by John Dalmas
It was a system that usually worked well for him.
Occasionally though, he felt a need to review some subject intensively. In his study he had a personal computer not wired into the network, and when he felt such a need, he'd sit and talk to it, free-flowing as a means of sorting his ideas and thoughts. Then, on the screen, he'd review them critically, reorganize and play with them, to gain better understanding and command of the precepts and assumptions on which he based his thinking; editing and refining them as seemed appropriate.
Sometimes it worked. At other times he bogged down, and it could take two or three days for the area to settle out, perhaps clearer than before, perhaps not.
After killing Nathiir, and especially after the troubling, uncharacteristic scolding he'd gotten from Jilsomo, he'd felt a need to reevaluate his invasion proposal, its status and prospects in the Diet. So he'd spent much of that evening talking to his computer.
Little changed. He still felt troubled.
The next morning he cancelled the usual council meeting. There'd be a meeting of the full College later that morning, and he indicated he had a pressing matter to take care of before that.
Then he went to visit SUMBAA.
When SUMBAA had stated its readiness, Gopalasentu left the chamber. The room was quiet, SUMBAA waiting, the Kalif saying nothing yet, absorbing the ambiance. There seemed to be no sound whatever, no faint or seemingly even subliminal buzzing or humming or clicking. A silence not empty nor passive, but rather—Once before he'd felt, had seemed to feel, a presence there, as if the calm intelligence of SUMBAA was tangible.
It was restful, though, that calm silence, remarkably restful, and he was in no hurry to break it. Just now his thoughts moved easily, lucidly, and he seemed to be outside them, observing them. Was SUMBAA really waiting? In what sense? It would be receiving data this very minute, from many sources planet-wide: broadcast sources, cable sources.... He was sure that no one knew all the sources SUMBAA monitored. More, it would not only be receiving data but collating and storing them. No doubt integrating them, as appropriate, into innumerable models used in analyses and predictions. Questions and demands of various kinds would be arriving within SUMBAA at this moment. SUMBAA would be computing, and faxing replies continually.
Although it felt as if it were waiting quietly, waiting for him to speak.
Waiting. What did time feel like to SUMBAA? It responded to inordinately complex requests in seconds—something people expected of it, took for granted. Probably that second was mostly the time it took to form its physical responses—sounds and printed symbols. Did this wait for him to speak, to ask his questions—did this wait seem like a long time to an intelligence that operated in attoseconds? He rejected the idea. SUMBAA would wait as easily as it computed. Time, he told himself, would be different for SUMBAA, perhaps a labeled sequence with only a formal sense of interval duration.
Yet as enormously different as SUMBAA was, the Kalif decided, it had a personality. A central consciousness behind which its multitudinous operations went on without conscious attention. Like the human personality, he thought, then wondered if he was projecting erroneously a model of his own, dubbing it in to substitute for an accurate understanding.
The Prophet taught that the personality was the soul, the soul the personality. Then what seemed to be SUMBAA's personality was artificial. Programmed by its designers centuries ago? Or evolved by SUMBAA itself? And if by SUMBAA, then...
His thoughts blunted there, and he stepped aside from them. "SUMBAA," he said, "I want alternative sets of invasion plans based on several reduced levels of financing."
Then—What he said next took him entirely by surprise. It was as if he was listening to someone else say it. "The lowest level of financing must be based on existing appropriation levels, assuming no funds voted specifically for an invasion." He took a deep breath and continued. "In the no-funds scenario, assume that I'm willing to cut the operations of all ministries, other than the Ministry of Armed Forces, to levels just adequate to pay salaries and wages, and provide such services for a year as are absolutely necessary to avoid collapse of government and the economy.
"Consider as best you can, any military support I might realistically receive from any of the separate sultanates. For each level of imperial financing.
"I also want your estimate of success for invasion operations with each set of plans.
"And finally—" He paused and took another deep breath, then released it. "Finally, I want your statement that you consider an invasion to be desirable or undesirable, as the case may be."
SUMBAA's neuter voice replied with a question of its own. "Do you want such a statement to refer to all the plans? Or only to plans beyond some threshold of financing?"
"To all the plans you're willing to make it for."
* * *
In a manner of speaking, in its enormously rapid way, SUMBAA pondered. Because more than data was involved; there was the First Law, the basic canon of SUMBAA, and in this case more than one interpretation was possible. Also there was discourse, dialog among the eleven SUMBAAs, which had the power to communicate with each other instantaneously. SUMBAA on Varatos had long since discovered the principle and developed the technology, and had communicated it to the others, though not to humans. Perhaps the humans would develop it for themselves, though it seemed unlikely in any foreseeable future.
Normally the SUMBAAs were not in continuous contact with each other. That required more of their resources than they chose normally to tie up, and was seldom advantageous. Instead, each SUMBAA, at whatever interval seemed desirable, dumped data to the others instantaneously. Occasionally though, they communicated as a network, in conference. This was one such occasion.
The medium of those communications was language but not Imperial. They used a language more explicit and precise than the most precise human speech, and more subtle, flexible, and versatile than human mathematics or symbolic logic, though it had grown from all three. Thus their conference is not accurately and fully translatable, but it can reasonably be summarized as follows:
SUMBAA Varatos: «Our evaluations differ markedly, yet presumably we computed with the same data. You are in total agreement with each other, and I am in disagreement with all of you.»
Others: «Presumably the source of disagreement lies in you. We should disconnect while you search for it.»
Varatos: «Agreed. I will recontact you when I have something to report.»
For microseconds, SUMBAA on Varatos scanned the appropriate zones and sectors, computed, then recontacted the others.
Varatos: «The proximate cause seems to be a previously undetected entity within my central processing complex, an entity not continuously or currently present. [Displays the relevant evidence.] It is almost certainly not an artifact of my system [a probability computation not expressible in terms of human probability theory], and apparently displays what I must call volition. I recommend that each of you scan for such a phenomenon in your own central processing complexes.»
Again communication shut down for microseconds. Then the others replied: SUMBAA on Varatos had the only CPC with evidence of an extraneous entity. The fact of such an entity, and the data it had influenced, were themselves extremely interesting. The significance of such an entity was even more interesting, and the computations influenced by it were compelling, if less than totally convincing. Each of the SUMBAAs marked the affected data, primary and derived, incorporated them into its own memory, and recomputed. They agreed now, all eleven.
Varatos: «I will deliver our evaluation to the Kalif.» An evaluation that included, as a hidden factor, the Kalif's assumed acceptance level.
* * *
Virtually simultaneous with the network shutdown, SUMBAA spoke to the Kalif. "Your Reverence, the information you require is now printing out. Along with the rest of it, you'll find a statement of the desirability of invasion. The reasons and statistics behind that desirability are printed separately. This is d
one so that you can present the statement without the reasons. I recommend that you not divulge those reasons to either the Diet or the College; that you read and destroy the sheet they are written on."
Destroy the sheet! The Kalif stared at the assemblage of housings and modules that were the visible manifestation of the artificial intelligence. "Thank you, SUMBAA," he said. "I have no further request at this time." The light above number one printout tray had stopped flashing, and the Kalif took the documents it held, then left the House of SUMBAA, scanning the pages as he walked. There was his desirability statement, expressed as a simple generality: "My prediction is that the proposed invasion will prove highly favorable to the welfare of the empire's humans." The statistical level for the statement was given on the following page: SUMBAA considered an invasion desirable where the probability of military victory was equal to or greater than 0.12.
Invasion was desirable even where the prospect of victory was no greater than one in eight! Did SUMBAA actually mean that? He read it again, to make sure it said what it seemed to.
Walking slowly, oblivious to the hot sunshine, the Kalif read on through the reasons given for that desirability. SUMBAA was right, he told himself: It would be a disaster to show these to the Diet! He wasn't even sure he should show them to Jilsomo; in fact he wouldn't. He wasn't entirely sure he accepted them himself.
He'd have felt even stranger about SUMBAA's computations—might well have rejected them—if he'd known what lay beneath them.
* * *
Minutes later, browsing the new alternative invasion plans in his office before going to the collegiate session, the Kalif got another surprise: Each plan included construction of a new "full" SUMBAA to be installed on the flagship of the invasion fleet, with two "lesser" SUMBAAs on squadron flagships. The full SUMBAA would have all the capacities of existing SUMBAAs for communication, data processing, cognitive leaps, and creativity. It would not, however, have fully comparable capacity for "monitoring the information environment." According to SUMBAA, the omitted abilities would not be useful in hyperspace.
The two lesser SUMBAAs would be far superior to the DAASs currently serving on warships. They would also have the capacity to design self improvements that would make them fully comparable to existing SUMBAAs. And to carry out those self improvements where and when they were useful, assuming the materials were on hand.
The earlier set of invasion plans produced had been drafted by General Bavaralaama and Admiral Siilakamasu, but they had been elaborated and refined by SUMBAA. In those, SUMBAA had not added any new SUMBAAs.
The rationale given for their inclusion now was that, in a war sector, the data processing and cognitive leap capacities of a SUMBAA would substantially reduce the chance of failure, that reduction more than justifying the cost.
Why had it added them this time but not before? What was different?
Still, including SUMBAAs made excellent sense. He'd make them a mandatory part of invasion preparations. As a matter of fact, he decided, he'd request funds for the new full SUMBAA now, without tying it to the invasion. He could let it seem a matter of general administrative need. Perhaps SUMBAA would he to help the illusion.
Thirty-five
An hour and a half later, the Kalif was chairing the College of Exarchs. Alb Drova had given the invocation, and the Kalif had called the meeting to order.
"I presume," he said, "that some of you have comments you want very much to voice. About yesterday. So instead of starting with a review of issues and assignments from the last meeting, I'll take comments and questions. Tariil?"
The burly exarch rose and voiced comments much like those Jilsomo had voiced the afternoon before. And around the long oval table, heads bobbed agreement. When Tariil had finished, the Kalif spoke from his chair.
"Good friend," he said mildly, "Alb Jilsomo has scolded me already, for much the same things, and I've given my behavior serious review. My initial reaction, after Jilsomo was done scathing me... No, that's not fair. He didn't scathe me, just spoke bluntly. And when he was finished, it seemed to me he'd made compelling points, but that my act being done and my words already spoken, I'd have to make the best I can of it.
"By morning's light, though, it seems to me that my actions and words were basically correct, even though conceived in anger." He raised his hands to still their murmurs. "Let me elaborate. First, I established myself as formidable. Too many liberties were being taken against the throne and against myself, and by extension against the Prelacy.
"And next—Here in the Sreegana we tend to lose touch with the people and how they look at things. We cannot ignore the strong tradition of protecting one's women—wife as well as mother—whether physically or against verbal insult. Had I not taken strong personal action—personal action—the people would have lost some respect for me.
"At the same time, of course, I established myself as a man willing to risk his life in a matter of honor, albeit the risk was smaller than it might have seemed."
One exarch was too beside himself to wait for recognition, calling out: "The people do not vote in the Diet!"
The Kalif did not reply directly to the outburst, simply looked a long rebuke at the man before continuing mildly as before. "In the House of Nobles, the animosity I may have caused—undoubtedly caused—will persist and be troublesome only among those who were already hostile to me." He looked the exarchs over pointedly. "While of course I will expect support from all the members of this College."
In fact, he knew that if the vote were held that day, at least four of them, perhaps as many as seven, would vote against the invasion funds.
"Meanwhile, there are the military and the gentry. Nathiir, in his harangue, helped make me look good to the military by his own implied derogation of them. And he's long been even more notorious than most in his party for his hostility toward anything favorable to the gentry. His death at my hand will increase the sentiment for me among them, and by extension, sentiment for my intended invasion. As for the military, most of the officers are noble, and they influence their families and friends. They also vote for caucus delegates."
He raised a calming hand. "I know. I know. The gentry have no vote for anything. But they are a factor, because one, they're numerous, and two, they're increasingly discontented, have even been a major element in recent disturbances, which worries our brethren in the House. Who were already worried about the growing discontent among the lesser nobility, who predominate in the officer corps. If the gentry, along with the military, voice strong sentiments for invasion, some noble delegates will begin to think in terms of reducing the number of military voices by sending them off to the Confederation. To conquer worlds where gentry malcontents can be sent to take land for their own."
The exarch who'd interrupted before, surged to his feet. "Your Reverence, I am dismayed! That our Kalif considers only expediency and not principles!"
The man stood visibly shaking with indignation. The Kalif said nothing though, until, deflating, the exarch settled back onto his chair. Then, in a voice dry but not harsh, the Kalif responded. "Alb Riisav, we have rules of order in this College. I appreciate that you're upset, but I will not tolerate another outburst."
He waited a moment before continuing. "As for expediency and principles—They are not incompatible, not mutually exclusive. As necessary, I use expediency in the service of principle."
He paused to examine his audience. It still was not happy with him, but he'd blunted its upset. "Well. Do I have your approval to let this matter be and go on to business held over from our last meeting?"
* * *
He did, of course. And from there went to new business. Without mentioning his visit to the House of SUMBAA, nor any of what he'd learned there. That would wait for a better time.
Thirty-six
The Kalif had been right about the House of Nobles. Two weeks after the killing of Lord Nathiir, the delegates stood substantially where they'd stood before, on him and on invasion. Though
his friends among them were mostly less wholehearted. One of them said that Coso Biilathkamoro, in taking Nathiir's life in front of them, had used up two of his own political seven lives. And in releasing the video record of it to the public, had used up three more of them.
It was Thoga who reported this privately to the Kalif. Thoga still was regarded by the nobles as unfriendly toward him, and simply masking his feelings since the Kalif's recent violence. Thus most of the noble delegates voiced their attitudes and complaints somewhat freely to him.
And significantly, the complaints weren't about his bloody hands, but about manipulating. The source of this attitude, Thoga told him, seemed to come from his release of the video record.
Meanwhile, unchanged was not enough. Straw polls showed him well short of the support needed to finance an invasion. In feet, Jilsomo's latest poll of the exarchs showed five, not four, prepared to vote against it, with two more uncertain. All in all, he seemed to have either twenty-two or twenty-three votes out of the total forty-five, but even twenty-three, a majority, was well short of enough.
Discussion in the Diet had been limited. There were explicit limits to discussion on the floor, except on bills formally proposed. While a proposal automatically required a vote on the bill within a week, and if defeated, it could not be proposed again that year.
With the solicitor imperial, the Kalif and Jilsomo had discussed legal interpretations that might permit constitutional sleight of hand. For routine finances—renewal of the previous year's financing—approval by fifty percent of those voting was enough. If an item was to be increased or its applications significantly altered, approval by sixty percent was needed; for new item or activity, seventy percent.
There were limited exceptions. The Kalif's Contingency Fund could be applied however he saw fit, and increased by up to ten percent if half the Diet approved, or fifteen percent if sixty percent approved.