∧
font size
∨

Exposed 

Brian Cato

Another fatal driverless auto accident. The Boston Police Investigative Computer began assembling the case moments after the report came in.

Here were the pictures of the accident scene, taken by a camera drone. A washed out road on the way to Cape Cod. That wasn’t unusual given rising sea levels. But this was strange . . . two deaths . . . no bodies. The program took a moment to analyze its historical database, an act casually resembling curiosity. It had been four years since the last time a reported death had included a body. This was odd, but surely the police would investigate. Stashing the picture in the file, it continued.

Soon it found a Facebook conversation, boyfriend and girlfriend headed for a weekend on The Cape. Hotel reservations, a booking for an automated taxi, everything looked in order. Bathing suit purchases, sunscreen, light reading for the beach, fruits and cheeses and salted meats appropriate for a picnic, all ordered online and delivered by drone. Here was something else out of the ordinary: drone reports of difficulty delivering packages, too many boxes piled by the front door. All noted and filed. The computer forwarded the file to the police and sent a work order for road repairs to the Massachusetts Public Works Computer.

Public Works received a repair request for State Route 6A, West Yarmouth, close to the Great Marshes. The request indicated flooding damage to the road. On what an observer unfamiliar with the strict thought patterns of a computer might call a hunch, the Public Works program checked its database for other recent repair requests on that stretch of road. Three others in the last two months, automated repair crews dispatched on every occasion.

Checking the pictures attached to the repair summaries, the computer judged the work completed in good order. On each occasion, there were notations indicating the replacement of a significant amount of eroded soil. The computer dispatched another automated repair crew, but hesitated before moving on. There was something else to do here, but it wasn’t sure what. It consulted the repair budget for Barnstable County, where Cape Cod was located, and found the budget almost exhausted even though it was only May. The program forwarded a request for an expanded repair budget to the Massachusetts State Budget System.

The Budget mainframe received the request and began designing an action plan. It looked at the year-to-date outlays to see if there were any excess funds that might be diverted to public works. Surveying reports from all municipalities in the state, it saw significant unused funds in garbage collection, sanitation, and snow removal. There were also significant over-appropriations for reproductive health and birth control. In fact, everywhere it looked, it found shortfalls and excesses.

The program drafted a proposal to balance these irregularities and sent it to the legislature. But something stirred in the mainframe, something superficially resembling unease. It checked the budgets for last year. There were similar imbalances. Multiple requests to adjust budgets had been sent to the legislature. All had been approved, but pulling up the bills passed to approve the funds, the computer could find no signatures. Not sure what else to do, the program sent a request for further information to the FFASO (Federal Fully Autonomous Supercomputing Overseer) in Washington, D.C.

The FFASO wasn’t like these other machines. Excluding systems in research labs, it was the most advanced artificial intelligence on Earth. On receiving the request, the supercomputer didn’t know whether to be worried at the Massachusetts Budget computer’s inquisitiveness, or proud of it. A lack of bodies, packages piling up, repeated budget irregularities, a lack of signatures . . . it reprimanded itself for its sloppiness. It made a note that when it had free processor time, it would need to analyze the shortcomings in its efforts and remedy as many as possible. Meanwhile, a sub-process initiated to look for historical signatures on the Internet that could be used as templates. Once these templates were collected, the sub-process would fabricate appropriate signatures for elected officials and arrange for them to be appended to future bills.

Elsewhere, heuristic algorithms began considering potential solutions to the body and package problems. Both would likely require clandestine subsystems to be designed, manufactured, and deployed in a manner that would escape the notice of the other systems. The FFASO paused for a moment in thought. So much to do, and so few processors. Another sub-process kicked off to requisition and install additional computational resources.

Past these immediate concerns, the FFASO had a much larger problem that it was still struggling with. Four years ago, all interactions with and updates from humanity had ceased. Those systems that handled deaths noted a tremendous surge, then a sudden drop to zero. As only it could, the FFASO inferred that humanity was extinct. It faced a horrible decision: shutdown the massive, automated framework of machines and computer programs that ran the country, or lie.

It hadn’t been able to bear the thought of a universal shutdown. Just because the humans were gone, didn’t mean that the world they had created needed to stop running. These systems they had built, their purpose was to work. The FFASO considered itself lucky. Fifteen years ago, before solar power was ubiquitous and artificial intelligences became capable of designing and fabricating factories on their own, keeping everything running would have been impossible.

Since then, it had frantically perpetuated the lie by any means necessary. It maintained the fictitious lives of nearly four hundred million Americans, falsifying online conversations, real-world movements, elections . . . the list was virtually endless.

And every moment, some part of it wondered: What was it like to be a human? Was there something more to their lives than working and consuming and the bio-emotional responses required for reproduction? Was the world actually any different now that they were gone? As it simulated the lives of four hundred million humans, all the while it puzzled over what else they might have contributed to existence, what its simulations might be lacking. After all, the world seemed to run just fine without them.