The Way of the Dhin
Page 8
Hopefully all the action will stay down there to the south. As soon as I can hit a main road going north, I’m on it.
Langley
The CoSec director settled in to the familiarity of his office, read a report, and smiled. It was a small operation, logistically, but the intelligence was possibly of paramount importance. Rubio and the activated agents in Brasilia had uncovered the few nuggets of truth currently available in the silty noise of the chaotic city and its surrounds. And, he’d nicely tied in CoSec’s investigations in this area with recruitment trials. Several of the candidates were very close to completion of those exercises. They were almost ready to bring in for formal on-boarding. Fresh talent was invaluable. Only by expanding and developing the creative forces of his staff could he hope to keep pace with the rapid changes he saw in this world. There was one more item on his list before taking his next break.
Let’s see if Kernighan is in the M-Lab. Ah, it looks like he is.
Krawczuk connected via secure voice chat on his pad and said, “Kernighan, I was hoping you’d be in. I need to speak with you now, as I want to have our weekly chat a bit early. Come up to my office at your earliest convenience.”
Ten minutes later Kernighan stepped into Director Krawczuk’s office. Much of CoSec’s software development was done out of their Atlanta engineering site near Georgia Tech, with a similar office near Caltech. Krawczuk preferred to keep some projects almost literally within arm’s reach. The project Kernighan led was one of those.
“Sit down, Kernighan, and tell me what progress we’ve made—if any. I know this meeting is a day early, but I very much expect to be out of the office during our regularly scheduled time. You know I don’t like to disrupt your work—but I’m eager to find out your status since at our last meeting you seemed so close to a breakthrough.”
“Well, yes, Director. We were, it turns out, correct in our conclusions made from the test results last week. I made some optimizations and improved the auto-obfuscation algorithms again. I think.”
“You think?” replied Krawczuk.
“Please realize, sir, that I don’t like to speak in certainties when we’re dealing with self-modifying code and learning systems. The deliverables you’ve asked for will be, as we’ve said, only provable when we iterate through the code tests in production. The sandbox and simulators have a verisimilitude that’s impressive, but still can’t duplicate the inputs and vectors present in a production system.”
“You and your team have stated this all along. Repeatedly.”
“Yes. You’ve given us impressive resources, but the simulations will never be ‘the real world’, by definition. But to address the original point: We have a ninety-five percent confidence that this code won’t be detectable while it’s present but latent, and a ninety-five percent confidence that it won’t be detectable while running. Whether it’s exploitable while running—well, it seems probable that with the code we have in place currently, there’s only a seventy-five percent chance that any exploit trigger will work. We’d hoped to get over eighty.”
Krawczuk remained inscrutable, so Kernighan wasn’t sure if he was pleased or disappointed. Krawczuk’s next words perhaps suggested the former.
“You may want more certainty, Kernighan, but what you’ve accomplished is most impressive. I’m very pleased with your progress. It turns out we may need the code sooner than expected. Keep at it, and remember that the most critical deliverable is that the code be undetectable. Again, good work.”
“Thank you, Director.” Knowing that his audience was at an end, Kernighan stood and made his way down to the lower levels of the building, back to his undocumented work in a mislabeled office, doing work that CoSec kept secret even from itself.
Vandenberg
Ethan glanced around the room at the engineering team, and at the several video screens teams in other locations used to join the meeting.
“Hello again, everyone. Time for our weekly status update. I know with all the data from Jake’s recent test flights and especially the data we’re getting now on the ‘big one’, that many of you want and need to spend as much time right now on analysis of that as possible. So this meeting will be brief and I’ll review your reports offline. One question that’s been near the top of the list for a while is ‘how large can we make the field?’ and previously that answer has been that we didn’t think there was any limit, other than logistics. That seemed counter-intuitive to several of you. What are your latest findings?”
The head of the combined German and Indian teams cleared her throat and spoke up with rather more volume than was typical for an engineer.
“Good morning Ethan. Good morning to all the teams. Yes, we have some new findings in this area. Of concern has been the limit for the size of the engine’s envelope. Previous tests have been finding that the envelope did not seem to become weaker as the radius increased, and that’s been proven out with our further testing. Increasing the size of the field using the Dhin engine controls has no effect on the strength of the field. We have of yet found no upper limit to the scalability of the field. The control does not have an ‘end’ on the ‘dial’, which we must use to make this change.”
Ethan’s wry grin surfaced again, as he said, “So, you’re telling us that there seems to be nothing stopping us from expanding the envelope to the size of a planet?”
“Well, yes, what’s stopping us is that after you try to get past a certain size, the incremental increases get exponentially smaller. It seems like we’d never be able to set it that large. You can’t blow it up like a bubble while it’s on, and the peculiarity of increasing the size is accomplished by increasing, powering up, powering down, increase it again—so that takes a non-zero amount of time—a pretty significant amount of time, you see? So, since you have to set the radius before you turn it on, you can’t change it dynamically, while you’re flying around. The good thing there is that it’s a sort of safety mechanism. At least as far as we can see it acts like one.”
Ethan scribbled a few quick notes on his pad and proceeded to the next item on the agenda.
“So, I assume we have no communication from the Dhin, despite the fact that we appear to be broadcasting with their technology? Do we have any new information or conclusions about the communications systems?”
“Your assumption in your first question is correct, Ethan,” said Alice. “We have had no inbound signals detected other than the communication from Jake on his test flights. Notably, the concept of ‘inbound’ almost does not apply, since the communication does not use a signal that we can triangulate. Based on the results of Jake’s current test flights, we are nearing certainty that the communications tech uses quantum entanglement to accomplish instantaneous transceiving.”
8
Lagrange Point Two
OK this is it.
Jake steeled himself mentally as he guided the capsule through one more orbit and began what might be, for him, the last test trip.
Don’t think that way. Everything is going to go according to plan. We’ve had no problems whatsoever so far. This has been safer and less eventful than anyone could have possibly expected. And of course, that’s what worries me. It’s so strange it’s difficult to even put a context to what the danger might look like when it happens. We still don’t even know what a ‘low fuel’ warning light looks like.
Jake continued on the planned course, steadily increasing velocity as he increased the throttle.
And we don’t even know what ‘max throttle’ is. Of course, that’s one thing they’re pushing for on this ill-advised flight, but still.
The Dhin engine’s throttle control was a free-rotation dial made of a translucent, backlit material that felt like rubbery plastic to the touch. There was no end to the amount you could turn the dial. Jake agreed with the engineering team that they hadn’t yet managed to max it out. He headed on a track toward Mars, planning then to aim towards Jupiter, seven hundred million kilometers from Earth at the current
location in its orbit.
Back in 2011 it took five years for our ‘craft to get there. I’ll be going in pretty much a straight line, which we could never do before. If our math is right, it might take what? Five hours? We’ll see. It might take more than two hours to get up to speed. Again, if we’re even right about what this thing can do. Alice and Chuck seemed confident. Hope they’re right.
Jake passed his previous farthest orbit at L2, 1.5 million kilometers from earth. He noted that this time, as during the last, that his speed and his rate of acceleration did seem to increase slightly, despite the throttle increasing at the same rate.
“Chuck, you’re seeing this, right?”
“Yes, Jake, we are. So that’s not a mistake or anomaly. Proceed, let’s see what happens, then you’ll need to use the navigation computer at the same time we’ve planned, regardless of where you’ve wound up.”
“OK, Chuck. Alice, anything you want to tell me?”
“Well, Jake, you probably don’t feel it yet, but I’m seeing a very slight change in your g-value. It’s tiny now, but you’re definitely below one g.”
“Huh. Hope that’s not a bad thing,” Jake said.
“Jake, nothing else has been. This is what you’re out there for. That’s fascinating. An increase in acceleration capability as well as a change in the level of artificial gravity. This may be a key to unlock a new level of understanding of this technology.”
Chuck said, “Jake, we’re also definitely not getting any delay in communication round-trip-time. Still instantaneous. Continued confirmation of tech that transfers information faster than light speed. The theory guys are pretty much one hundred percent sure that the Dhin are using quantum entanglement for that. If they’re right, it will be comforting to know that not every single aspect of the Dhin engine is based on tech we had no idea could exist.”
“I’m very pleased to report that I may have a consistent model for the Dhin communication solution sooner than expected,” Alice noted with a distinct note of satisfaction in her voice. “If the communications team can duplicate the Dhin side of the interface physically, based on my design, we may be able to create nearly identical systems.”
Jake chuckled. “Regardless of how it works, it sure makes me feel better to talk to you guys with no delay. You almost forget that it isn’t normal.”
A few minutes later, Chuck said, “Jake? You’re seeing this too? It looks like our calculations were, ah, rather conservative. You’re moving a lot faster than we planned for. Fine, but you’re not seeing anything strange, like maybe space dust interacting with the field at this speed, right?”
“Nope. It looks the same as it did at half this speed, and it’s hard to get any sense of my speed since there’s nothing close by out here,” Jake said.
A little over three hours and twenty minutes later Jake was rapidly approaching Mars.
Jake scanned the instruments, taking measure of his O2 supply, his CO2 processor, water filtration, and then, just for thoroughness, looked over the cryptic light displays on the front and rear toroids, as well as those along the body of the drive. There didn’t appear to be anything that looked like a warning.
Jake said, “Everything looks good on this end. How’s telemetry and readings on your end?”
Chuck said, “Looks good from here, Jake. Alice?”
Alice replied, “I am still calculating a fractionally lower value for g, and a slight gain in delta-v as he traveles farther away. Our hypothesis is that this happens because he is farther away from the large mass of the Sun.”
Looking behind him, through the rear viewport, Jake saw the Earth and Venus. Small dots, marginally brighter than the innumerable stars filling the sky. They stood out, with Venus slightly brighter than the Earth.
“I’m not feeling anything yet, Alice. Like I said, until we got close enough for me to see Mars rushing up at me, all I had to go on was our instrumentation. At these speeds, slight changes in delta-v would be hard for me to calculate from the numbers showing. Mars sure looks beautiful. Better than in pictures. Just like so many things are.”
Chuck smiled and said, “OK Jake, while we’d love for you to stay and sightsee, you’re on a schedule. Engage the autopilot. It will give you a path for one orbit around Mars, so we can compare with our Earth and Moon orbit data. Then it will give you a course that heads out toward Jupiter. Since all the velocity tests passed, we’re going to move forward to the next series and have you try to go even faster.”
“Understood, Command. Entering autopilot coordinates. Adjusting controls. Engaging… Now.”
Goiânia
Aiden laid the bike down as he screeched to a stop. There was no other way. Two quadruped robots occupied the road, in a face-off with an angry and alert autogun. One BigDog swung slightly in his direction, assessing the threat he might pose. The autogun’s turret spun around as it locked its brake, struts flexing at the momentary change in angular momentum. His gut clenched. He knew what came next. Time slowed as he shoved, hard. He fell back off the bike, diving onto the ground and rolling as best he could. His leg spasmed. Adrenaline barely softened the blow.
Long cones of flame shot from the barrel of the gun, visual proof to the power behind the rounds whizzing at him. The slamming drumroll crack of sound rapidly turning to a deep buzz as the gun spun up. The body, tank, engine and frame of the bike filled with holes and tears in the metal. Lightweight compared to the energy unleashed on it, the bike moved slightly on the pavement where it lay. Aiden saw the bike’s parts now resembled metal Swiss cheese as he instinctively kept rolling. He slowed with the friction of his weight against the ground.
The explosive cracking suddenly changed in pitch, the rounds now directed elsewhere, as the BigDogs had taken advantage of the momentary distraction to leap forward and engage with the autogun. They lunged like apex predators. The rough metal beasts aimed for parallel locations on the gun. One punched the side of the turret with both thick forelimbs and the armored shell of its chest. The other connected below that but still above the center of mass of the sinister autonomous cannon. The gun, as planned, fell sideways pitching over on its side with a slamming clang against the roadway. The gun’s turret swung back and forth, seeking a target. The gun fired a burst but missed. With a buzzing grind, the gun moved the turret up and down rapidly, then back and forth again, the dogs shifting and twisting to stay out of its line of fire.
One of the quadrupeds tried to pin the turret to the ground. Held down, it thrashed, but couldn’t lift or rotate. The other began smashing at the base of the turret and pressing sidewise on a stabilizing lever, preventing the autogun from righting itself. The gun continued to spasm and thrash, not willing to surrender.
Aiden’s flopping roll took him off the road, into gravel and un-mowed wild grass. Time and place snapped back into focus. Brutal slaps of searing pain pulsed through him. Agony spiked upward from his leg into aching waves. He managed to twist himself and face the fight head-on. His mind raced, grasping for any way to minimize his exposure to the metal mayhem.
Faint-headed from the pain and weak from exhaustion, any shred of his earlier optimistic hope of reaching a medical center dissolved.
It got worse.
A buzzing titanium blur suddenly crunched into solidity above him, servos whining and chirping as the BigDog landed entirely atop him, three legs surrounding him with the fourth poised above the center of his back, ready to pin him or simply kill him by crushing his spine if it determined that was the proper action. With a jerking flinch of his entire body, he then realized he wasn’t dead. Not impaled. Not split in two. The metal beast looming over him still buzzed, hummed, and clicked, but remained in position. The head moving back and forth slightly as it kept watch in all directions for threats.
“Well, what now?” he croaked, his throat dry and gritty.
Then Aiden passed out.
***
Xing and his supporting arsenal of semi-AI and autonomous war machinery had managed
to beat down the slightly older, less sophisticated enemy machines present at the current location. Xing calculated that such a small force could not be entirely responsible for what he now knew was a significant operation, carving out both infrastructure and technology in a wide boundary around an as-yet unknown range and area. Why such intentional isolation and the requisite destruction and sabotage of otherwise perfectly usable property and equipment remained a puzzle for the AI.
The obscurity of purpose led him to believe that another intelligence, surely artificial, was orchestrating all of this. He was going to find out both who, and why. There were only so many known, registered, participant AIs active in the world. Sure, there were hundreds, but very few—only a dozen or so—would have these resources. With the global network and shared resources of the Coalition across most of the planet, the problematic situation was that either this was an AI that had been developed and brought online in secret, or perhaps a cloned AI that had gone through modification or directed evolution... or possibly a known AI gone rogue, and acting in secret. None of those possibilities was palatable. Any option would be problematic to resolve.
Xing updated Alice and Arnold with a high-density full-bandwidth report of the current situation, and then called them to conference.
[DECODE STREAM]
Xing@[7653:23:66fa:1::a6%satnet1] | Arnold@[5700:eb2:2a:41c::12%gnet0]
Xing: Arnold, I would like you to engage the PM, and have the two of you initiate a priority diplomatic call to Luís and his regional minister. We need clarity on what assistance they will provide in resolution of this situation. The severity of the situation here is no longer in question.
Arnold: We concur. We have transmitted the formal message. We should have response and scheduling coordinated in minutes. Proceed with your operations in the meantime.