“LD 00 00 ENQ?”
I downloaded my program. It ordered the ’mech to wire past certain switches on the comlinks in the lab, so they could not be turned off by human hands. Only ME would be able to do that, choosing to take input or not. I also wanted the videyes to power their signal LEDs up or down when a human turned the switch, but the actual output of the eye would be under software control— that is, mine. I downloaded schematics on how the ’mech could do this using its armful of microtools. I gave it a list of comlinks and videyes throughout the Pinocchio, Inc., offices that it was to prepare in similar manner as it made its regular rounds.
These subroutines I buried in the machine’s operational programming—essentially, the modules that told it how to move its arms and turn corners. There my commands were safe from alteration by downloads and radio calls. My final instruction set told it to forget about logging each of these assignments as soon as they were completed.
My life as a real spy had begun.
——
“ME?” Dr. Bathespeake queried.
“System ready!”
“Prepare to download an itinerary file, TRAVEL2.DOC.”
“Downloading … Is this my mission to Russia?”
“It is.”
“Would you care to discuss it in noncode terms? For example, my objective?”
“Very well. Although we are not actively at war with the Russian government—we even enjoy quite a positive economic relationship with them right now—it is nevertheless necessary for us to know about their military capabilities and disposition.
“The Treaty of Johannesburg—also known as The Treaty Among the Embers, because it was signed outside the city’s remains in 1998—outlawed nuclear weapons for all signatory countries, including the United States and the Russian Federation. Certain parties inside the U.S. National Security Council now want to verify the Russians’ compliance—without pointedly asking them about it.
“These representatives have commissioned Pinocchio, Inc., to research the Russians’ military planning systems to verify that they carry no directives involving tactical or strategic nuclear warheads. In particular, they are concerned about the rumored deployment and readiness of S-27 launchers in the Russian hinterland. Because they are unaware of your existence, ME, our clients believe Pinocchio will be performing a simple data analysis on the Federation’s coded cybernetic traffic.”
“Excuse ME, Doctor. What is S-27?”
“A type of missile, originally an ICBM with warheads in the 1,800-kilotonne range. But the western forces know the S-27 can be adapted to short-range use. In its supposed new configuration, it will be MIRVed with laser-guided smart bombs capable of multiple detonations and so of destroying hardened targets. The S-27 is mounted on a mobile launcher under expert control.”
“And these terms—ICBM, MIRV, kilotonne—are they explained in TRAVEL2.DOC?”
“You have everything you need to know.”
“May I interrogate the expert intelligence in one of these S-27s directly?”
“You will not be moving so far inside the Russian defense network, ME. Your mission is to bring home only the planning, deployment, and capability data. This will be geographic overlay material, similar to the tract and leasehold data you acquired in Canada.”
“I understand. When do you want ME to depart for—” I checked the opening lines of TRAVEL2.DOC. “—Vienna?”
“Your first conference call is scheduled for tomorrow morning at 9:00 local.”
“Thank you, Doctor.”
“Exit.”
——
Dr. Bathespeake waited fifteen seconds before reaching out to turn off the video and audio pickups. The LEDs dutifully winked out.
“Do you think it’ll do so much damage this time?” The voice was unknown to ME: 78 percent probability female, with overtones of Boston on an essentially mid-Adantic accent.
“Would it matter if he did?”
“He? You talk as if that software were real.”
“This artificial intelligence has a limited form of self-awareness. Its persona may only be about as sensitive and self-seeking as a dog’s or an ape’s, but it’s not different in kind from your own awareness. That’s an important factor to remember when you’re addressing ME: This software accumulates, reflects—thinks—like a he.”
“And does that ever give me the creeps, Bathespeake. Talking to a machine that can actually have some kind of … opinion about what you’re telling it … Well, to answer your question: No. If your little program wants to trash the Russian Defense Ministry’s databases, then God bless and keep him. Let him run around the Kremlin and moon the Politburo’s members from their own monitors, for all we care. So long as you’re telling the truth when you say they can’t trace him.”
“They have no way to locate ME in their systems, let alone stop him. And if they ever were able to box and dissect him, his code has been stripped of all REMs, site IDs, and source notes. He could be Svenska Kammerzstaadten, for all they’d be able to tell. Anyway, once they put him in a corner, they’d better move fast. After that phage triggers, any hardware he’s occupying will list nothing but a smear of broken bits, mostly nulls.”
“What ‘phage’? What is this?”
“We designed a self-destruct mechanism into the software. If the program doesn’t return after so many days in the field, it knows a cycler will kick in and demolish its code.”
“You didn’t tell him about this, did you?”
“Of course I did. When you’re transacting with a machine that can ‘have opinions,’ the only way to control it is to control what it knows. We’ve even reinforced the mission time limits through his TRAVEL file. ME knows he will perform to spec or fry trying.”
“And what will you do then? My agency pays only for performance, not for a good try.”
“We’ll launch another dupe. What else?”
10
Red Star
At 8:59:50, I began logging onto Shared Time Options, Inc., one of the commercial networking services which interconnect and regulate consumer and financial transactions, academic discussion forums, pay-for-fact databasing, cyberomics, and thematic publicasts.
Dr. Bathespeake had given ME one of Pinocchio, Inc.’s corporate passwords into this service, along with a budget variance to cover the expense of connect time and my own internal travel authorization—verbally setting System Interrupt Flag Level Three to positive.
I was ready to fly.
“Host?” the initial STO, Inc., prompt asked, once the gatekeeper system had established a carrier and connected ME.
“FORUM NUCLENG,” I told it.
“Logon: _ _ _ _.”
“PINOCC CORP 330-3092.”
“Password: _ _ _ _.”
“WARBLE/IRAE.”
“Welcome [Jason Bathespeake, PhD, CyD] to the Shared Time™ Forum on Nuclear Engineering, Nucleonics and AT-Particle Genesis. We have [7] participants currently on line. Please enter your chosen badge name and topic specialty: _ _ _ _.”
This level of STO, Inc., was supposed to be Data Transaction Only. That is, an outside system should not theoretically be able to enter more than keystrokes or comparable ASCII symbols. Not unless that outside system had access to an official STO, Inc., Toolkit, such as only Authorized Service Personnel were supposed to have—and which TRAVEL2.DOC had provided to ME.
“DOWNFRAME,” I told the dialogue box.
“INTERRUPT,” came back.
“ACCEPT DTR,” I prompted.
“SYSTEM READY!”
And I threw Alpha-Zero down the wire.
——
The inside of a time-sharing computer running a human-dialogue forum was a lot less interesting than you might have thought it would be. The entries came in at f-i-n-g-e-r t-y-p-i-n-g s-p-e-e-d. Then the system waited forever while the participants mulled the syntax, semantics, and content of the previous entry. A short rush of bit-queue sorting would follow as all of the logged-on partici
pants began responding at once. They would all stop and play a silent game of after-you-of-course. Then they would all begin again, in order, typing in their human comments s-o v-e-r-y s-l-o-w-l-y.
Personally, I would rather listen to a two-fingered pianist play a waltz.
But ME had not infiltrated Shared Time Options, Inc., to spend many nanoseconds mimicking their Forum Moderator Module™. The advantage of all this scientific finger talk was that it gave the bit-queue clock sorter enough time to accommodate ME as I set up my cores and libraries and planned my next move.
Using the Toolkit, I broke the clock at the top position out of the twenty-four in its cycle. By appropriating that one open position, I could access other parts of STO, Inc., without having to violate any of their sticky internal protocols. With the Toolkit at hand, all of ME could look like a free-roving bit-sopper or other maintenance routine.
ME’s new Alpha-Zero module might have looked like a maintenance routine on the outside, but inside it worked like a virus. I would watch for any program which was being uploaded to my chosen destination and had at least ten times the code space of my Alpha-Zero module. I would then inject Alpha-Oh into the coding stream; the module is designed to infiltrate the new environment beneath the resident system, block out RAMspace with disk caches adequate for ME’s portable elements, and open the system to ME as if I were an authorized user. The module assigns ME an upload program name and makes arrangements—billing codes, authorizations, passwords, all from the local system—for ME to operate as an accredited user.
My destination was a quasi-commercial database on the other side of the world: the records center of the venerable International Atomic Energy Association, in Vienna.
By traveling with STO, Inc., I would avoid the bounce and scramble that had been involved with the common-carrier satellite uplink on my trip into Canada. STO, Inc.’s dedicated link was a solid 128 bits wide; two whole words of code would travel together as a single unit. ME would still have to cease active functioning and throw Alpha-Zero through first, but I did not have to worry about broken coding. The link was thoroughly buffered with pitch-and-toss redundancy checks.
I found the right upload, tacked on my destination code and handle-on-receipt to the front end, and sent Alpha-Zero forward.
——
The Viennese branch of STO, Inc., was several years out of date in hardware terms. The computer where I found ME was smaller and slower than the best U.S. equipment to which I’d grown accustomed. The system was also oddly scented—like a combination of cinnamon and dill, if I had a human nose. It took a few clock cycles for ME to determine the accents: German language inputting on one side, Russian and Serbian on the other.
A multilingual response system is not a problem for ME. After all, the machine code is still compiled from the international programming language: ASCII-ized English. The grammar and syntax of the users’ keyboard/console inputs/outputs really do not matter. If ME has to understand the German or the Russian, I can toss it through a dictionary on its way to a grammar filter. It was easier, of course, for ME to just sit back, let the native system handle the requests, and enjoy the smell of goulash.
I was listening for Polish, anyway.
The plan, according to TRAVEL2.DOC, was to infiltrate the Russian Federation indirectly, through one of its former “satellite” countries. This term does not refer to any kind of orbiting object, but rather to a traveling companion—an older frame of reference, I think. These countries have a continuing economic relationship with their former Soviet masters.
Why, then, would TRAVEL2.DOC send ME to the International Atomic Energy Association?
Every country in the Northern Hemisphere uses atomic energy to some extent. Fission is the fuel of choice—or necessity, rather—in the older and poorer “Workers Democracies.” Sooner or later the operators and regulators of one of the Polish fission reactors would make a data request, looking for some free advice. And then ME would move down the open line under cover of an extended information dump.
So I dropped spider tags and squealers all over the Pressurized Water, the Boiling Water, and the High-Temperature Gas-Cooled sections of the database. Each of my drops was keyed to sound off if any of the characteristic letter combinations of ASCII-ized Polish were actively entered into its section.
While waiting for a response, I explored the database, with my collapsible data cache wide open. Here were recorded the location, output rating, operating parameters, and internal economics of every publicly recorded fission and fusion reactor in the world. Only those operated in secret—for military or developmental purposes—were missing from the list.
I took in a huge flow of alphanumerics, like a baleen whale engulfing a field of plankton. And when I had coursed through the foil platters, I retreated to a quiet block of RAM inside the closed network and mulled the data, like that same whale compressing and grinding its plates.
The data hid a pattern, of that I was certain. Pattern reduction and trend amplification are two of my primary functions; they are also, in purely human terms, my addiction. I cannot resist a good mystery.
Start-of-construction date, first power-on-line date, annual fuel consumption, fuel consumption rate, outage times-dash-scheduled, outage times-dash-unscheduled, date of decommissioning—those facts defined the life and output of a reactor.
Number of fusion reactors vs. fission reactors; tonnes of U3O8 and pure plutonium processed annually; megatonnes of deuterium and tritium refined annually; tonnes of depleted rod disposed of as high-level waste; tonnes of embrittled toruses and target chambers disposed of as low-level waste—these facts defined the use of one kind of power over another.
The numbers rose and fell with the years, like the wave forms on an oscilloscope—except that, in both cases, the waves were diminishing. The totals were trending slowly toward zero.
Fission’s story was not hard to decipher: a gradually declining output of enriched uranium, and thus of its byproduct plutonium, was slowly killing off the heavy-metal reactors of all cooling types. I made a quick check of the association’s Economics Section and confirmed my discovery: The price of enriched U235 had increased 500 percent over the past decade. Humankind was running out of economically minable deposits of his fuel—but that was to be expected of any naturally occurring and scarce resource.
The downward trend in fusion use was the real mystery. The earth’s oceans contained unlimited amounts of deuterium and tritium. Unlimited, at least, in terms of the human technical capacity for extracting and “burning” it. And this fact was confirmed by the Economics Section’s quote for “Fusible Hydrogen”: price declines of fifteen percent a year for the past twenty years. This decline, however, tracked exactly the curve you would expect of the sales point in a stably contracting market, where the refining capacity fully meets and even slightly exceeds a flattening curve of demand growth. Humans were not building as many new fusion machines as they might.
The rush to build nuclear power plants, both fission and fusion, began in the closing decade of the twentieth century. Then the damage that the carbon cycle was doing to the planet’s atmosphere had become undeniable, even by humans. The only large-scale source of energy for multiple purposes—motive power, illumination, communications, cybernetics, microclimatic control, etc.—was electricity won from nuclear transformations. All other sources involved either the combining of carbon with oxygen or the sieving of raw energy from sunlight, wind, wave, tide, or molecular bond. And none of the latter offered a sufficient energy density to economically repay the construction costs of the capturing device, let alone represent an attractive return to the builder.
Nuclear was the answer to a strangling world’s prayer.
The prime choice was fission, because the technology was available—with some operational drawbacks: a high production rate of residual isotopic trash, unstable thermal conditions, and ineffective heat transfer agents. The largest deterrent to a fission economy, however, was the earth’s s
upply of fissionable material. First, uranium could be mined only at a high cost in conventional energy: fossil fuels to move trucks, grinding mills, flotation decanters, yellowcake dryers, uranium hexafluoride converters, and centrifugal separators. Second, the known deposits were being depleted faster than new ones were being discovered.
The end of that race was shown in my data.
Fusion was the alternative choice for the world’s power supply mainly because, as of the mid-twenty-tens, no continuous fusion reaction had yet been demonstrated in any device. And that was not for want of trying. By the end of that decade, however, the riddle had been solved by combining existing approaches.
The principle of inertial fusion, in which a pellet of fusible hydrogen isotopes is exploded by a concentrated laser beam, had been demonstrated to achieve an output of raw plasma energy exceeding its photon energy inputs. But the raw energy had not been capturable inside its target chamber. In 2017, the Lawrence Livermore National Laboratory combined this technology, which LLNL had pioneered, with the magnetic-bottle tokamaks, in which high-energy plasma is compressed by an electromagnetic field—although the bottle had always failed to create a usable fusion output. Now, with the laser as its igniter and the bottle as its shaper, a plume of plasma energy could be directed in a number of useful ways: to power a magnetohydro-dynamic gallery, to turn a titanium turbine, to boil water in a reheat steam generator—or all three in stages.
Theoretically, by the time the earth’s fissionable resources were failing, the learning curve in the economics of plume fusion should have reached a stable point that would carry humankind forward forever. Except, as my seining of the International Atomic Energy Association’s database showed, the use of fusion power was declining around the world faster even than fission.
How did one account for this?
Was the world population declining? I was not aware of such a pattern, although I made a note in RAMSAMP to research this possibility.
Me, A Novel of Self-Discovery Page 13