The possibility of digitising the human mind is one of those questions that will only be closed by its successful achievement, I think; there’ll always be an argument for its possibility, because the only way to disprove it would be to quantify how personality and mind actually work, and if we could quantify it, we could probably work out a way to digitise it, too. (That said, if someone can chop a hole in my logic train there, I’d be genuinely very grateful to them, because it’s a question that’s bugged me for years, and I haven’t been able to get beyond that point with my bootstrap philosophy chops.)
Philosophical digressions aside, low-grade not-quite-proof-of-concept stuff seems to be the current state of the industry. Via NextNature, New Scientist discusses a few companies trying to capture human personality in computer software:
Lifenaut’s avatar might appear to respond like a human, but how do you get it to resemble you? The only way is to teach it about yourself. This personality upload is a laborious process. The first stage involves rating some 480 statements such as “I like to please others” and “I sympathise with the homeless”, according to how accurately they reflect my feelings. Having done this, I am then asked to upload items such as diary entries, and photos and video tagged with place names, dates and keywords to help my avatar build up “memories”. I also spend hours in conversation with other Lifenaut avatars, which my avatar learns from. This supposedly provides “Linda” with my mannerisms – the way I greet people or respond to questions, say – as well as more about my views, likes and dislikes.
A more sophisticated series of personality questionnaires is being used by a related project called CyBeRev. The project’s users work their way through thousands of questions developed by the American sociologist William Sims Bainbridge as a means of archiving the mind. Unlike traditional personality questionnaires, part of the process involves trying to capture users’ values, beliefs, hopes and goals by asking them to imagine the world a century in the future. It isn’t a quick process: “If you spent an hour a day answering questions, it would take five years to complete them all,” says Lori Rhodes of the nonprofit Terasem Movement, which funds CyBeRev. “But the further you go, the more accurate a representation of yourself the mind file will become.”
It’s an interesting article, so go take a look. This little bit got me thinking:
So is it possible to endow my digital double with a believable representation of my own personality? Carpenter admits that in order to become truly like you, a Lifenaut avatar would probably need a lifetime’s worth of conversations with you.
Is that a tacit admission that who we are, at a fundamental level, is a function of everything we’ve ever done and experienced? That to record a lifetime’s worth of experiences and influences would necessarily take a lifetime? Emotionally, I find myself responding to that idea as being self-evident… and it’s the intuitive nature of my response that tells me I should continue to question it.
9 thoughts on “Personality back-ups: immortality through avatars?”
The possibility of digitising the human mind could be disproven by proving the existence of something that can’t be digitized, like a soul, but that’s another big can of worms. Simulating the brain down to the quantum level without achieving consciousness may effectively prove it impossible, though not logically prove it. Logically proving a negative is hard.
Personally, I think it will be possible; I just hope uploading becomes feasible within my lifetime.
As to the rest of the article, to record a lifetime’s worth of experiences through conversation would take a lot longer than a lifetime, because conversation is low bandwidth, along with the feedback loop of experiencing the conversation. We would need a way of reading memories straight out of the brain, with a bandwidth faster than realtime.
It should no more take a lifetime to download a lifetime’s worth of experiences than it should take as long to read a book as it did to write it. In terms of the “reality” of an uploaded mind… we have to default to the Turing test. If there’s no way anyone can tell they’re NOT talking to the “real” person, then it IS the real person.
There is non-material component of our brains. Duplicate the structure of the brain, the effects of the neurotransmitters, the consequences of hormones, and you duplicate the person. From that instant of duplication, however, the two minds will begin to diverge.
I don’t find the idea of building a simulation/model of my personality interesting at all. And if “encoding” the experiences I had in my lifetime requires that I spend a significant bit of such lifetime training the model, the whole thing becomes even a bit nonsensical. After all, what use would such a model have? To remain to my family as a (probably a bit grotesque) substitute for me when I’m dead, so that they can have with it the conversations I was not available for because I was busy training my backup?
Let’s instead talk about building artificial minds that are not a copy of our own…
This sounds like a fairly involved version of a chatbot, which have been around for a while. These are essentially speech-pattern mimicry bots, which take in vast amounts of text data on a subject, organize it, then simply ‘look up’ the most probable output to a given response based on past interactions.
To say this is approaching becoming human is to say that a machine that analyzes vast amounts of seismographic data, and is then capable of generating a realistic seismographic reading of an earthquake is becoming an earthquake itself. It’s the shadow, the syntax, the pointer, without the pointed-to.
There is no way to prove that another entity is conscious, all one can prove is how far people will bend over backwards in order to believe that a machine is becoming a human.
It is unlikely that personality backups, or any other kind of “Dollhouse-esque” mind recording, will be possible. Consider water in the ocean: just one simple molecule (H2O) with some trace elements and some sodium chloride. Motion of those molecules via diffusion, turbulence, etc., is fairly wall understood. Does that mean we can measure, let alone reproduce, the motion of all the 3.3×10 to the 22 power water molecules in a 1500 cubic centimeter volume (approx. the volume of a human brain)? No, we can’t get anywhere near that resolution, and even if we could, the measurement process itself would likely disturb the motion of those molecules. Now take the human brain and its infinitely more complex and varied array of molecules, neurons, and brain structures- can we even map that accurately at a level necessary to instantaneously replicate it (if we even could replicate it which we can’t- we can’t even *grow* a liver or a decent tooth yet)? No. It’s just that people, especially those without any experience in the relevant scientific field, drastically overestimate our capabilities and underestimate the complexity and sophistication of millions of years of evolution.
One of the incoherencies in the arguments both for and against the ability to digitize the human mind is the assumption that human memory works like a camera or an audio recorder. It doesn’t, not at all. Human memory records some sense perceptions of events in a cluster of memories (somehow, we don’t yet know the physical mechanism or how the clusters are arranged spatially). Then the memories are recalled, usually only approximately, and somewhat interfered with by current sense perceptions and concepts that are in the foreground at the time of recall. Then the memories are restored with the modifications that were made on recall (plus probably more modifications resulting from any thinking about the memories that occurred after recall. When the memory is next recalled, it’s the modified version that comes back, and that in turn is modified before being stored again. There’s no canonical cop of a memory that stores the “original” version.
Maybe we are our memories, but those memories aren’t accurate records of the events of our lives. Instead, as time goes on, they relax towards a narrative we tell ourselves about the events of our lives. In other words, we create ourselves in a dynamic process that constantly changes us. No simple record duplicate that process.
The possibility of uploading the human mind into something more permanent than a human brain is a subject that has been explored in considerable detail by my old friend, Prof. Robin Hanson of George Mason University. See, for example, http://hanson.gmu.edu/uploads.html . Interested readers may also enjoy his regular blog, which can be found at http://www.overcomingbias.com/ .
There is a remaining question : if you upload your brain in a computer and if you die, are you still alive in the computer or are you definitively died and is it just a mere copy of you running in the memory of a computer ?
Echoes of “Learning to Be Me” by Greg Egan.
Quoting from the source:
What is your most successful work – not in terms of financial reward, but from a “personal satisfaction” angle? Why?
“Learning to Be Me”. It was a very simple story, but I think it did exactly what I’d intended it to do.
Not online: you’ll either have to get “Axiomatic” or Interzone #37.
Comments are closed.