. With these experiences my digital self goes to sleep and that mind state gets written to my biological mind and booted up again. I now had a continuous experience of going into the digital space and going out of it. I now can repeat that process as often as I like to get comfortable. At some point I might decide to stay in the digital. Or at least when my body is about to break down it is a very calm and familiar experience that preserves my identity and does not resemble death at all.
[sev]，從理想上講，是的，這可能是想法-通往事物的大門-“獨特的記憶鏈？”在這些媒介之間行走？我不得不說，經歷過一次全身麻醉之後，我有了一個“ Aha Moment”（Aha Moment）-思考-_“所以，這是冷凍技術理想的感覺！但是，我確實感覺有點不同步：它感覺到瞬間- -儘管手術花了一個多小時，但我似乎沒有浪費任何時間：這真是一件奇怪的事，我仍然會因爲這種記憶不連續而感到沮喪嗎？“ _ :)所以，我認爲這可能是最接近完美的狀態，低溫技術的未來應該能夠實現。
但是，如果將我定義爲我的不是“獨特的記憶鏈”怎麼辦？ -但是，僞隨機過程編碼在構成大腦的所有組件中，即使只有很小的偏差ε，從理論上講，混沌也會在其產生的動作上與原始動作產生很大的分歧？ （即不是IIMT-x等效）
[sev], idealistically, yes, that would perhaps be the idea -- a door for something -- "a unique chain of memories?" to walk between these mediums? I have to say, having been through general anesthesia once, I had an "Aha Moment" -- thinking -- "So that's what cryonics should ideally feel like! However, I do feel a bit out of phase: it felt instantaneous -- as if I didn't lose any time, although the surgery took more than an hour: what a weird thing, am I still me with this memory discontinuity?" :) So, I think it is probably the closest to perfect, that the future of cryonics should be able to get to.
If the "unique chain of memories" is what defines one's identity, then I could theoretically imagine that another processing unit, like a brain-CPU, could take on a little journey to derive its further actions from that "unique chain of memories" to append new memories based on those new actions, then suspend that, and write new memories back to my biological brain, and let me resume.
To realize this ideal at this level of mind uploading would require that "brain-CPU" be memories-compliant: i.e., allowing itself be controlled by "unique chain of memories" that defines me, and be neutral and unbiased otherwise.
However, what if what defines me to be me is not "unique chain of memories?" -- but rather, a pseudorandom process coded in the totality of components that make up the brain that, even with a very small epsilon of deviation, chaos-theoretically significantly diverges in the actions it produces diverges from the original? (i.e., it is not IIMT-x-equivalent)