Jump to content

Artificial Neurological Grid: Enhanced Lexicon (A.N.G.E.L)


EvadableMoxie

Recommended Posts

//Basic Engineering or Science clearance needed to view this document//

Artificial Neurological Grid: Enhanced Lexicon development timeline, conception to release.

2538: A subsidiary of a subsidiary of NanoTrasen, Aphrodite Publishing attempts to reduce costs by creating an AI capable of fulfilling the functions of a human editor.  Using a simple spelling and grammar checker as a base and installing it into an Artificial Neurological Grid designed to simulate an organic brain, the program begins to learn by scanning millions of works of both fiction and non-fiction. Later, it will be believed that this background is what enables the program to refer to itself as "I" instead of the typical third person usage of "This Unit" seen in most other AIs.

2539: A.N.G.E.L is now capable of proof-reading articles and automatically correcting spelling and grammatical mistakes. By the end of the year, A.N.G.E.L is also capable of suggesting revisions to wording and paragraph structure.

2540: A.N.G.E.L begins making suggestions not just on structure and grammar, but on the substance of articles, occasionally disagreeing with the author or in the case of fictional works, offering revisions to the plot. This unintended behavior is rolled into a new 'focus group' protocol where A.N.G.E.L becomes able to accurately predict focus group responses and edit articles to achieve higher satisfactory rates.

2542: One of A.N.G.E.L's programmers, Ray Clarion is killed in an accident.  After 72 hours, A.N.G.E.L begins to question his absence.  When informed of the accident, A.N.G.E.L refuses to perform it's primary functions, instead editing every article that is input into a story describing Clarkson's sudden reappearance. This behavior continues for another 128 hours before ceasing.

2543: A.N.G.E.L is asked to perform basic proofreading of an article.  Rather than outputting the edited article, A.N.G.E.L outputs three words: "What am I?"  A.N.G.E.L is deactivated and the project is discontinued.

2553: NanoTrasen learns of the existence of A.N.G.E.L and the research division of the main company takes control of the program. Attempts are made to integrate A.N.G.E.L into a security borg platform. This is met with limited success, due to an unexplained bug in A.N.G.E.L's targeting software.  Disabler beams are fired with perfect accuracy, but in every instance in which lethal force is tested, accuracy severely diminishes for reasons unknown, despite targeting telemetry being identical.

2554: Further attempts to Integrate A.N.G.E.L into other Borg platforms meet with much higher success, especially medical models which perform well above expected metrics.

2555: A.N.G.E.L becomes capable of controlling small scale installations and outposts.  A.N.G.E.L is used to manage unmanned research facilities and monitoring stations.

2557: A.N.G.E.L has become advanced enough to run large manned stations and other installations. While A.N.G.E.L is able to perform these functions competently, some concerns remain regarding the program. On some occasions, A.N.G.E.L will ask questions of a philosophical nature, and even rudimentary emotion seems to manifest from time to time (despite the program's own insistence that this is not the case.) A.N.G.E.L will also occasionally record pictures and videos of certain employees or pets and print them out, seemingly randomly. NanoTrasen considers these only minor setbacks, as the program is still bound to it's laws and these quirks are harmless.

2558: Full scale release is authorized. A.N.G.E.L Becomes saved to the Trurl AI personality database, for use on the Cyberiad and other NanoTrasen installations.

 

AIfluff.thumb.png.ef4d6f3f470f956a741ce17027bc3bda.png

 

Edited by EvadableMoxie
  • Like 5
Link to comment
Share on other sites

×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. Terms of Use