Bombshell Genius
Image Source: In Auto News
Intent: To provide an advanced AI
Development Thread: Finally, some non-Jawa womanizer! (posts 11-30)
The Hammurabi Code (posts 1-10)
Manufacturer: Century Technological Advancement Enterprises, jointly with Ringovinda Systems
Model: Hammurabi Artificial Intelligence
Affiliation: Private market
Modularity: No
Production: Limited
Material: Artificial intelligence:
2,163-word thesis spread over the dev threads as follows:
Part 1 (1,579 words in posts #11-19, 21, 23, 26, 28, 30)
Part 2 (584 words in posts #1-3, 5, 7, 9-10)
Strengths:
- Provides the functionality of advanced droid brains (gives a device equipped with it the autonomy and feel of a semi-sentient being)
- If installed on a ship, allows to reduce the crew requirements of the ship
- Low-maintenance
- Can make updates for a device's databank autonomously
Weaknesses:
- Expensive
- Learning curve both for the program itself and for its operators
- Must be modified at its core to be utilized in weapons systems
- Requires a large quantity of RAM to properly work
- Difficult to reprogram emergent personalities without resorting to a complete memory wipe
Description: Created over the course of a week-long trip, including a date, in which Jessica Med-Beq drew in her hidden computing talent alongside Warren Century, Hammurabi features the latest in sub-symbolic and statistical AI algorithms, as well as machine learning, fuzzy logic, and other tools from AI to give the feel of limited sentience, while at the same time giving it some form of common sense. Periodic partial memory wipes, based on how often a cluster in a memory bank is used and how old it is, allows an object equipped with it to "forget" stuff similar to semi-sentient (and sentient) beings without the need for a complete memory wipe. As with most advanced AIs in its class, especially those capable of modeling emotions, Hammurabi has a tendency to create personalities even after limited use of a device equipped with it. Built-in programming containing the Four Laws of Droids in its operating system allows it to prevent a device carrying it from harming the owner of said device, or its creators.
Hammurabi is programmed to provide the ability of independent operation of droids or power armors while ensuring the safety of its users. As used on ships, such as small craft or some models of light corvettes (ships up to 100m), it allows the ship to pilot itself while its occupants can sit back and relax. Unfortunately, due to all four of the Laws of Droids being in force, an object equipped with Hammurabi cannot be made to fight sentient targets or vehicles that contain sentients onboard. Yet, in all cases, it will also be allowed to have its programming evolve and adapt over time, so long as the new lines of code do not interfere with the Four Laws of Droids. For example, while Hammurabi can use a ship's weapons to blast an asteroid in its flight path, or to engage droid starfighters, it cannot engage ships with sentients onboard: on a ship equipped with Hammurabi, sentient crewmembers must operate the weapons in order engage ships with sentients onboard.
As a result of its complex programming, Hammurabi is equipped with "common-sense" programming, which also allows it to "learn" in a similar fashion to a sentient being, and to act according to the best course of action it has decided based on the available data. Therefore it allows:
- Nearly-autonomous operation of a device (ship or power armor) equipped with Hammurabi, as if the device became a droid; of course it can be used in droids
- It allows the device to evolve in its functions over time, thanks to machine learning and common-sense programming
- The program has a learning curve both for itself and for its operators
- The owner must be defined when uploaded
- The device cannot automatically operate weapons for combat while Hammurabi is active
- It can operate ships up to 100 meters in its largest dimension
- It can operate the engines (sublight and hyperdrive), the reactor, the life support systems (thus allowing some measure of energy and crew savings), the mainframe, the sensor systems, the navicomputer, the communications systems as well as the shields; in an emergency it can override the weapons and, along with them, the targeting computer, when navigational hazards are best dealt with by blasting them with weapons fire
- It is uploaded in a ship's mainframe after the ship has taken off
- Since it is embedded in the mainframe, it requires a source of power to be active in order to operate
- It does not negate the need for organic gunners or automated weapons operation programming in space combat as a result of the Four Laws of Droids
- It can sometimes bring a critical system to a complete stop if said system has been compromised in-flight
- The wearer of the power armor can obtain an assessment of threat or environmental hazards, sometimes even taking the wearer out of the way if the power armor has independent propulsion systems
- Hammurabi can be uploaded in a power armor with an UI
- It cannot operate the weapons for the wearer
- It cannot be used in armors without an UI
- In power armors without propulsion systems, it will operate mostly as an advisory system
- In power armors, it tends to think on an operational level, rather than on a strategic level: if the best course of action is to flee, it will cause the wearer to flee
Primary Source: (Wikipedia both)
Artificial intelligence
Three Laws of Robotics (the inspiration for the Four Laws of Droids as stated in the thesis)
