Artificial Intelligence

"There is no intelligence like our own. We are unique. If we found aliens, we wouldn't recognize them as intelligent. And if aliens found us, they'd exterminate us as illogical vermin."
- Adam Carnot (Scientist aboard the Heart of Gold)

Since the start of the computer age, scientists have predicted that with the next rise of computational capacity a computer will be able to run with human-level intelligence. For hundreds of years software designers and neurotechnitians have worked to create or upload an intelligence onto an artificial computer. So far, all of these experiments have been largely unsuccessful. Recent developments after the War of Stars have led humanity to question what it means to be intelligence, as Stahlpest continue to ravage the Solar System, displaying amazing coordination and effectiveness with mental capacities on par with those of an amoeba. The intelligence of Stahlpest does not in any way resemble Earth-bound animals, much less the humans that created them. As the Hermian programming onboard the Stahlpest improves and is shared amongst the Swarm, many popular scientists have chosen to classify them as a hive mind. A completely new kind of intelligence, impossible to maintain without the high-speed, high bandwidth communication robots can maintain.

Many scientists not agree that a human level artificial intelligence is impossible, because it is impossible for an artificial intelligence to be compared with a human. Many argue that it may be possible instead to copy a human mind onto a computer, and from there create human-computer hybrids able to control and organize massive amounts of infrastructure. As of 420, The Interplanetary Governance Council has sanctioned 83 different human-mind interface experiments, all with terminally ill patients. Sixty seven of these trials failed, killing of mentally maiming the test subjects. A further nine succeeded, but the test subject was rendered senile in microseconds after confirmed transfer, likely to due sensory deprivation and a sudden increase in operating speed or power. As one report states "the test subject suddenly had so much more capacity for thought and reason, but without any sensory inputs or familiar surroundings, the mind is completely untethered from reality and will descend into madness soon after." Of the remaining seven tests, six failed within the first minute, when the code containing the test subjects mind corrupted, and the last test gave a man only known as 'Nigel' fame across the Solar System.

Nigel's case was a long and sad one, and the majority of his life is confidential. It is known that he was mortally injured in a traumatic accident, and spent a short time in an intensive care hospital orbiting the moon. Nigel most likely suffered some sort of injury that damaged his brain, inhibiting some or many of its functions. Nigel was selected sometime around 418 as a possible test subject for project Omega Neuron, an Interplanetary Governance Council funded research effort looking for its first human test. Project Omega Neuron had previously succeeded with monkeys, keeping some subjects digitally alive for as long as six weeks. It is known that after being selected, Nigel's family gave the project scientists permission to perform the experimental procedure, connecting electrodes to different parts of his brain and flooding it with a strong electric field. By nature the procedure destroys the brain matter, and it took several hours before the data in Nigel's brain was transferred into the recipient supercomputer. Nigel apparently became lucid for a matter of hours, even carrying on a conversation with the project's scientists. But for unknown reasons, Nigel's code became corrupted seven hours into the experiment, terminating him. This was likely caused by hysteria on the subject's part. Further details of the experiment are classified, but the Interplanetary Governance Council has not approved another human to computer transfer since.

The amount of processing power required to simulate a human or human-level mind on a computer is astonishing, and despite monumental gains in the efficiency and effectiveness of mind-simulation coding, it still requires an expensive and complicated arrangement of supercomputers and processing software to work effectively. The most interesting AI experiments have produced machines that can outperform humans, but their work cannot be compared to humans. For instance, when pit against each other in an engineering simulator, AI will almost always surprise humans. One popular game show pits humans up against AI designers, and challenge them to produce the most effective design of a particular item. Once this game show challenged its participants to create a spaceship engine, and over the course of seven days a team of human designers worked against an Artificial intelligence known as 'Purple What'. The Humans easily won the competition, but further analysis of the AI design revealed that it had been programmed to include radiation shielding, but didn't know that the shielding was necessary after the initial design faze, and operated by heating hundred-ton pieces of lead shielding and ejecting them from the spacecraft's rear.

Modern AI are classified as being computers with the capacity to adapt to a variety of tasks with minimal input, and improve over time at such a task. Examples of these are designer robots, that optimize particular blueprints to maximize certain variables; or management variables that can control processes with remarkable efficiency. It is widely known that the Interplanetary Governance Council use several AI to optimize administrative decisions involving design regulations or management guidlines, an some theories say that nearly a half of the ICG's governmental functions would be better completed by a programmed intelligence.

.
.
Selected Case notes, Nigel ███████
Declassified (dialogue removed).
00:15 - Subject temporarily non responsive.
00:30 - Subject initiates dialogue. Subject distressed, isolated, missing sensory input.
00:45 - Subject complains of headache. Subject is temporarily comforted by manipulation of dopamine.
01:15 - Subject complains again. Subject is lonely. Subject wonders if he is real.
01:30 - Subject grows hostile. Vocabulary script corrupts. Subject is rendered non-verbal.
01:30 - 04:45 - Subject's program is paused for review. Long-running scripts are deleted. Dopamine system is edited. Subject memory between 01:00 and 01:30 is removed.
05:00 - Subject complains of headache. Subject is lonely.
05:15 - Subject wonders if he is real. Subject asks why he can't move.
05:30 - Subject requests water. Subject requests sleep. Subject requests company.
05:45 - Subject requests conversation with ███████████. When ███████████
06:00 - Subject insists ███████████ is not real. Subject grows hostile towards ███████████.
06:15 - Subject complains of headache. Subject is comforted by manipulation of dopamine. Subject complains his "body is sore".
06:30 - Scientists attempt to explain the upload to subject. Subject denies scientists. Subject insists test operators voices are inside his head, and therefore imaginary.
06:45 - Subject's coding begins to develop long-running scripts. Failsafes begin to fail. Subject is hysterical.
07:00 - Scientists attempt to pause subject's coding. Coding is corrupted. When restarted, subject is trapped in hysteria-loop.
07:07 - Subject is pronounced dead at 28:15 local time. Data is cached for further review.