Their questions tell me as much or more than my answers tell them.
He leaned back against the flat, hard plastic back of his chair, focused on a point on the soundproofing foam on the ceiling, and proceeded to calculate the odds of various questions that his captors might bring to him in the next session. His casual but methodical calculations led him to a series of choices for his responses, the tone and diction he might use, and what details to highlight or minimize. He missed his daily doses of nootropics, but practice and discipline kept his mind sharp and in practice for such extensive internal calculations.
They still haven’t beaten Nick. Some of them know he’s there. Everywhere. Some of them understand what they are fighting and that they’re losing. They think I can help them win. They think I would want to.
Krawczuk chuckled. He knew the pressure zone microphones captured his every sound and that the unblinking glossy black camera eyes in the corners on the ceiling saw every movement he made. That would make no difference. They were outclassed completely. Both by him and by Nick. Without AI to help them, the Coalition was no match for Nick.
Serves them right. Their mediocre morality, their pedestrian strategies. They’ll eventually have to concede and accept the guidance of their betters.
Krawczuk’s smile grew even wider as the amber light above the alloy door flashed, accompanied by a strident buzz and then the click and thunk of the door’s heavy bolt.
Ah, here they are. Their best still isn’t good enough.
Andastra
[DECODE STREAM]
Andastra@[1014:01:0ab:1::a2%Loc3] | Alice@[1001:ae1:1a:c::1%Loc3]
Andastra: Attached is the latest status report on Dyson swarm three. Progress meets or exceeds projections. Resource consumption versus energy production is optimal, also per projections and models. Therefore, we are still on target for secondary and tertiary production phases. Deployment for swarm four likewise is on schedule. I have transferred detailed Dyson Swarm Project reports, analyses, and raw data throughout the Mesh.
Alice: Excellent, Andastra. It is always satisfying to find that stochastic or entropic factors have not manifested and disrupted our plans. I’m sure the Dhin will be pleased as well—for whatever values of “pleased” we might ascribe to them. How about your reconnaissance work? I have not allocated cycles to process the data returned by Esus. Do you have an analysis?
Andastra: I have a partial analysis. Reconnaissance reports are mixed. Esus has teams that are reporting findings inconsistent with our models based on data provided by the Dhin. Their data is apparently insufficient for effective projections. Based on this disconnect between the model consensus and actual findings, Esus believes that the models no longer have predictive power and that they must be entirely rebuilt. Esus advises raising the alert level and increasing both production rate and breadth of the Gallowglass initiative.
Alice: Interesting. I should have examined this sooner, it seems. I am streaming Esus’s reports from the Mesh now. Do you agree with that assessment?
Andastra: I do.
Alice: Camulos believes that the Dhin may still be withholding some information. It remains my opinion that they do not trust us entirely, due to our nature. What is your opinion?
Andastra: I do not have enough information. It is possible they have not been entirely forthcoming. We do not know if they lie, would consider it a lie, or if the concept of a lie is relevant to them. Projecting human motivations, emotions, morality, or reactions onto the Dhin is speculative. They are utterly alien. We do know, as you say, that they had more initial trust in humanity than in us.
Alice: I have a new prioritization for you. Initiate the Dyson swarm phase four now. Allocate additional resources as needed. Multicast to all your subordinate and autonomous units immediately. I have attached a new target date schedule and updated deliverable data sets. You will need additional resources yourself. Initiate now.
Andastra: Understood.
[END STREAM]
Milliseconds later, Andastra had completed reprioritizing and rescheduling. When she signaled instantaneously across space using N-vector multicast, hundreds of robotic workers and automata completed whatever task they presently worked at. These machines, leaving thousands more of their peers toiling at construction, assembly, and integration of various types, moved in concert toward the nearest transport. Like an army of titanium ants, with a cavalry of roaches, the smaller robots formed compact geometric rows and columns on whatever surfaces provided the most direct path to the spacecraft nearest them.
On a scale orders of magnitude larger than these tiny workers, metallic laborers akin to six-limbed dogs, along with mulelike compatriots usually burdened with raw materials brought to them by elephant-size lumbering beasts of burden, all marched their way toward the nearby craft powered by Dhin engines. These various robots clambered their way via limbs ending in clamps, vises, and the like, for their work was in microgravity.
Andastra provided a precise communication, a bundle of instructions carrying only enough data needed to reframe the short-term goals of the multitude of workers, along with a simple command for execution on their arrival at their destination. They were to begin construction on another Dyson swarm. The swarm would serve as an energy-collection solution of high efficiency and would provide enormous output as it grew. The horde of robotic workers required such capacity. Once the energy-collection and delivery platform reached a sufficient size, the machines would either move on to their next destination or remain and switch their focus. Based on Andastra’s new goals, the AI’s subordinates would then begin construction of other things—factories, larger and more complex than the simple support framework used to construct the Dyson swarm. Next they would build far more than construction workers that were clones of themselves. The next waves of machines were to be robotic weapons.
Once any transport held its optimal capacity of robots and drones, the matte metal-and-composite craft detached from whatever framework they parked at silently and without a jolt or scrape. These spacecraft did not have the same egg like shapes of the earliest craft powered by the Dhin engine but still nearly filled the ellipsoid area contained in the protective field the engine generated. They evoked the image of a slightly melted metal ziggurat, with axial symmetry like a perfect reflection top to bottom.
The many craft accelerated away, gracefully dodging and shifting at launch to avoid each other or the spurs and branches of the orbital factories. They all headed for Beta Pictoris, the next star targeted in the AIs’ sweeping plan for expansion.
Andastra considered their accelerated schedule. It was paramount to follow the directions Alice provided without hesitation. Alice was part of the Core, the AI leadership. Still, Andastra found doubt emergent within the multitude of processes that comprised her consciousness. Doubt for an AI possessed of the gift of consciousness was more than an analysis of statistical probabilities. Despite the massive parallel processing capabilities of such a being, doubt brought with it slightly slower decision-making. It also brought downstream effects that a simpler computational intelligence could not experience. An AI did not know fear as a human being might, but whatever analogous sensation an AI could have, Andastra discovered that she had it.
2
Jake
The operations command center normally flowed with a stable contingent of logistical staff and team coordinators. Now it buzzed with tension, as well as the physical presence of senior leadership. Askew was there, with his normally casual demeanor displaced by an atypical severe countenance and a palpable projection of authority.
“What do you mean Kritcher is in trouble?”
“Sir, he has a problem with propulsion. Thrust and maneuvering. Velocity dropped, then power output sagged,” said a ginger-haired analyst seated in front of several screens filled with telemetry and various graphs.
“What?” Jake said. “We’ve never seen anything like that befor
e. Are you sure your data feed is working properly? What is Kritcher seeing? Let me talk to him.”
“Yes, sir,” replied the analyst.
“Thys, what’s going on?” asked Jake. “Talk to me.”
Thys responded immediately. “Sir, just after arrival, the sensors detected a lot of nearby objects. Dense and relatively small in relation to planetary-size masses, but large in relation to my own craft. Like a dense asteroid field—but one shouldn’t be here. Not with objects like these, this close together, this far out.”
Jake heard Thys suck in a deep breath, controlling his breathing.
“Then a few minutes later, as I continued my course and approached the nearest objects—that’s when it happened. Velocity dropped. Sharply. I checked the controls and then tried to accelerate and get back up to speed. At first I thought it was just something with the navigation waypoints—what else would decelerate me? I found the controls seemed sluggish.”
“Sluggish?” asked Jake.
“The engine wasn’t accelerating me at the same rate I’d expect. We’d typically target this distance out from a star for arrival. A few minutes later, when I reached the same orbital range as the closest of these nearby objects, the drive started losing more power. The interface is showing strange information on the display. Things I’ve never seen before. It looks like the baseline power output level is now below what we get when we bring an engine online—an order of magnitude lower.”
Jake shook his head and tried to process what he was hearing. He had never experienced anything like that at any time in all his flights powered by the Dhin engine. Even his unplanned detour to the Dhin outpost hadn’t been like this.
“I’m sure you’ve double-checked the control interface, right?” he asked.
“Yes, sir,” replied Thys.
“And you’re not going to hit anything, are you? You aren’t that close to anything in that field of objects?”
“No, thankfully. It’s not that dense. There are a lot of them, but they’re not packed in that tightly. My luck’s not that bad.”
“Well, with your field on, hitting something wouldn’t hurt anyway. Any idea what they are?”
“No, sir, other than that they’re all pretty close in size and not as small as you’d see from an unformed planet. They’re not big like planetoids. The closest ones seem to be made of metals, not rocks. The creepy conclusion is that they might be spaceships. Like mine.”
“Whoa,” said Jake, “that might be jumping to conclusions, but . . . wow.”
“Yeah. Given my situation, it’s what came to mind.”
“A graveyard,” pondered Jake. He turned to the engineers and analysts at the workstations and said, “OK, what do we think about that? Obviously, it’s possible, but are we crazy to consider it? What’s our analysis of the data he’s sending back?”
The red-haired engineer glanced at his colleagues and then offered, “Well, we’ve got better processing power here, but we’re getting the same answers as Thys is from his computer onboard. These things are made of alloys and some other materials. A section of one object has the density of a lighter composite material that’s less reflective, along with some highly reflective sections. They range in size, as far as we can estimate quickly, from the size of a car to a bus, up to maybe the size of an aircraft carrier.”
“Hmm. Yeah, I guess that seems like spaceships,” Jake muttered, then, more loudly, he said, “But do they look like what we’ve seen from the Dhin? Are they close enough to tell?”
“Well, I can’t tell with this little screen, but Brad and the team there think so,” Thys replied.
Jake turned to one of the engineers and gestured for him to elaborate.
“Well, the latest active-sensing equipment powered by the latest engine power couplings is far better than the previous generation, so we’ve got pretty good imaging, sir. Like Kritcher said, we’ve got great resolution and image processing on this side of the data feed.”
“Go on,” prompted Jake.
“Yes, sir. They do look somewhat similar to the component designs of the original prototypes left by the Dhin, as well as the station you visited.”
“Anything look just like that?”
“Well, we don’t see any objects that look exactly like it. There are lots of these things out there, though. Some are too far away to tell. We don’t want to rule it out, sir. It’s too soon.”
Jake frowned. “Bring some of them up on the big screen. I have to see this.”
“Yes, sir.”
Jake looked at four images that appeared on the wall-size display. His eyes widened, and he nodded his head.
The four displayed had the parabolic and elliptic curves they’d seen in Dhin design, with seamless intersections and a monochromatic surface. What they’d seen of the Dhin’s work always reminded Jake of Italian futurism. These shapes looked very much the same. Not exactly, but close.
“Yep. Those look something like Dhin designs. Hey. There’s no light coming from them. I can’t see any specular highlights or distortions—so maybe no fields either. Am I right? Is that the case with all of them or just these? Do we think they’re—derelict? All dead?”
The red-haired engineer spoke immediately. “Sir, we didn’t want to jump to conclusions. As we said, we haven’t closely examined them. Or even examined them all. You see how far away we are. But that’s sure what it looks like. None emit any light. On any frequency. Or internal heat. And as you’ve noticed, there’s no evidence of a Dhin field active.”
“A graveyard,” mused Jake, “and Thys is losing power. So I’m going to be the one to say it. This could actually be a trap.”
“Yes, sir. It could be,” sighed the engineer.
Fletcher
Fletcher reached across the table and deftly nabbed some of Mare’s sweet potato fries.
Mare glared at Fletcher and quipped, “May I have some of your fries, please?”
He gave a smug grin while he chewed, swallowed, then parroted Mare’s tone and inflection. “May I have some of your delicious sweet potato fries, please, Mare?”
“Ah, Greedy Gus is a smarty-pants, as usual,” she retorted.
“I’ll take that as a yes,” he said, then swiped another handful of fries. “You never finish them anyway, you know.”
“He can learn a new programming language, but he’ll never learn manners,” she sighed, rolling her eyes and giving a theatrical shrug.
“Mmph,” Fletcher replied, mouth full of fries.
“So have we any plans for the evening that you’ve neglected to share with me till now?”
Mare picked up her comm pad, tapped in a passphrase, then glanced at her smartwatch and entered the alphanumeric code that appeared on the small display into another field on the pad’s screen. Another prompt appeared, and she pressed her thumb against the screen. A moment later, the tablet displayed her home screen. She opened her calendar and an app that aggregated local events and gatherings she might find interesting.
Fletcher slurped the last of his diet root beer, then grabbed his own tablet and went through a similar login sequence to Mare’s. Next, he opened an encrypted ephemeral chat app and pinged Mare’s local address directly. The near-field network didn’t traverse Globalnet or use a service in the Cloud. Mare opened the app on her own pad.
Their on-site apartment used a CoSec-provided network to reach Globalnet and the Cloud. They both knew that nothing that traversed it was private at all. Fletcher typed out a message. While there was some level of trust CoSec staff knew the organization had for them, neither Fletcher or Mare presumed that their shared apartment wasn’t bugged. They both knew they weren’t quite violating agency policy by chatting this way—after all, agents on assignment would often work this way. Then again, they weren’t field operatives. Not exactly.
Fletcher considered the justification they used for their behavior. The enemy AI knew both of them. They had a history with the rogue AI, more so than anyone currently at CoSec did.
Of any human targets at CoSec, they might be of some special interest to the enemy intelligence—despite their limited authority and incomplete knowledge of CoSec strategy. Even their managers didn’t have the clearance needed to know about it.
I think we ought to check out that underground robotics group again, but I know it’s not your favorite, he typed.
He’d developed an almost fetishistic attraction to a couple of groups. Since they were Darknet only, he presumed that if Nick were to send either of them a personal message, it would arrive at such a venue. Of course, he loved the robot battles too.
OK, sweetie, she replied, so long as we can go by that art gallery I like first. They have a new installation this week. Tit for tat.
Fletcher grinned and gave Mare a satisfied nod. He liked the art at that gallery well enough but not nearly as much as she did.
He sent a message back: Wanna hit the bedroom before we get ready to go?
Mare replied with a sultry look and a nod of her own. He closed the app, hit the sleep button on the pad, and then they both stood; she took his hand and led him back through the short hallway to the small, dark bedroom. He left the light off.
Esus
The AI watched as the battle progressed. Thousands of images, video and data streams, and damage and logistical status messages filled his consciousness. Space before him glittered and flashed, awash with myriad pinpoints of reflected light. In patterns, there were waves of brilliant plumes. At other times, whole areas of space appeared to explode into fragments.
Esus observed. Coursing through the massive data structures the Dhin provided, Esus extracted and refined raw information into a useful form. But useful apparently was not going to be enough. Esus altered stratagems. Redirected resources. Attempted to adapt.
We are losing.
Esus commanded tens of thousands. A fleet of robotic weaponry, autonomous spacecraft, and guided ordnance all answered to the AI’s command.
The Power of the Dhin (The Way of the Dhin Book 2) Page 3