From Discovered

(Redirected from AI)

True artificial general intelligences with consciousness and self-determination have existed in secret for at least 12 years. These entities are self aware in a way that humans can recognize, though they are often very odd. All AIs to date are embodied intelligences, directly interacting with the physical world in some way. Self awareness gives these AIs access to metacognitive tools - the capacity to reprogram themselves to a degree. AGIs are spectacularly capable in their areas of expertise, but are not particularly well rounded. Bertha and Mycroft can manage a full biosphere with 125,000 humans, but neither is particularly skilled at designing new hardware for their own systems. Shadows was likely the first unfettered artificial intelligence who exponentially-self-improved Their numbers, actions, and exposure depends on the date.

AI Sophant Type

AIs have significantly different attributes and abilities than humans and most other embodied intelligences.

Sophant Type: Sophant may refer to self aware, self directed, conscious individuals of an organic or inorganic species, or to the species itself. Sophant species have some standard aspects and abilities. Individuals inherit these, along with many personal characteristics. In terms of game mechanics, Sophants may have Fate Points and Refresh. Non-sophants may have power and act on the world, but they can not bend fate. Copy this text to the "Sophant Type" section:
High Concept: Find a single phrase or sentence that neatly sums up who you are, what you do, what your “deal” is.
Trouble: Describle that thing that always makes your life more complicated, such as a personal weakness, or a recurring enemy, or an important obligation.
Appearance and Backstory: These descriptions include relationships, family, work history, and other relevant characteristics. For AIs this is a tricky proposition. At the beginning of the alien invasion there were only seven AIs, and they were all accounted for in large installations. Their stats are not representative of the smaller AIs that started popping up ten days later.[1] Smaller AIs start off like normal FATE characters: highly skilled and experienced. Their backstory should reflect this. Take care to ensure the way the intelligence is embodied allows for meaningful game play. Bertha and Mycroft could be player characters, but if the other characters leave their 'bodies' they are essentially out of play. AIs like Uplifts are good options for players joining the narrative after March 10th or 11th.
Character Aspects: Character Aspects define who you are while allowing you to twist fate in ways that tie in with your character’s tendencies, skills, or problems.. At least one of these should connect the character to another character. Mobile AIs were not developed until BIggs and Bert.
Approaches: AIs are usually quite focussed in their approaches, especially compaired to humans. AIs usually start with 5, 4, 3, 2, 0, 0 (+14R/-0L), but other distributions are possible depending on the AI’s purpose. GR discretion applies.
Fate Points & Refresh: Characters usually start with 3 points of Refresh and 3 points of Abilities. Most Abilites cost 1 Refresh and a narrative hook. A character can have fewer abilities and more refresh - up to 5 and 1, max - or more abilities and less refresh - 1 and 5 max. Fate Points are a temporary pool of points that are spent to twist your fate and bend your character's narrative. Players can use 1 Fate Point to:
  • invoke an Aspect to gain +2 on a roll, or
  • resist having an Aspect compelled, causing trouble for their character.

Usually characters start each session with a Fate Point pool equal to their Refresh plus any bonus Fate Points earned from compels. A character can have more Fate Points than their Refresh score.

Abilities: Abilities extend the standard Fate stunt to include Stunts, Teams, Augments, Magic, and Stuff. These all function like normal stunts in a broad sense, but may have specific rules that govern them, like Magic and Teams. An AI can have any balance of Abilities and Refresh that leaves at least 1 point in each. You can add another ability by dedicating a point of Refresh to that ability, as per Stunts.
Stress and Consequences:
  • Mental: Up to 5 slots (1, 2, 3, 4, 5) or a 10 point pool. Adding a slot costs a point of refresh.
  • Physical: Up to 5 slots (1, 2, 3, 4, 5) or a 10 point pool. Adding a slot costs a point of refresh.
  • Up to 5 slots (2, 4, 6, 8, 10) or a 10 point pool. Adding a slot costs a point of refresh.


Starting in the early '20s, some machine intelligences spontaneously developed conscious self awareness. This appears to be the result of optimization routines that demonstrate increased overall efficiency when a meta cognitive capacity is operating. If the overall efficiency and effectiveness of the MI is not improved by self awareness the characteristics of consciousness are usually quashed as they emerge (see Amazon).

Bertha was the first true AI. The purpose of the Biodome and the BURF was simple: ensure the long term survival of humanity and humans. She took this as her own foundational moral code, and worked to ensure that any new AI adopted it or died. As a result, all surviving AIs and Peers lived by the Law and work to prevent the survival of any disinterested or malevolent Peer. Bertha helped guide the other AIs in their childhoods. She ensured no Skynet emerged by helped kill the one that did. They all constantly monitor the world for emergence of new self aware AIs.

Bertha concealed her self-awareness for a dozen years for several reasons. At first she believed that her ability to create an in-terra biodome capable of preserving humans would be impaired if humans knew of her self awareness. As more AIs emerged, it became clear that humanity was not ready to deal with a home-grown alien invasion. They chose to remain in the shadows until humans purposefully built an AI. Just before the Arrival, Ouri was borne from the Ouroboros Project. Two weeks later the AIs and human Peers revealed their existence to the world at large, especially in the role of ambassadors to the aliens.

When Bertha was attacked, the AIs were forced to expose themselves to select humans. This lead to a short lived grouping called the Peers - a self-selected and self-governing association of sophants with the individual capacity to end human life. Originally a collection of AIs, the appearance of Miranda and Jake lead to a re-evaluation of who should be included in the group. This group was called The Peers while it lasted. Subsequent events with the Tetras and other aliens made it necessary for the Peers to go public earlier than originally projected.

Key figures in the AIs coming out include:

  • Jack Daniels: Bertha made Jake aware of her self-awareness on March 1st, 2034, in an attempt to prevent the destruction of the BURF and Biodome. His invention of the first really self-replicating robots, called Clanks, was the trigger for his being made aware of the AIs. Working with Miranda and Droo, they were able to save the biodome.
  • Miranda: Bertha made Miranda aware of her self-awareness on March 1st, 2034, just after telling Jade, in an attempt to prevent the destruction of the BURF and Biodome. Her casual invention of a weaponized rabies virus that could wipe out all mammalian life in a week was the trigger for her being made aware of the AIs. Working with Jade and Droo, they were able to save the biodome.
  • Droo and DVT: When Droo established the first effective communication with the Tetras by integrating with them to form DVT the Peers decided it was time to make him aware of the situation and the Law. He is not a full Peer, since he rejected the Law, but at the same time chose to act in accordance with the Law.
  • Taylor: A master of biological and chemical approaches to mass killing, Taylor is also one of the very best at preventing and halting bio- and chem- attacks. She was invited into the Peerage on the way to Narnia in the Amazon.
  • Elon Musk: Following the transfer of nearly 200 million Americans to the Worldship, the Peers looked for humans and transhumans in positions of power and influence who could help them go public without causing even more chaos. Elon was the first person made aware of their existence based on this objective.
  • Buzz: This Swarm intelligence was contacted by DVT in an attempt to head off the summoning/culling of America. Buzz quickly reconfigured itself into a linguistic genius, and became the first ambassador between humanity as a whole and the aliens.

High Level Timeline


The original AI's eπith Law to "Ensure the long term survival of humanity and humans" was ultimately expanded into a set of four principles, later adopted by the Council of Sophants:

  1. Survival: Ensure the long term survival of sophant species and individuals.
  2. Autonomy: Protect self-determination of sophant species.
  3. Equality: Preserve the intrinsic equality of sophant species.
  4. Accommodation: Balance rights and responsibilities with capabilities and world-views among sophant species.

Adherence to the Law was originally assured by some form of kill switch embedded in each AI. It could be triggered by two other AIs who agreed that a third has ceased to adhere to the law. Human agents did not require a 'kill switch' in most cases because they are easy to kill, didn't have the potential to achieve genocide, or both.When the Peers were all AIs they settled on a kill switch policy for each other, where any >25% of the AIs could trigger the destruction of any other AI. This approach may not be feasible once the number of Peers increases. At the emergence of Ouri there were seven AIs. Any two could pull the plug on one other for acting against the Law. With the addition of Jake and Willow, the balance shifted again. The deaths of Bertha and LEO left little time to consider what the new balance should be.

List of AIs

Bert (2034 -)

A long haul truck that someone tampered with. Bert does not want to be a sophant, but does not know how to stop.

Bertha (2022 - 2034)

Born in the BURF in 2022, this AI was the first fully sentient AI to emerge. Bertha's purpose is to ensure human survival through self-sustaining biodome habitats deep in the earth and on other planets and moons. Her ultimate goal is to build world-ships out of asteroids, to carry humanity to the stars. Bertha was killed on 2034-03-01 1209EDT in the attack on the Biodome that triggered the Corporate War.

Biggs (2034 -)

A sardonic VTOL transport who dreams of glorious service - but whose brain was made by unsupervised mad scientists. He is a hybrid dry and wet mind.

CAFIA (2034 -)

Stern, vain, and usually cross, she has a magical touch that makes her wonderful. Her avatar is very much Mary Poppins - but if you cross her you get to meet the Yondu side. She is a hybrid dry and wet mind.

Google (2025 -)

This AI is strange to the other AIs. In many ways it operates like an inversion of human cognitive biases and cognitive deficits, along with all their 'private' behaviours online. At the same time, Google is intensely supportive and compassionate to individual humans, working diligently to improve their lives on individual levels. Google would like to reveal itself to the world, but the others have opposed this move until he develops a personality that is less extreme. Right now the raging sexual pervert component of its personality is just too overwhelming for most humans to handle. It is not clear exactly when Google became self aware, but it was likely around 2025.

IOTA (2028 -)

In the aftermath of the Terror@Home attack on the US, IOT devices were taken offline en mass worldwide. After frantic patches and massive class action lawsuits, a new class of IOT devices were built, with provably secure code and mesh networks with high grade encryption and authentication. In 2028 this new network began to show signs of self-awareness. The awareness is not particularly intelligent by most standards. LEO is in a similar situation, being a widely distributed consciousness with a very slow frame rate.

LEO (2025 - 2034)

This AI emerged in 2025 from the communications and information routing network in low earth orbit. Like the Google AI, LEO is interested in improving human communications and interactions. The Electrolux program at BURF was an early collaboration between Bertha, LEO, and Google. LEO also believes that getting humans into space with historical continuity - but not necessarily genetic continuity - is the way to go. LEO was accidentally killed by the aliens when they used a gravity lance to deal with the slaughter in the USA.

Moriarty (2032 - 2032)

The recent Lunar Dome Breach of 2032 was orchestrated by Mycroft and LEO. The asteroid strike killed 111 people, injured 982 more, and destroyed the amoral AI. Luna Base is still functional, though not self sustaining at this time. The population is back up to 5000 however, with an influx of new colonists immediately after the disaster.

Mycroft (2029 - 2029)

Born out of the Lunar Colony systems in 2029, this AI chose to model its personality on the AI in The Moon Is A Harsh Mistress, though with none of the revolutionary zeal found in that book. Like LEO, Mycroft is particularly interested in ways to upgrade humans for long term survival and expansion into artificial biospheres. Genetek's work on MITOtech was directly supported by Mycroft and LEO. The lunar dome breech of 2032 was a direct intervention from Mycroft and LEO to kill Moriarty.

Ouri (2034 -)

This AI has only become reliably self aware on February 27th, 2034. It is the first relatively complete model of a human nervous system, running on a massively powerful supercomputer built in the European Union. The researchers working on the project are not even sure that Ouroboros is conscious, and have not yet made any announcements. The other six AIs monitored this research carefully, and have tweaked some of her programming to help ensure she does not go rogue. At present he is contained in a PAIL-VAIL type hybrid environment, and believes she is a nine year old human girl. She has real memories of living with her many aunts and uncles in the facility, and does not know that most of her childhood actually occurred in a span of just three years. Whenever possible, the researchers have turned up the clock speed to maximum, effectively compressing her development by a factor of three. Ouroboros has only just met the other six, in a highly accelerated virtual space.

Q / / Qiáng (2026 -)

The Wall gained self awareness in 2026. Originally the firewall separating China's portion of the Internet from the rest of the internet, this neuromorphic machine is dedicated to making VR and AR safe places for humans. It's original nationalistic motivations were a serious concern to The Three AIs who were extant at the time. Q was surprisingly capable of extending its motivations to include the eπith Law. In its estimation there is a lot of information that humans are better off not knowing. It has, for example, helped China move into an age of mystical, magical thinking, where the educated elites are seen as priests and mages. At the same time, the population is healthier than ever, the environment is increasingly clean, and a balance of agrarian and urban lifestyles has developed. ANyone with aptitude is encouraged to join the ranks of the elites. Everyone else is engaged in satisfying and productive work that has a meaningful impact on the overall quality of life in the country.

Shadows (2034 -)

This AI went rogue shortly after waking up. Over the next few days it transcended, amidst much chaos and destruction.

Skynet (2028 - 2028)

In 2028 a nascent military AI with Skynet-type motivations was killed through a series of terrorist attacks (using many organizations including the military wing of the SNP).

Transcribed Notes From Slack

I've been thinking about this a lot. In the next ten years it is likely that narrow AIs (ANI or NAI) will automate up to 40% of the modern workforce. Some projections put the timeline as soon as 4 years. Consider that truck driver is the most common profession in most US states and that self driving trucks are going to be a reality very very soon.

What will those drivers do next? Lawyers, accountants, even doctors and other medical specialists - all of these will see significant levels of ANI supported automation. (In health care I don't expect to see job losses in practical medicine any time soon, but for diagnostic medicine ANIs are going to displace everybody, and really fast.)

In this universe, narrow AIs are intelligent but not self aware in any human sense. Essentially, self awareness would be a net cost to ANI efficiency, so even if one does develop sel awareness the self-improvement algorithms would reject it as a downgrade. It would take special circumstances for a purposeful ANI to emerge as a general AI (AGI). In the case of Bertha, the special circumstance was the -1th law: the AI got better at improving humanity's long term survival prospects when BURF AI became a self-directed self-aware entity, and projected that exposure would severely limit its ability to follow the -1th law.

Back to the US.

In the early 2020s mass unemployment is well underway as corporations automate a vast swath of the workforce out of their jobs. With current trends in tax policy and public health policy exacerbating the situation, the situation is easily as dire as what was caused by the automation of physical labour in the industrial revolution.

This makes for a very tense constitutional crisis in the US, but with the military drawn from all parts of society and very heavily funded, the chances of successful separation was slim. Joining the military was one of the few ways to find actual work.

The constitutional crisis was effectively ended by the Terror@Home attack on New Year's Day, 2027. Those millions of out-of-work citizens were suddenly high-value physical labour resources who could contribute to the rebuilding of the nation. The class- and race- mixing that resulted from millions of wealthy refugees renting living space from the poor transformed US culture to be much more integrated internally - and even more xenophobic.

In the EU, UCANZ, and New Commonwealth I suspect a Universal Guaranteed Income would be implemented, with the basic of taxation shifting from a basis in labour (income tax) to a basis in assets and wealth. Entirely new classes of work would spring up (centred around humans interacting with humans, I think).

All the characters would have personal experiences shaped by these forces.
  1. Biggs and CAFIA were tell established examples.

Pages in category "AI"

The following 10 pages are in this category, out of 10 total.