Michael Anissimov at Accelerating Future has an
excellent post on the potential benefits of universal mind uploading. Now, I don't mean to piss on anyone's parade here. I'm looking forward to being an upload as much as anyone. But I find myself a little put off by the 'gee whiz' tone of this article, as I am by a lot of a certain strain of Transhumanist though that tends to stress potential over consequences. Being the companion of a lovely and intelligent technophobe, as I am, provides a unique perspective on the hidden flaws in technotopian ideas; and part of the reason I embrace transhumanism as I do is that I see it as the best way of exposing and potentially preventing the dangers of accelerating technological development. If, as Anissimov suggests, working upload tech is available as soon as 2050, these are issues that we need to start thinking about how to address right away; and I think it's imperative that these critiques come from within the transhumanist movement itself, rather from a paranoid bioluddite or mainstream-corpocratic perspective, in hopes that they be taken seriously within the movement and remain as free as possible from authoritarian and anti-enhacement baggage. So all I'm doing here is trying to provoke some debate; because for every advantage Anissimov points out to near-universal uploading, I can more or less immediately think of at least two corresponding potential drawbacks:
1) Massive economic growth. By allowing human minds to run on substrates that can be accelerated by the addition of computing power [...] economic growth [...]will accelerate greatly.
Quite possibly. So let's think about how to avoid:
1a) Massive growth of inequality: Under our present capitalist-corpocratic economic system, this intensive growth in wealth will become increasingly concentrated in the hands of accelerated hyper-intelligent uploads, who would likely be composed primarily of the present wealthy, privileged elite of society. Their capability to influence the economic and political system to their own benefit would increase in proportion. The 'unloaded' - those who are too poor to afford uploading, or who just don't care to discard their meatbodies and become sentient computer programs - would form a vast impoverished underclass, valued only as menial labourers if at all. A second strata of 'cheap' uploads might be able to access the neolife only by accepting crippling conditions imposed by truly 'totalitarian' elites jealous of their newfound place at the lofty pinnacle of society. We must find a way to ensure that the uploaded economy becomes a new source of empowerment and affluence, rather than an entrenchment of privilege and oppression.
1b)Economic abandonment: With little incentive to do anything which benefits the slow and stupid masses of the unloaded, besides providing for their most basic needs or convincing them to join the neolife, the grand explosion of economic productivity occurs almost entirely in virtual space. Perhaps the demesne of the unloaded becomes a massive and permanent welfare state, or a neo-primitive economy where the meatlings are simply left to fend for themselves; or maybe they are rounded up and forced into the neolife to end their irrational drain on the limited resources of the physical universe? We must find a way to ensure that the unloaded can still participate as valued consumers and entrepeneurs in the uploaded economy.
2) Intelligence enhancement [...] By observing information flows in uploaded human brains, many of the details of human cognition would be elucidated. Running standard compression algorithms over such minds might make them more efficient than blind natural selection could manage, and this extra space could be used to introduce new information-processing modules with additional features. Collectively, these new modules could give rise to qualitatively better intelligence.
I would love a piece of this. Here's what concerns me:
2a) Mind slavery: Deliberately reducing, crippling, or engineering uploaded intelligence in order to create a 'productive class' of willing servants. And you know damned well that people will think of a way to do this, and if they think of it they'll try it too. It might be imposed by brute-force programming, or as a condition of 'cheap' uploading by those who can't afford the artificially inflated price imposed on it by godlike plutocrats. And if one has unlimited right to alter one's own spinoffs, one need not even do something so petty as to force or scam flesh-and-bloods into servitude; you could create an overwhelming army of mind slaves with a thought. We must ensure that mind slavery is unthinkable, with no safe havens whatsoever for the perpetrators.
2b) Intellectual stratification: If the ability to enhance intelligence comes at an economic cost, it will be unequally distributed, with the best enhancements going to those with the greatest wealth, thus becoming yet another tool for the entrenchment of privilege. The economic castes maintained in present society through crude methods like authoritarian force, capital accumulation, and systemic denial of opportunity could become literally encoded into the very minds of the human species. Even if all uploads can access the highest possible level of intellectual enhancement freely and easily, what of the poor unloaded, forced to compete with beings that can think rings around them? The best scenario I can think of is that a diminishing population of increasingly demoralized biological humans would somehow find a way to live out their days in dignity. We must ensure that everyone receives the full benefits that intelligence enhancement can bring, and that even the unloaded share in those benefits to the greatest possible degree.
3) Greater subjective well-being [...] With uploading, we will be able to see exactly which neural features (”happiness centers”) correspond to high happiness set points and which don’t, by combining prior knowledge with direct experimentation and investigation. This will make it possible for people to reprogram their own brains to raise their happiness set points in a way that biotechnological intervention might find difficult or dangerous.
I would jump at the chance to escape the depressive tendencies that have handicapped me for all of my life. Still, I worry about these possibilities:
3a) Persona engineering: Laying bare the innermost secrets of cognition and modeling them in easily manipulable software form has nearly infinite potential for abuse. This reinforces the potential for mind slavery, as many of these problems reinforce each other, but has broader applications. The creation of living weapons of targeted rage, of solipsists stuck in permanent dreams, the deliberate induction of all manner of psychoses and mood disorders, are only the tip of the iceberg when it comes to destructive persona engineering. Even well-meaning changes (such as removing the capacity for boredom) could have very uncomfortable if not outright destructive repercussions. We must create a framework in which to judge persona engineering as potentially ethical or unethical, and prevent or punish malicious acts of mind abuse.
3b) The perfect drug: Given the capability, some uploaded citizens will deliberately abuse the capabilities of persona engineering in ways that are harmful to themselves and others in pursuit of subjectively pleasurable, exciting, or novel experiences. They will retreat into solipsistic fantasy or damage themselves potentially beyond repair. The end-state of such a descent into self-abuse is the mind reduced to nothing but a subjective-pleasure maximizer. I advocate cognitive liberty and the right of anyone to do as much damage to their own mind as they see fit, but at the same time it's important to recognize that this is a potential problem (if for no other reason than the amount of computing power an exploding population of self-addicts could clog up). We must find effective non-coercive ways of discouraging and rehabilitating those who choose to abuse persona engineering to detrimental effect.
4) Complete environmental recovery [...] By spending most of our time as programs running on a worldwide network, we will consume far less space and use less energy and natural resources than we would in a conventional human body [...] By transitioning from a proteinaceous to a digital substrate, we’ll do more for our environment than anyamount of conservation ever could.
I'll certainly buy this as a possibility, and would be happy to see it happen; but given the way humanity has traditionally interacted with the rest of the ecosystem (including the ancient hunter-gatherers, who among other things burned forests to flush out game and drove entire herds of prey off cliffs) I find the following scenarios somewhat more plausible:
4a) Resource rush: No longer dependant on the biosphere for survival, but hungry for continual expansion of computing power for their increasingly sophisticated software and proliferating progeny, the wealthy and powerful uploaded society proceeds to loot the natural world for every resource it has left. Whatever forces in the uploaded community fight in favour of maintaining the biosphere as a sort of ancestral terrarium cannot stand in the face of such intensely accelerated demand. The needs of the unloaded population, of course, count for practically nothing; they end up exterminated as a barrier to resource appropriation or simply extincted by the demolition of their habitat. Once this planet's resources are exhausted, neolife colonies branch out to find new worlds to exploit. The end result: a universe tiled in computronium. Protecting the health of the biosphere and its unloaded inhabitants needs to be a fundamental value of uploaded society if we don't want to see it disappear.
4b) Ecological abandonment: Even assuming that essential values and lack of immediate demand for resource appropriation can combine to restrain the uploads from wrecking the biosphere, they might still not be moved to positive action against the effects of the industrial age. Taking a laissez-faire attitude toward the biosphere and the unloaded population which depends on and exploits it, they could just allow natural evolution to take its course and force generations of our descendants to live with the consequences of our short-sighted mistakes, or even stand idly by as the unloaded continue to pollute and factory-farm themselves into extinction. The biosphere will more than likely survive such neglect, but will emerge radically transformed from the natural world that we know. If the uploaded truly value the natural world which birthed them and the biologically based humanoids who would still depend on it for survival, they would need to use their enhanced intellects to find a way to negate the damage we have caused to it and restore it to a pristine state.
5) Escape from direct governance by the laws of physics [...] In a virtual environment, the programmer is the complete master of everything he or she has editing rights to [...] Any civilization that develops uploading would surely have the technology to develop virtual environments of great detail and flexibility, right up to the very boundaries of the possible.
This is one of the most immediately seductive potential benefits of uploading, as well as one of the most viscerally terrifying and fraught with pitfalls:
5a) Reality glitches, crashes, and bugs: If the history of computing has taught us one thing, it is that no software product is ever released perfectly free of unwanted 'features'. The number of problems that could be caused by buggy neolife software is practically limitless, and could end up causing the uploaded endless annoyance, suffering and distress. Having to download patches for buggy code is enough of a hassle in terms of video games and operating systems, let alone for your very existence. I would not upload unless and until I had some degree of assurance that reality glitches, at the very least, could be easily corrected and that any damage to my reality or mind could and would be quickly reversed.
5b) Tortureworld: Neolife has the potential to erase every drawback and limitation of human existence while infinitely enhancing all of its pleasures and fulfillments, but just as much it has the potential to become a neverending hell where suffering is magnified to unimaginable proportions. Whether out of mere sadism, or as a means of enhanced control and domination, the uploaded are almost certain to create the conditions under which the very substrate of reality can be turned into a method of torture. We'd better put a hell of a lot more effort into preventing neo-torture than we have into eliminating the torture we've already got.
6) Closer connections with other human beings [...] By offering partial readouts of our cognitive state to others, we could engage in a deeper exchange of ideas and emotions [...] we’ll engage in much deeper forms of informational and emotional exchange that will make the talking and facial expressions of today seem downright empty and soulless.
Again, to someone who has always had difficulty in decoding and interpreting social interaction this is a seductive promise. There are just a few potential problems:
6a) Mind rape: When the mind perceptions are embodied in software instead of flesh, it will become potentially possible to subject them to much more intimate and disgusting violations than are possible to inflict on the body alone. Even trying to consider what might be worse than mind slavery and reality torture makes me sick and angry; and you know someone's going to come up with something that will be deserving of the term. We have to find out what it is, and do everything in our power to ensure that no uploaded mind is ever subjected to it.
6b) Alienation and solipsism: Given the possibility of creating perfect personal realities, even down to the inclusion of nonsentient 'zombie' companions which react with perfectly plausible subservience, the uploaded may choose to retreat into hermetically sealed bubbles of narcissistic fantasy. Even given the possibility of persona modelling and mind-to-mind communion, the temptations of persona and reality engineering and the potential dangers of allowing increased access to the inner workings of one's mind could drive wedges between the uploaded and encourage massive human alienation. And there's always the problem of relations betyween the uploaded and the unloaded, which will inevitably become more rather complicated and less intimate as the flesh-software divide widens. We will have to take positive steps to maintain and encourage human intimacy and discourage solipsistic thinking.
7) Last but not least, indefinite lifespans [...] By being a string of flickering bits distributed over a worldwide network, killing you could become extremely difficult. The data and bits of everyone would be intertwined — to kill someone, you’d either need complete editing privileges of the entire worldwide network, or the ability to blow up the planet.
Possibly so, though I hope I can be forgiven for having somewhat more faith in transhuman ingenuity to solve the problem of murdering digital entities, and point out the following possibilities:
7a) Deletion and fragmentation: The more secure the system, the more redundant the backup, the more eager some people are to prove their superiority by taking it down. Even in a cloud-computing, globally distributed, holographically stored neolife system, someone is probably going to come up with a way to eliminate every trace of a specific individual (or a particular group of entities) with enough time, intelligence, and motivation. Given such a scenario, however, the easier course might well be to create software weapons which don't try to 'kill' but merely to 'maim' beyond any hope of recovery, deleting massive pieces of personality and memory. Casual murder may well be eliminated, but I'm not convinced that assassination, terrorism, genocide, or simple malicious 'griefing' could be entirely rooted out; though we're certainly going to have to give it our best shot.
7b) The monster next door: Finally, in a neolife of inviolable immortality and worldwide connectivity, there is the problem of what to do with those who violate the prohibitions against willfully malicious acts. If killing an upload is really impossible, the death penalty is off the table. Would there be any way they could be effectively sequestered from the rest of society? Could their persona be engineered to remove the potential to harm others (and is that something that can be ethically justified)? I have no doubt whatsoever that some people, for any of a variety of reasons are going to try to enslave, cripple, torture, violate, and kill other uploads. We have to start thinking right now about how those people are going to be dealt with.
Again, the intent of this litany is not to try to discredit the value of uploading or any of the listed potential benefits, but only to open up some potential questions for debate and inspire my fellow transhumanists to think about how to address these potential problems. If relatively safe, easy, and inexpensive uploading is indeed available in less than 5 decades, I would seriously consider taking advantage of such a service; but I'd be a lot more confident if we had at least figured out some potential solutions to these problems.