arguing on mind uploading

[This is day #17 of my 30-day microblogging challenge. You can and should still join!] (Note: keep in mind when reading ahead that this is meant to be a rough draft!)

Earlier, I was having a debate with a friend (D) who is against neurocomputational functionalism and “mind uploading”. It went something like this:

- For context, there's a few "mind uploading" companies out there that aspire to generate high-fidelity digital clones

- D argues that the concept is flawed, because it will most likely not be feasible, either technologically or economically, to produce an actual high-fidelity copy on an exclusively digital substrate. Mainly because it can't recreate the physical sensations downstream of hormones and other chemical processes.

- I argued that this doesn't matter, and that the copy would still be valuable as long as it could accurately predict what the reference person would say, or how they would behave (for example, while animating a digital avatar). For example, we could probably create such a predictive model by recording the person with 24/7 cameras and voice recordings, and having them explain their thought processes and beliefs out loud.

- D replied that this model would be too simplistic because we would only be evaluate the copy based on language and movement prediction. It would not capture the individual's internal experience, and would therefore neither be "alive" nor conscious. It's difficult (if not impossible, for now) to capture an individual's specific feeling, such as "love" and "excitement". He's skeptical that the model could achieve a high level of accuracy without recording physiological data.

- I was going to push back by arguing that such feelings do trigger external expressions, but then I realized it isn't always true. For example, many people feel strong emotions while keeping a neutral face. I disagreed with the last part about the requirement for physiological data for an entirely behavioral model. However, I could see their point about it not "feeling" like an accurate copy of yourself if it doesn't "feel" like you.

- I still insisted that keeping such an "empty shell" would still feel valuable, because it'd be better than nothing

- He asked "Why would this be valuable to you? You'd be dead, and this would be an incomplete copy for you. Why would you care if it continued after you die?"

- I replied "because I'm human and I have a self-preservation instinct. Even if the copy captures 20% of what feels like 'me', that's better than 0%", and I pointed out people's desire to have children as a legacy, even if they're non-identical copies. This desire comes from wanting parts of you outlive you.

- He also asked how I would evaluate their accuracy without accessing their "internal state" *(which they could never fully replicate on an entirely different substrate)* -> I replied "by having the Reference confirm if they think they would have reacted in the same way as the clone in the same situations"

- He asked "where would you draw the line in terms of how extensive an evaluation to perform before trusting the accuracy?", and argued that the copy would immediately diverge from the reference, because existing in an entirely digital substrate would automatically make it an "alien" version of the reference.

- (The last part came from me agreeing that the evaluation would be difficult, as even an entirely physically-identical copy would probably start quickly diverging from the Reference due to having a different lived experiences, from even just existing in different physical coordinates. I compared this to comparing your current self versus the "you" that existed 5 years ago)

- In general, D's main pushback against my arguments has the shape of "where do you draw the line?" (re: both data capture and evaluation).

- D's main concern is that mind uploading requires destroying the brain in order to map it. Even if the companies don't actively encourage it, failing to deter customers from ending their lives under the belief that their digital counterparts are equally real would be tragic.

- D believes that too much focus is being put on into digital mind uploading, and that this diverts attention from other potentially more promising approaches related to keeping the brain alive past the body's expiration date