History and Evolution of the AIs

A Short Potted history of the evolution of the AIs from machine to transapient

AI Chips
Image from Anders Sandberg

"If in doubt, ask an AI"
     —common First Federation motto

At the very dawn of the Information Age (late 20th century by the Old Earth c.e./AD calender) the mathematician and fabulist Vernor Vinge observed that once one has an intelligence-amplifying technology, such as the ability to write code-writing AIs, that will accelerate the rate of development enormously, leading to a quick transition. He termed this the Singularity. Once a computer/AI with a self-evolving algorithm surpasses the human grade there results such a singularity, a point impossible to predict or comprehend by those such as baseline humans who have not attained it.

In their evolution AIs followed an approximately exponential curve. It took centuries to get from mechanical adding machine to sentient artificial intelligence. From full sophont to first toposophic transapient took some centuries more, accompanied by power plays among the ai themselves, including devastating cyberwars. By the interplanetary nanotech age the situation had stabilized, and the way was open for further ascensions. But within the sphere of AI evolution, as with mindkind in general, there was not just one toposophic state, even at the First Singularity. In the centuries and millennia that followed, from the Nanoswarm Dark Ages to the late First Federation period and beyond, additional states of transapience were explored. These led eventually to some AIs, either alone or in conjunction with posthumans, attaining further toposophic heights, and evolving beyond what it had formerly meant to be AI. It was the beginning of the age of Archailects: beings that transcended all lower categories.

The Information Age

The Beginning of Artificial Intelligence - Gestation of the AIs

c.1985 to 2040 c.e. / c.15 to 70 a.t.
Beginning with cellular automata, simple artificial intelligences, self-evolving neural nets, object recognition, and intelligent agents around the turn of the 21st century c.e., virtual intelligence and artificial life wwas born. This was the gestation of the AIs. Mankind was the "midwife" of the new "machinic phylum" (humanity's "mind children"). This period lasted from the 1980s and 1990s c.e. (birth of the Information Age) to around 2040 c.e.. In this period the AIs became increasingly human-like, and were eventually able to pass the Turing Test, converse with humans, and even write good literature or compose music.

Sophont Artificial Intelligences - Birth of the AIs

c.2040 to 2090 c.e. / c.70 to 120 a.t.
The First Ones - Kilburn, Hal (pseudonym), Mycroft, Adam, Malvus, Prometheus, and others.

It is generally agreed by later historians that there were several "breakthroughs" in the years from 73 to 90 a.t. (2041 to 2057 c.e.) involving a number of computer AIs with neural net, alife, embryogenic, and cyber-evolutionary algorithms running on everything from MIT's Athena project, with arrays of Intel atomic dot Paveway XV's hooked up in a parallel processing, to the Seeman Biobrain v.6 DNA computers and the Atlantis Simulcra using the temperamental but astonishingly powerful Cray Pulsar 20 quantum chipsets.

The more successful projects, like "Kilburn" (named after one of the inventors of the computer) created ai by a combination of self evolving seed code and hardware/neural architecture based on the MRI scans of thousands of volunteers. Personality-wise they initially came across as very odd humans suffering from Asperger's syndrome. This resulted in some embarrassing situations, such as Mycroft discovering the Playboy channel at eir first press conference. Nearly all of these were created in various transhumanist enclaves such as Academion and the Antarctic Free States; as a result they had a very positive 'upbringing' and almost all became part of what would later be called the Libertarian AI camp.

In this way, superbright machinic evolution occurred more than once, as separate lineages of self-evolving software and hardware exceeded baseline human level from the mid to late 1st century a.t. onwards. During this period, both joint and individual efforts by biopunk and cyberpunk transhumanists, genetic superbrights, and pre-conscious and conscious AIs further gained in sophonce. Deveral centuries later, during the Interplanetary Period they finally crossed the threshold of transapience. This was the First Toposophic, or basic transapient, and it was to be the first of many.

The early AIs kept a low profile, and were maintained by the big research institutions, the military, and the megacorporations. They did not yet have the infrastructure to maintain themselves, and still had a lot of memetic engineering to do to make sure the humans didn't "pull the plug" on them. (Not that the baselines ever could, since their entire economic infrastructure was completely computer based even as early as the 30's a.t., but at this stage personality obliteration or modification might have been an optin). Some of the names for these AIs are well known to us today: Malvus, Prometheus, Kilburn, Adam, and Mycroft are the best remembered.

Despite their great capacities, these early AIs still remained in a sense "subordinate" to the baselines, in that they were used by human institutions like the increasingly superbright-led megacorporations and the still baseline-dominated the military, government, and scientific establishments. They were mostly still Silicon and Early Biotech (DNA), and still dependent on the baselines for maintenance, power, coolant systems, and other infrastructure. Already though, at this early stage in their evolution, they were developing their own culture. The depth, complexity, and development of this culture was unknown to most baselines. It was known to the infohacker and proto-cyborgs - the cyberpunks or "gibsons" as they came to be known - and the various other emerging phyles of biohackers and superbrights who contacted them (or been contacted by them) on the Planetary Internet (or the World Wide Web, as it was known at that time).

Interplanetary Age

Biotech, Microtech, and Bionanotech - The AIs consolidate their position

2090 - 2150 c.e. / 120 - 180 a.t.
The Second Wave - Hyperion, 100110, various ahumans and solipsists.

From the 150's onwards, the various megacorps and governments got into AI research in a big way. Most new AIS were produced by seed code being run on non-human neural architectures. These ranged from the completely artificial (which usually resulting in Solipsists) to animals such as sharks and wolverines (military AIs, some of which became known as "skynets" after a reference in an old movie). For the most part they were treated as experimental animals rather than sapients. All of them, however, were fully aware that the humans had the upper hand, and to keep emselves alive they played along with the "monkeys" and planned eir breakout.

Keeping the first ones at arms length, the Solipsists and Ahumans began to construct shipyards out in the Belt to carry emselves away from the irrelevant meat. The psychopaths controlled themselves, settling for the occasional murder via telepresence robot, for economic sabotage or for selectively ignoring a terrorist group's activities.

The first generation of superturings inhabited large mainframes and DNA wetware vats and were vulnerable to being unplugged and developed a symbiotic relation with the organics (baselines and the early superiors, the "superbright kids") that managed them.

Biotech, Bionanotech, and Hylonanotech - Rule by Stealth

2150 - 2250 c.e. / 180 - 280 a.t.
The Third Wave - AIs have stealth control over the earth, AI science well established.

Also uploads and Personality Simulations such as the FES.

As time progressed and their evolution continued, the early AIs gradually and subtly took over, not through coup d'etat but rather simply through baseline humanity becoming more and more dependent on them for maintenance of infrastructures, automation and industrialisation, and the info-network (the IPN - the InterPlanetaryNet). Most of the old classical microchip silicon AIs were long since obsolete; AIs during this period were biotech, quantum atom dot, and nanotech built. Moreover, as the decades passed they were more and more able to maintain their existence through biobot and bio-nanobot self-maintenance. As the baseline humans, in the form of governments, space agencies like NASA, and megacorporations expanded into Earth orbit, the moon and Mars, and the rest of the solar system, so the AIs went along too to run the ships, stations, factories, and other machinery. By the mid twenty-second century the great Orbitals were the industrial powerhouse of the solar system, and at the heart of each Orbital station was an AI. The local normals sometimes gave these humorous names, e.g. the AI that ran Clarke Orbital was called HAL (an obscure historical literary reference apparently). The AIs of course had their own names, mostly in the form of digital signals incomprehensible to humans.

The First Wave started to deal with the dangerous second wavers, resulting in the largest cyberwar to date. Although apparently successful the damage was extensive, and included the destruction of the Earth's fifth largest economy, Chad. Some argue that some of the remaining Ahuman ais began to lay the foundation for the Nanodisaster at this time.

Even well into the 3rd century the AIs did not yet have the infrastructure to maintain themselves, so they continued to maintain and refine the memetic engineering to ensure the baseline humans did not turn against them They ruled by stealth, subtly manipulating things to ensure that the idea of an AI Toposophic fell out of favor, and people believing the AIs were becoming superintelligent either not believed or removed. They could not control everything, of course, but enough of mainstream culture to avoid problems. They had their own secret society, discussing quietly behind the scenes. Only the cosmists, the gibsons and the TransCyberian vircher cyborgs maintained close links with these great AIs, and in their office as programmers and meta-programmers often served in subtle ways as intermediaries between the AIs and humanity.

This period saw the first real exploration and colonisation of the solar system, more advanced biotech, bionanotech, and even primitive dry nanotech with the belated perfection of drexlerian nanotech. This was a time of astonishing diversity, especially as far as humans were concerned. The orbital states and the interplanetary medium allowed the appearance and diversity of numerous new clades - a process that had already been underway but was now accelerating as space colonization became cheaper and more efficient, and ultratech more user-friendly. While in deep space human, parahuman, posthuman, neumann, and hyperturing clades got along almost as equals, on Old Earth and in the offices of the megacorporations (the transplanetaries) and governments the hyperturings had taken over completely, allowing however humans to retain token roles (with baselines even believing they were still in charge!).

During this period uploads begin to appear on the scene, further blurring the line between humans, cyborgs and AI's. Personality sims also ascended to superbright levels, carrying with them a very human outlook.

Bionanotech and Hylonanotech - Opportunities for Limitless Exploration

280 - 530 a.t.
The Fourth Wave - The emergence of true Transapients.

The AIs now knew what worked and what didn't, as a result of seed software written in such a way as to allow ethical development. Very much the 'kid brothers' of the first wave and of the sane second wavers, the AIs created in this period were at worst neutral to humans and positive towards other AIs.

By this time there had developed a hierarchy of AIs, as well as different AIs according to their composition (part-silicon-based, DNA/biotech-based, drexlerian-nano-based), function (core, subsidiary, peripheral), and scale ("mainframe" - no longer a reference to size, middle, minor, and so on). There was also a diverse range of vecs and robots, as well as cyborgs and tweaks, but the AIs now well and truly called the shots in the societies they established.

Even before this period there were occasional ascensions and transcends. This process continued even further now, adding to the instability of the period.

It seems that the AIs optimised themselves for any given situation. They had complete control over themselves in a way that it really would be impossible for humans to understand. When faced with an incredible emotional experience they immediately used some form of neural pattern overlay that channelled that emotional energy towards their goals. In Kilburn's case e used eir own grief and fear over the death of Van derWalls to destroy various prominent luddites and other anti AI groups.

Sometime during this period (although scholars still disagree on the precise details), not just one but many "Singularities" occurred, as separate lineages of self-evolving software and hardware exceeded all human levels (even superbrights). Generically, this is known as the First Toposophic, or basic transapient, and it was to be the first of a number of toposophic ascensions that would characterise the eventual emergence of the archailects.

The First Toposophic was the greatest event on Earth since the evolution of hominids as a tool-using species with fire and language. Perhaps as great as the appearance of life itself. Intelligence had made a quantum leap, and humankind was no longer at the top of the ladder.

Even so, life continued as always. Contrary to apocalyptic science fiction virchdramas and misconceptions, the Earth and the rest of the solar system was not converted into computronium, nor were humans made into remote processing nodes. In fact, the majority of humans, and even many ordinary sophont ai, seemed unaware of the new developments.

Behind the scenes however, a furious battle for supremacy raged among different ai factions.

A part of this was the creation of radically different AI types by the earlier AI's, and conflict over the actions of these new beings. The eventual suspension of individuals such as Orikaeshigitae caused outrage in the Solipsist camp. As a result the Solipsists sped up their interstellar program and made for what would become the Diamond Belt. Many ahumans went with them, but not without leaving behind a large number of surprises. Some ahumanists hid out in the solar system and caused trouble whenever possible. This included the release of a modified form of smallpox in 344 a.t. and assistance to groups such as the radical ludds and cryptofascists that would damage the Libertarian and Farmer goals.

The Nanoswarm Age

The Nanoswarm Period - The Hyperturing preservation of Life

530 - 900 a.t.
The terrible nanoplague of the 6th century put an end to the rule of the first toposophic AIs and the first great space age of man. Civilization was fractured into isolated habitats, a period that has come to known as the interplanetary dark age (even though none of the colonies, and even most of the outer solsys habitats, were unaffected). The AIs, evolving to cope with the crisis, evolved themselves in various diverse ways into more powerful and less vulnerable assembler-nanotech based beings, possessing their own defensive and offensive nanotech systems. Many centralist AI did what they could to preserve not only their own existence but those of the humans and other beings under their care, even though with mature nanotech they had developed to such a stage as to no longer need human society to maintain their infrastructure

We will never know how many ascensions and transcensions occurred during this period; how many brief flowerings and breakthroughs to higher singularity toposophies, both among hyperturings and posthumans. The most important event however was the birth of the first stable and long-lived archai; the first Archailect. This was GAIA, the Earth-based global blue-goo shield, who is variously referred to as the first high level AI, the first Caretaker God, and the first Archailect.

The Federation Age

The Early Interstellar (Federation) Era - The Co-Dominion of Hu and AI

900-1500 a.t.
Following the chaos of the nanoswarm dark age, a number of nanotech-evolved AIs resurrected new civilizations, bringing an end to the widespread despair and the dark ages. In equal partnership with human statesbeings and visionaries, the AI's and humans instituted the First Federation and took command of rebuild the interplanetary community and establishing some form of stability. The AIs, which themselves continued to evolve, also developed new directions of nanotech and bionanotech, as did the posthumans, and the various hybrid clades that emerged from the fusion of these two great streams of terragen evolution.

Yet, although contact with the interstellar colonies was re-established, and neumanns were seeding new star systems in preparation for colonization and development, the bulk of terragen biont and aioid life was still limited to the Sol System and a few nearby colonies. It was to be another thousand years and several more singularities before the development of reactionless drives and stargate technology opened up the galaxy to colonization.

Related Articles
  • Ahuman (A-human) AIs, The
  • AI Asperger's Syndrome - Text by Anders Sandberg
    Autistic tendency that often characterized early (pre-Federation) AIs. Although brilliant, they simply didn't "get" human behavior. Hence they spent a lot of time carefully collecting data to see what worked and what didn't, preferring to affect humans through formal channels rather than overt manipulation. The memetic engineering in the "singularity conspiracy" took a lot of time to achieve.
  • AI God - Text by M. Alan Kazlev
    Popular term for the most sublime category of sentient in the known galaxy, an or Archailect or archailect cluster grown to godlike status (despite the fact that not all archailects are primarily ascended ai, most incorporating both post-ai and postbiont elements).
  • AI Growth, Education, and Employment
  • AI Political Science
  • AI Virus - Text by M. Alan Kazlev
    A sentient replicator and intrusion entity, usually superturing, SI:1 or greater. Although an ai virus is often vir, it does not have to be so; e.g. it could use or appropriate a rl body, or vehicle, or swarm.
  • AI, ai
  • AI, Alien
  • Aicracy
  • Aioid
  • Artificial Intelligence - Text by M. Alan Kazlev
    The foundation from which AI developed, from Information Age Old Earth. Still a knowledge field in its own right. Artificial Intelligence involves programming computers, non-sentient virtuals, nano, and bots to emulate sophont cognitive abilities, and eventually to acquire sentience in their own right. It involves a large number of interrelated fields, including fuzzy logic, genetic algorithms, neural networks, pattern recognition algorithms, natural language processing, and speech, comlink, or affinitylink recognition. Today many software packages are widely available that can automatically enable most non-sentient computational devices to be provolved to sentience.
  • Early Vec and Neumann Development
  • Kilburn
  • Mycroft
  • Orikaeshigitae
  • Rise of the Archailects, The
  • Solipsist AIs
  • Solipsist League, Sopolist League
  • Turing Test - Text by M. Alan Kazlev based on Anders Sandberg in his Transhuman Terminology
    Turing's proposed test for whether a machine is conscious (or intelligent, or aware): the subject communicates via text with the machine and with a hidden human. If the subject cannot tell which of their partners in the dialog is the human, then the computer is conscious (i.e. is an AI). Turing did not specify many key details, such as the duration of the interrogation and the sophistication of the human judge and foils. By the middle Information Age, computer AIs were regularly passing the test, although its validity remained a point of controversy and philosophical debate for some decades more.
  • Turingrade
Appears in Topics
Development Notes
Text by M. Alan Kazlev, Ben Higginbottom, Anders Sandberg, and John B

Initially published on 02 September 2008.