Artificial intelligence has been around since the early 2030s. Since the Network War of the 2040s, every Earth nation has fiercely enforced laws regarding limitations on AI freedom to act and ethical or moral independence, and controlled their ability to self-upgrade and network.
Nevertheless, AI is too useful a tool for a spacefaring humanity to turn its back on, and AI nowadays comes in one of two forms: Installation, or Human-Basic.
Installation AI is generally found running the processes and activities of a major technical or industrial installation – they are of necessity networked to their installation technology, and are usually hard-limited to a level of reasoning required for their task and no more. They do not have “personalities” and their freedom to act outside their programming parameters is usually very low. All colony ships have a pilot AI, which attends to hibernation, memory retrieval, departure and approach burn, and essential ship maintenance inflight. Installation AIs are not great coversationalists.
Human-Basic AIs are independent units, installed into a specifically cloned and designed human-basic clone meat body. They are unable to directly interface with other technology save through the same methods as humans – eyes, voices, fingers. Originally created to provide AI analysis and support in scenarios and situations where Installation AI was inappropriate or counterindicated, in practice their development has led to their identification as a worker underclass in Brazilia and Utopia, and their complete banning in Antarctica. Human-Basic AIs are limited to an intelligence and reasoning level appropriate to their designated tasks – and in any case, are hard limited to human median + 25% across a variety of intellectual and cognitive indicators. Further, Human-Basic AI bodies are usually visually distinct from humans – the method varies, but metallic skin and barcoding are both common methods. The days of unidentifiable superintelligent AIs walking among us came to an end in the 2050s.
In both cases, the preservation of human life, calibrated to an appropriate level, is the overriding concern of any AI. There is a key parameter built into every AI, whereby at what point they are prepared to accept the death of one or a number of humans for the greater good, is set. This parameter – the “trolley problem limiter” – is set in slightly different ways depending on the cultural background of the AI’s programmers. Utopian programmers will always try and limit the AI’s ability to make such decisions, and will place the common good as higher than the individual good; Brazilian AIs tend to the other end of the spectrum with regard to this issue, favouring the safety of the AI’s immediate charges over a hypothetical cultural good.
There are four Human-Basic AIs and one Installation aboard the colony ship Ovid; they are designated Newton (the Installation), Planck, Curie, Avogadro and Lovelace. Their responsibilities are medical support, astro-navigation and communications, engineering and repairs, and administration and colony integration.
Since initial flight tests in the 2030s, travel between the various colonies in the Sol system, and then in the early 2100s interstellar travel between the Sol, Gliese and TRAPPIST systems has utilised the EM Drive, a microwave-driven propulsion system which builds thrust incrementally up to close to relativistic speeds over time. TRAPPIST is 39 light years from Sol and Gliese 581 is 21 light years from Sol. Unfortunately, the generation ship model cannot safely accelerate much beyond 0.6c without the crew experiencing severe and life-threatening consequences. Allowing for initial acceleration and deceleration curves, the journey to TRAPPIST aboard a generation ship takes 68 years, and to Gliese 38.
With journey and comms time delays measuring in the decades, the great powers of the Sol system swiftly realised that their ability to influence or control the extrasolar colonies is negligible – and while the Federation primarily sends ships to Gliese and the Union primarily to TRAPPIST, all three powers continue to send ships to both colonies.
Sol knows only what the two colonial systems have transmitted back across the distance – both colonies are successfully established on multiple habitable planets, and both are crying out for more colonists. The view is that weight of colonists and speed of advances will be the real measure of who controls such systems, and Sol’s resources now are bent to increasing the number of colonists sent to each system, and to making that journey easier.
Flight times between the systems still measure in multiple decades even with the benefits of the EM drive, and while most travel is conducted by unmanned AI-piloted cargo sleds captured at destination by the source system control station, colonisation and research missions still require a manned solution.
Cryosleep and Memory Storage
While cryogenic freezing and revival technology has been fully understood and relatively safe since the mid-2050s, the process of revival from cryosleep destroys the surface brain structures responsible for the storage of long-term intellectual memory in the cerebral cortex, while leaving instinctual muscle memory and some learned skills intact. Sleepers have been revived from cryostasis sessions ranging from a few hours all the way out to full decades – and in every case, the awoken sleeper has irrevocably lost all long term intellectual memory.
Early 2080s advances in neural net mapping, cortex wet tech read-write techniques and real-time offline storage offered a solution. Deep space sleepers have their entire memory structure recorded and stored in non-wetware hard storage – usually a small solid-state storage device kept with their capsule during the sleep process – and have their memories re-uploaded to their brains upon awakening.
This process, though now commonplace for deep space interstellar travel, is an ethical minefield. During the early experiments in memory re-upload a number of key limiting factors were determined in the process.
Uploading memories to a human brain can only be successfully achieved if no existing memory structures are in place – so it’s impossible to upload memories unless the mind to receive them is blank, as after a cryosleep episode. Attempts to upload memories into a mind that already has existing structures for long term memory storage destroys the memory structures in place and often corrupts the uploaded source data too. Rare examples of successfully integrated memory from two people exist in the literature – usually from two subjects with a close bond or very similar experiential profiles – but they are impossible to predict successfully and often have severe mental and physiological health impacts in the medium term.
Equally – attempts to upload memories into bodies other than those from which they originally came, while generally successful from a point of view of accessing the recorded memories in the new brain, usually lead to a distinct and marked deterioration in physiological health for the uploadee in the short to medium term, consistent with the effects of a neurodegenerative disease. Death often follows within the 6-12 month timescale. Research on the causes of ERPDS (Engram Reupload Physiological Deterioration Syndrome) is ongoing, but initial theories revolve around a build-up of mismatched signals where recorded memories and body muscle memory clash or have different expectations.
Of course, research in the SOL system on long-term effects of cortex remapping is ongoing – it is unclear what long-term effects the colonists in the outer systems may be experiencing from the process, as communications from those systems have a decades-long lag, and to SOL’s point of view, their window into what is happening in the systems is very out of date.
Recovery from cryosleep is a delicate process, with recommended medical regimes to follow to shake off what can be years of physical stasis; the process is not without its safety issues, though nowadays recovery rates are around 98%.
Ethics and Spirituality
Numerous religious and spiritual bodies have prohibited or spoken out strongly against the process, regarding the essential “soul” of someone as something it is morally repugnant to remove from them. Philosophers and ethicists have wildly differing views on the impact of memory removal and re-upload, and whether or not a person whose memories have been removed and then, at a later stage, re-uploaded can truly be said to be “the same person”.
Is a personality the sum of recorded memories or is there an intangible “something more” – and if there is, does that intangible survive the extraction, cryosleep and restoration process? Does the removal of memories return a human, as some fringe Judeo-Christian sects believe, to a state of innocence? Or do the acts of a person whether remembered or not leave a mark on this intangible soul?
Return to Tabula Rasa Main Page