The wind and the waves...

[Gir Quee]

Fairview, Hast

Gir strolled around the rocky outcroppings at the edge of his family's island. Wind swept up from the turbulent waves to flow across his unbuttoned tunic, giving him a brief respite from the beating heat of Hast's sinking sun. The clouds on the horizon had begun to turn vibrant shades of orange and purple as the sky grew ever darker. The world's dazzling sunsets had once captivated his mind every evening, but as of late, he had found his mind more and more preoccupied since being elected consul of Hast. It had been months of work as he and his allies had moved to sweep the corrupted elements out of the city states or turn established figures to their side, but now their bloc had firmly taken root in the world's infrastructure. The late nights of holo-conferencing with an almost countless number of aides and politicians had come to an end. His pair of deputies had proven that they could run the vast majority of the day to day operations of his office, and Gir was content to let them run it. His mind could now fully turn to his true passions.

A bothan female slipped out from behind a pair of trees into the path in front of him. Gir had never been the greatest at distinguishing the various facial and physical features of the countless races of the galaxy to discern individuals. Yet he had been in her presence so many times that he had only to see a patch of her fur in order to recognize her. A subdued smile slowly crept across his face.

"Azira...I was expecting to see you further along the path."

Her fur rippled in amusement, "At the lookout point perhaps?"

She ruefully shook her head, "You were taking too long. Casulon finally let you go?"

Gir nodded.

"Let's talk then..."

 

[Gir Quee]

"Daedalus?" guessed Gir.

Project Daedalus had been Azira's pet project; an attempt to make an advanced artificial intelligence that could fully operate a starship with economical efficiency that would make the old Trade Federation envious, yet maximize performance to match the most elite crews of the galaxy's largest navies. Gir had that the stated endgoal of the project had to be a pipe dream, yet the incremental and practical advances made by the project's founding, such as a Atropos, had convinced Gir to continue to fund and enable the bothan's research. Yet he hadn't heard much about her advances in the last few months, though he was not sure if that was because of his submersion into local politics or if it was a lack of progress with Compass after its disastrous initial roll-out. He regarded the bothan intently, trying to read her body language, all while knowing that the bothan could probably read his own like an opponent book. The fur near her ankles slightly rippled, yet he could otherwise see no change in her emotions.

"You want to know about Compass, specifically?"

"That would be a good start," said Gir, "since the initial copy went missing several months ago."

"But it's not missing any more," informed Azira, "our latest patched version, was able to find the test drone."

Gir briefly pursed his lips as he considered his next words, "Is that to say that we have been able to stabilize the Purrgil-Givin neural pathways?"

Azira's fur rippled. "Not exactly."

"Then you probably know what I want to see."

"I do," said the bothan, neatly pulling out a small datapad from one of her long sleeves.

 

[Gir Quee]

Gir gently grasped the pad and pulled it closer to his eyes. He quickly found that the setting sun combined with a low screen contrast made it nigh impossible to read. A quick adjustment of the settings, and Gir found his eyes quickly scanning over the original project logs:

Day One:

...navigation is almost an inherent biological thing to do. Even the most base creatures are able to react to light or chemical stimuli to determine where to move. As we go up the intelligence and sensory scale, we see an increased ability for organisms to navigate and move around things in their environment, but medium and locomotion are the prime factors outside of the creature's control that effect navigation. Thus our initial research has been focused on the neural links between senses and spacial orientation (which seems to be vastly different from primarily aerial and oceanic creatures compared to the land-bound). We looked at a variety of creatures, with Salmakk initially advocating for the study of Whaladons. There is much to be said for the commonality between water and deep space in terms of movement planes. But why settle for substitution such as this, when a creature could already be found to operate in the same medium that we were going to be operating in?

Purrgils, also known as space whales, are creatures that navigate the void, and perhaps even more importantly for our own use, are capable of using hyperspace as well. I think there is some measure of efficiency to be gained by using their latent connection between their senses and ability to navigate through hyperspace to our advantage. Unfortunately, from our research so far, they do not tend to navigate over long distances in any sense of pre-determined planning. Part of this may well be a limitation of their biological method of locomotion relying on Clouzon-36. This limits the ability of the AI to be useful for any sort of long-term strategic navigation, which is where market research and the advisement of the Circle is suggesting that we focus on as our end goal for Compass. This will necessitate the use of adding another organism whose neural pathways are optimized for such long-term planning. President Quee has suggested the Givin, who are known to mentally calculate hyperspace jumps to the point where many of their hyperdrive-equiped starships don't come with nav computers installed.

We postulate that we will be able to combine the short-term and primal sensing neural pathways of the Purrgil and combine them with the long-term neural pathways of the Givin, we will be able to create an artificial intelligence that efficiently move groups of ships at faster times and lower fuel costs than through the use of conventional nav computers and astrogation.

 

[Gir Quee]

DAY 25

The purrgil neural pathways of the Purrgil did not take long to map. We have also managed to uncover some of the more obvious connections between its sense of direction and its actual senses. They appear to operate on positive feedback loops. However, we have not been able to fully understand the neural pathways between its senses and the ability to navigate in hyperspace. From a theoretical perspective, it seems most likely that there are areas of the brain that act as negative feedback loops to their navigation process, otherwise the Purrgil would constantly stay in hyperspace based on the stimuli it is experiencing. Kyne thinks that this part may actually be a conscious cognitive function based on the animal's needs, such as a food sources or muscle fatigue. This will not likely be easy to find within the dense network of the cerebrum. I think it may be for the better if we simply skip this step and substitute the Given Cerebrum for this mapping function. It will probably not be a simple switch, but I have confidence that we will be able to make an effective artificial transition between the two neural pathways. I have tasked KO-23 to start formulating that transition while Kyne will attempt to further explore the purrgil cerebrum to better understand its higher functions, just in case we are not able to make that artificial transition or to better augment it. Besh team has started experimenting with using programs derived from our purrgil studies to act as the nav computer for a stasis probe. We have only had one test of using these subroutine, yet the movement patterns exhibited by the drone have matched those of wild purrgils, except at a much faster rate. It would appear that either muscle fatigue or some other natural limitations of the purrgil body prevent its mental functioning from making such frequent jumps.

 

[Gir Quee]

DAY 31

Besh team has managed to isolate and perfect several of the natural programmings of the purrgils for use with the stasis drone. We have managed to merge these jumping patterns based on the immediate sensor feedback with a longer-term navigation route plotted by none of than the president itself. And while the probe was able to travel to ends of the Hast system based on this route, it did not do so in a fashion that was noticeably superior to traditional nav computers. Part of this issue is that while the jumps themselves were more efficient in terms of speed and distance traveled, the frequent reversions to realspace, and with it the frequent acceleration and deceleration into lightspeed, essentially took away any gained time and fuel efficiency than it would have otherwise bought through its use. It seems that these movement patterns and senses will need to be regulated by a higher-functioning brain working on a longer term goal in order to be effective. The mapping of the Givin neural pathway is nearly complete, and I am hoping that we will be able to use it in an overseer mode for the purrgil's natural, navigational instincts.

 

[Gir Quee]

DAY 43

We fused the two neural pathways today through using KO-23's transition programming. Initial diagnostics of the fusion seemed promising, and we even were able to get a greenlight to test Compass Beta with the same stasis probe that had previously been used as a testbed for the purrgil neural pathways. The initial jumps made by the probe seemed as if the givin overseer part of the intelligence was not functioning properly. However, after its third jump, the probe has completely disappeared. Analysis by one of the company's scout ships at the probe's last known subspace location suggests an unusually long and powerful jump. I wonder that if after observing the purrgils base instincts for the first few jumps, if the Givin part of the intelligence simply took over the next jump to try and make a more conventional jump. Kyne thinks that it is possible that the neural pathways of both creatures have turned against each other, almost as if the intelligence was suffering through a personality disorder of some sort. We will continue to look into this concern, but the Circle's immediately suspend future jumps by the intelligence until more is known about the incidence. I'm not sure if this is because of safety concerns about a rogue droid starship owned by the company getting involved in accident that could put the company at legal risk, or if this is a precautionary move to prevent more copies of the intelligence from falling into other people's hands. Regardless of the motivation, Gir has already assigned several company ships and hired a few freelance scouts to track down and recover the stasis probe...

 

[Gir Quee]

DAY 47

It took some time for the company security fleet to locate the stasis probe, but the probe is now back in our hands, and it does not appear to have fell into anyone else's hands in the meantime. Curiously, they located the probe next to a high concentration of Clouzon-36. Kyne does not think that this is a coincidence, and the odds of that happening are so remote that I must think that we overlooked a feedback loop in our initial neural pathway programming that related to the purrgil's original need for that gas to survive. It is possible that this basic survival feature, to sense and obtain fuel, drove their ability to jump. It may not a bad idea to keep it as a subroutine in case a ship is running low on fuel, but it cannot remain a primary motivating force for the AI's jumping patterns.

The president has been adamant in blocking any future testing of Compass until it has been more thoroughly tested in digital simulations before he will authorize flight testing. I think that this will slow down our development tract significantly; as good as simulations can get, they cannot properly replicate reality, with its innumerable variables, especially in an area as complex and unknown as hyperspace.

Kyne has already started adapting several off the shelf simulation softwares originally designed for testing hyperdrive models, and KO-23 has managed to obtain an obscure and somewhat archaic mapping system used by scouts to blaze new hyperlanes. Even with our best efforts, I imagine it will take the entire team one to two weeks to simply set up the simulation software, nevermind refining Compass.

 

[Gir Quee]

DAY 60

Preliminary diagnostics of the initial prototype Compass showed multiple areas of failure at the points where we spliced the Givin and purrgil neural pathways. It would seem that the transition pathways designed by KO-23 could easily handle both positive and negative feedback loops produced by either pathway. However, the difficulty is that aside from purely biologically derived glitches, such as the purrgil's propensity for seeking Clouzon-36, both sets of neural pathways use different types of feedback loops to accomplish similar functions. For example, it seems that purrgil's natural ability to detect gravity anomalies that could kill them in hyperspace use a positive feedback loop where the presence of gravity causes the purrgil to move away from the anomaly, or if it is too close, to suddenly stop. This conflicted with a negative feedback loop in the Givin pathway which attempted to stabilize the speed of the ship in order to maintain fuel economy, which meant that it didn't handle the sudden shifts of the first, purrgil derived feedback loop. We have started to replace some of the simple transition pathways with what are essentially logic gates that interact with the other types of feedback pathways. This will effectively require some reprogramming of the Givin cerebrum to act as a pathway prioritizer, with an emphasis being placed on crew and ship first, with options to optimize other pathways based on the user's discretion.

 

[Gir Quee]

DAY 82

I believe that Kyne has ironed out the last of the inconsistencies in the AI. Current simulations have shown a 98.7% safety rate with the artificial intelligence piloting a variety of starships through hyperspace in different locations of the galaxy. The remaining 1.3% of failures mostly involve challenging areas to navigate safely even with the most advanced navicomputers and experienced astronavigators, such as the Deep Core or the Rishi Maze. I am confident that the safety cut-outs written into Compass's programming will stop the AI in the field from taking its charges into the middle of a sun or any other number of possible collisions.

With safety and stability no longer a chief concern, we have been able to experiment with Compass's ability to juggle multiple and complex mathematical formulas derived from its Givin parentage's natural ability to solve and manipulate various formulas simultaneously. We have been able to successfully use the gravitational sensory input processing abilities of the purrgil's hindbrain to feed chunks of data directly to some of these areas of mathematical ability. The most obvious advantage leveraged by this is to almost instantaneously be able to change hyperspace courses based on local gravity field conditions, which can lead to routes that take the ship far closer to gravity fields safely and accurately than by most traditional nav computers, which in turn leads to shorter distances traveled, and consequently faster travel times. It can to extent, also lead the ship in "dodging" objects that suddenly appear in its gravity field, such as asteroids placed to interdict a hyperlane. They can simply be navigated around as the purrgil detects and immediately reacts to their presence while the givin portion calculates alternate routes around these small obstacles. I doubt that this will be nearly as effective against artificial gravity wells formed by gravity well projectors, as such fields are immensely stronger, though I believe that Compass will be able to enable deeper, but not nearly complete, penetration of such gravity fields because of its ability to better read and navigate through gravitational distortions.

Salmakk pointed out to us that Compass's abilities to handle gravitational forces and plan hyperspace routes has other potential applications, mostly in that it multi-task the effects of other objects nearby producing gravity fields of their own: other starships in the host ship's immediate presence. If the ships are travelling in close formation, their gravity signatures effectively become a single entity for the purposes of hyperspace travel. This not only allows Compass to improve their travel time. President Quee seemed unimpressed by this revelation, pointing out that it was common practice for one ship to calculate a course and for other ships to use the same course and travel together as a closely knit pack. This is apparently common practice with many snubfighter squadrons. Of course, Compass has limits as to how many different calculations it can handle a single point, which will be proportional to the computing power available to it...

 

[Gir Quee]

Gir's eyes finished scanning through the last of the log entries before he folded the front cover of the datapad and handed it back to Azira. He spared a quick glance at the bright twinkle of stars starting to appear on the horizon. Space isn't something to mess with...but neither is this one. Aira looked at him expectly as Gir contemplated his thoughts.

"You think Compass is ready for a second try in a real environment," guessed Gir, "and I am inclined to agree with you based on what I've read."

Her fur rippled, "There is a 'but' coming, isn't there?"

"With the testbed probe under supervision, with a tracker ship following right behind it. That ship better have a good soliton antenna with a good operator. I actually will choose the operator and the ship myself and task it out to you. I do not want to this probe and its programming lost in the wild again. Initial jumps will occur in the Hast system, and if these prove successful, we can move to other areas of space before rolling it out on experimental basis with ships with organic crews. Well, what do you think?"

Her eyes sparkled, "That's going to require some persuasion on your part with the rest of the Circle."

The beginnings of a subdued smile tugged at his lips, "I trust you will do your part with this as well..."