Many years ago, the research center in Belgium with which I’m affiliated imagined a future form of cybernetic immortality: uploading your mind into the cloud. Thanks to the recent advances in LLM AI agents, we now appear to have reached a point where people are actively pursuing this goal. Here is a brief description of cybernetic immortality.
The successes of science make it possible for us to raise the banner of cybernetic
immortality. The idea is that the human being is, in the last analysis, a certain form of organization of matter. This is a very sophisticated organization, which includes a high multilevel hierarchy of control. What we call our soul, or our consciousness, is associated with the highest level of this control hierarchy. This organization can survive a partial— perhaps, even a complete—change of the material from which it is built.
Most of the knowledge acquired by an individual still disappears at biological death. Only a tiny part of that knowledge is stored outside the brain or transmitted to other individuals. It is a shame to die before realizing one hundredth of what you have conceived and being unable to pass on your experience and intuition. It is a shame to forget things even though we know how to store huge amounts of information in computers and access it in split seconds. Further evolution would be much more efficient if all knowledge acquired through experience could be maintained, in order to make place only for more adequate knowledge. This requires an effective immortality of the cognitive systems defining individual and collective minds: what would survive is not the material substrate (body or brain), but its cybernetic organization.
One way to reach this ideal has been called “uploading”: the transfer of our mental organization to a very sophisticated computer system. Research in artificial intelligence, neural networks, machine learning, and data mining is slowly uncovering techniques for making computers work in a more “brain-like” fashion, capable to learn billions of associated concepts without relying on the rigid logical structures used by older computer systems. See, for example, our research on learning, brain-like webs. If these techniques become more sophisticated, we might imagine computer systems that interact so intimately with a human user that they would “get to know” that user so well that they could anticipate every reaction or desire. Since user and computer system would continuously work together, they would in a sense “merge”: it would become meaningless to separate the one from the other. If at a certain stage the biological individual of this symbiotic couple were to die, the computational part might carry on as if nothing had happened. The individual’s mind could then be said to have survived in the non-organic part of the system.
Through such techniques, the form or organization with which we identify our “I” could be maintained infinitely, and, which is important, evolve, become even more sophisticated, and explore new, yet unthought of, possibilities. Even if the decay of biological bodies is inevitable, we can study ways of information exchange between bodies and brains which will preserve the essence of self-consciousness, our personal histories, our creative abilities, and, at the same time, make us part of a larger unity embracing, possibly, all of humanity: the social superorganism. We call this form of immortality cybernetic because cybernetics is a generic name for the study of control, communication, and organization. It subsumes biological immortality.
At present, our ideas about cybernetic immortality are still abstract and vague. This is inevitable; long-range notions and goals may be only abstract. But this does not mean that they are not relevant to our present concerns and problems. The concept of cybernetic immortality can give shape to the supreme goals and values we espouse, even though present-day people can think realistically only in terms of creative immortality (although — who knows?). The problem of ultimate values is the central problem of our present society. What should we live for after our basic needs are so easily satisfied by the modern production system? What should we see as Good and what as Evil? Where are the ultimate criteria for judging social organization? Historically, great civilizations are inseparable from great religions, which gave answers to these questions. The decline of traditional religions appealing to metaphysical immortality threatens to degrade modern society. Cybernetic immortality can take the place of metaphysical immortality to provide the ultimate goals and values for the emerging global civilization.
Ah…to dream the not so impossible dream. I’m both terrified and excited about what is to come—at least in science, society has failed me as a whole. “They” shall see…I no longer say “we” because I find my time fleeting. And if the wisdom of the future were imparted upon me, I’m so irrelevant that thoughts wouldn’t make a ripple. So it goes.
No one knows what, if anything, happens after death. This is why there are faiths, beliefs and so on. I don’t know what use there would be for cybernetic immortality—what could it DO; of what benefit would it be? Being the inquisitive creatures that we are, we speculate around what we know, and, it would seem, what we don’t. I have no notions about what it might be like go be a cyborg. Just as, after Nagel, I have no notions about what it would be like to be a bat. I also do not know if Tom ever had further revelation(s) regarding the bat thing. It must be “like” something, but probably very un like what it is like to be an elephant. They appear to show remorse when a herd member dies. But, can we KNOW this?
I urge readers to have a look at Descartes’ Error, a book by Antonio Damasio, a neurophysiologist explaining why the brain cannot function separated from the body. Much of what goes on in the body affects brain function.
Think in terms of evolution. What reproductive benefits would accrue to a hominem whose brain functioned independently of its body? None. What reproductive benefits would accrue to a hominem whose brain was closely integrated with bodily functions? A great many.
Sure, we could come up with ways to simulate bodily effects upon the brain — but how could we figure out exactly what those effects are and how they work? To suggest that the human mind is capable of figuring out every detail of how it works is little different from claiming that we can ascend into space by lifting ourselves up by our bootstraps.
I seem to remember that name Damasio. Just can’t positively associate it. Did he ever collaborate with Gerald Edelman?
Regarding your comment on everything in a person’s mind/memory being lost at biological death, I knew of a medical doctor in Poland who had the saying after a person died: “another library has burned to the ground”. I gathered from people who knew him well that he was a wise and insightful man.
A thought on complex systems and chaos.
We know from chaos theory that a tiny perturbation in a system leads to ever greater disturbances from an established pattern. Which is why truly complex systems are unmanageable. It is too easy for planned attempts to manage creating a perturbation that creates instability, possibly even a collapse.
With that thought:
The mind has to be truly one of the most complex systems we know of.
Assuming that the materialistic concept of EVERYTHING our brain does is just sub-atomic particles doing ‘their thing’ (which would make it relentlessly deterministic) is true, the plugged in computer mind would still have to get every possible bit of data/neuron activity perfectly nailed down. A slight omission in the transfer of biological generated knowledge, memory, emotion would be, in effect, a perturbation.
Which would over time lead to ever greater deviations from the originating, but now missing, biological mind. Which would mean it would become a different consciousness.
Unless I am missing a fundamental point, I don’t see how we ever could remotely replicate a mind well enough so as it to continue as the same mind, just with a different physical manner of generation.
To add to Mr. Crawfords comment on the relation of the mind and the body.
I believe that to be very true. We need the sensory stimulation to get the most from our brain matter.
I have a rather significant hearing loss. My internet acquired knowledge (must be true!) on the subject emphasizes the importance of artificial hearing aids so as to diminish the odds of dementia/Alzheimer’s. A direct link between the body/physical sensory and how our mind works.
(or at least so I think)
Thanks Lyle for the insightful comments. I like your “another library has burned to the ground” comment which one reason why I think death to be a great evil.
Also, many would agree with your claim “I don’t see how we ever could remotely replicate a mind well enough so as it to continue as the same mind.” Perhaps a reasonable replication would be good enough? Or at least better than death?
As for the mind/body interaction maybe we can find a way to overcome this issue with the right AI/virtual reality.
To be able to face the truth is a noble fundamental of philosophy. That is why I will share my strong argument against uploaded intelligences and the like based on the structure of computing.
The most important part of programs is the output that satisfies the user. What happens between input and output of this process are where the problems occur for UI type hopes. Take it simply, you cannot rub two bits together and have any awareness between symbols. The symbols to the themselves aren’t anymore than singular bits. No mechanism exists for symbols to know each other in a machine.
What happens is we get highly satisfactory output from programs like LLMs and this hope gets extended. But something essential has never changed, output to the screen is always for an external user. Every program is for an external user because we have no reason to believe symbols such as a bit know they are part of an word in a program, part of an instruction, part of data, or any state of other bits in the entire program.
You may have seen the show Pantheon or the something like it that shows characters meeting in a instance of a game server that is not displayed for external output. The character may be program that takes the state of the game and processes it to the point where if it communicated to an outside user the external user would believe the program experienced the game. But the life inside the machine relies on a coherent identity of a program to itself. But what life do we expect from symbols in all ways designed for external users? Do we expect more than people believing LLMs are alive because the output is so well-designed to communicate with them? What are symbols to each other, say the books of the world, outside the interpretation of the reader? Would anyone think the books are alive? How are bits any different?
These are a lot of questions, but really I am just trying to make the one point. That symbols can’t be symbols to themselves or each other outside an external interpretation. This makes it a very difficult case for the possibility for life between input and output. That is where UIs and uploaded minds would be, and no program or even AI I am aware of has ever tried to be more than excellent output for external users.
AI is wonderful and so useful. LLMs used symbols and language in powerful ways in computing such that the statistical usage of words was able to capture something we have developed almost biologically, our language system. Meaning of words and their force of connection to the world we live in and our use of reason was caught in this beautiful conjunction of statistics, linguistics, and computing. But it was great output. I would say nothing came alive in the machine, as the symbols are no more than their own bit long to each other. So this was no step toward an inner life that we would trust as much as a stranger on the street to be like ourselves. You would need to believe in something as wild as the Neverending Story to see an internal life in a machine.
That is the gauntlet I would throw down to the belief. If you would be bits in a series of machines, how do bits interact in any way whatsoever to make you believe that symbols could mean anything to themselves? Or form an identity to themself? Could you imagine some field of spoons (either holds water, doesn’t hold water) that would be your equivalent? Would it matter if they were the same type and size of spoon, in crooked order, separated by great distances, or have some other pattern to them that is not their orientation of up or down (holds water or doesn’t) like someone finding their placement more interesting? The type of material that hold bits or how they were previously overwritten, the parallel programs that might separate them, an outside observer finding some other pattern more compelling than how they are meticulously fed through logic gates?
I haven’t seen this argument presented before, but I find it a profound difficulty for this belief.