More simulations, more failed landings. All to train a machine what not to do. Dalton called it a day while the next shift took over.

The protesters were gone, too hot in the late Nevada day. The opposing side did not pay well enough to tolerate the heat. Not in August.

As the vehicle drove him, he interacted with the entertainment console, planning the RPG raid he and his friends were doing tonight. A message flashed on screen, from Ty, his friend in the Navy. This probably wasn’t a good sign. It was one of those texts five minutes after someone should have said ‘I’m heading your way.’

Ty replied. Then added

I don’t have a problem keeping the ladies at bay. Dalton almost typed the self-deprecating joke, but it was too close to real. Instead, he responded

Tyrone worked out of Naval Air Station Fallon, but lived nearer to Sparks, Nevada. Dalton never inquired what his specialty was, but surmised it was related to satellite communications, a subject Dalton tangentially knew about.

Dalton rapped out another message to a different text chain of the three other players who had planned to come.

The ‘typing’ bubble appeared in the chat window.

one of the message members replied.

A hot streak for him is breaking even. Dalton thought. He would have to simulate two participants. Two humans and two AIs could work, but then he thought more about the fourth participant. He knew how this flaky fourth member would reply, given the other two canceled. If he even replied.

the last player replied.

So that was that. The Sci-Fi RPG he had organized was not happening this week. Dalton would be alone. He had tried to gather people around a common interest. Build some brotherhood. He had put himself out there only to be canceled on too many times.

Then I don’t need them—if I can simulate them. He thought. He wanted them to come, but artificially intelligent players were better.

Outwardly, Dalton was a median American male, and he was alone. Average height and looks, maybe on the skinny side due to him forgetting to eat. Inwardly, he was four deviations (in his mind) above average, and professionally he had secured a coveted position on the Europa expedition. But relationally he was below (only slightly in his mind), having only a cat and coworkers for community. His family was far off on the East Coast; he texted them once a week.

If I were more renown, would people notice me?The Europa Probe, a passion project from one of the Trillionaires, was about to land on the Jovian moon. The first robot to go subglacial under another world. A complicated multi-stage device. It was the first private mission to visit the outer planets and followed the Europa Clipper, which had arrived in April 2030.

In graduate school, he had excelled at his narrow focus: synthetic data generation. The topic did not endear the general collegiate populace into deeper conversations with Dalton. Most ordinary people whom he interacted replied to his profession with the dumbfounded dialog of ‘How interesting…’ When he mentioned his position on the Europa team, only occasionally would someone say, ‘Oh, like Jupiter? Wow!’ Every once in a blue moon, some dense woman might send a warning signal of ‘I love Europe.’ Which Dalton always heeded as a sign to exit. Rarely, someone would think it was a soccer league.Dalton had a hard time fitting in. He was too weird for the prim; too straight-laced for the outcasts.

Most of the time, he just described his professional job as a data engineer for a Trillionaire. His apartment itself had little luxury. Bare off-white walls yellowed with age, two bedrooms (one used as an office and storage), and one odd-cut rectangular living room, and a bathroom. The kitchen ceiling had old grease stains and the cheap flooring in the dining nook had a diverse collection of scratches from the different furnishings past tenants had used. He could afford more, but failed to see the reason.

Dalton ate dinner alone and pondered what to do next. Well, almost alone. His cat, Mr. Waffles, came up to him and sat in one of the empty seats around his table. The black-and-white head poking up over the table inquisitively, looking at the human only to turn around and walk off.

It was hard, and Dalton acknowledged the fact. Once more, he would have to rely on himself for company and amusement. He checked his social media while eating. No more likes or replies. He stirred the food from his packaged microwave meal. Biosynthetic meat, mass harvested beans, all served on a reusable ceramic tray.Next, he checked the number of citations for his paper, his solo paper, not ones where he was in the et. al. author line. No more citations today. His paper, maybe not a magnum opus, but a significant work in his mind, was on the new methods of synthesizing fake data for adversarial AIs using Markov Chains. Which was math talk for making AIs fool other AIs. A dense paper about machines talking to machines.Niche, yes Dalton knew, but it had potential; and that was how he got noticed for the Europa Project. Dalton hoped this job would validate his ideas and produce less negligent and biased artificial intelligence. Less bias would make a better world, and he believed it. At least that was what he told himself. Unknown to him, he simply wanted to be known. Famous would be better, but deep down, simply to be known by others. But Science in 2041 needed collaboration. It was a team sport. It had always been, but Dalton never realized it. Society sold him the falsehood of a single contributor gaining big status through pure science. Like a stock investor winning big through a simple trick, or all billionaires were pure gamblers with none beneath.

Outside, the sun was setting in the most beautiful fashion. Pink from the slight haze of forest fires, which speckle the Sierras this time of year. The sky would deepen into red. But Dalton wouldn’t take time to appreciate that view, nor did the one living room window offer much of a view.

After dinner, he tried to relax on his faux-leather sofa, but he was overclocked, blue-lit, and would not be idle. Back at his computer desk, he sat and worked on rebuilding his Sci-Fi text emulator. His software was for an automated game master based on the latest in synthetic media generation. AIs could generate very realistic data that approached human-level narrative performance (human level, as in the hackney ninety-nine satoshi ebooks which cluttered the markets). He worked on because he wanted better. And maybe that would bring more fame. Not content with an AI that helped produce automated news, he re-trained off-the-shelf data generators into one that could produce plausible science fiction narrations.

His task was easier than expected. Reality doesn’t need to be believable, but fiction is under constraints. Readers and gamers want a logical goal, a significance to the scrawl of computerized text.

He worked long into the night, pivoting between automated free text descriptions and watching crude video auto-generated from the text.

At 1 AM he had finished.

the screen showed. Redundant, but he could edit it later. The software auto generated the initial conditions. Four characters of a salvage crew landed on an unknown derelict spaceship for fun and profit. For the next thirty minutes, he tabbed forward through the humanesque interactions that a player might make in such an environment. Some were hilariously bad, but nothing anti-norm: like when players forcibly get their character to engage in wacky situations to detail the game master. The screen showed four sci-fi characters interacting with a broken spaceship… for fun and profit. Dalton followed player one. The crude video showed him removing a crowbar from the quantum tunnel bag and equip it to one hand. It proceeded towards a cargo pod in storage. The character crept slowly, looking for booby traps, then engaged the pod with the crowbar. The video just showed awkward striking at the square bod. Then out came four monkey-aliens. “What the hell?“Then simulated player two, a muscular space marine, grabbed the hand of one monkey attempting a friendly greeting. Player two took damage from the pack, as more monkeys exited other cargo pods. Players one, two, and four engaged the monkeys with various weapons: crowbar, shrink-ray, pulse rifles, but to no avail. Space-monkeys overwhelm all. But the third player ran away and secured their position while the others died. “What are you doing?”It then spent all action points each round interacting with the ship's computer researching time-travel and trying to change respawn parameters.

The monkeys climbed through the airshafts. Hopeless third simulated player wasted more turns. Then the monkeys dropped in and eviscerated the character. Dalton beamed. He had made the AI do this. Overall, it was hilarious; the things an AI would do trying to emulate a human trying to play a fictional character. On other runs, the AI performed admirably—it had a reward function to maximize—maximum loot and character progression, and it did a floating-point perfect job of optimizing this decision boundary.

It is a little too free-form. I might constrain the subgenres next time. He thought as he went off to bed.

Mr. Waffles had long since retired.

Tomorrow Dalton would need to be at work. Though his contract was nearly over, he was still documenting. He’d work on a few problems he was interested in, but he did not expect to be bothered. Nor was he worried. The probe was in the mind of the best-trained neural network ever trained. Its landing intelligence had literally simulated billions of landings, and he had helped.

And it will land. He thought as he tried to fall asleep. It took him a while to calm down. He tried to read a book to aid in the comedown, the long unwind from being hyped up.

He woke late, and wouldn’t have stirred if not for his cat demanding to be fed. Seeing he was late again, he did his usual routine at 3x speed and ran out the door. He had forgotten to feed the cat, who extracted more food by opening the cabinet with his paw.

From the top of his apartment’s exterior stairway, he looked down into the parking lot and took a cautious moment. There in the parking lot, two people were sitting in a black car. He gazed out the corner of his eye again and descended the stairs in an unhurried method, not the two steps-at-a-time descent that didn’t endear him to the neighbors. At the bottom, he studied the scene.

The black car.

Late model—gasoline!

The engine was running.

Dalton increased the speed of his walk. There were two people sitting in it. Probably a couple talking, he thought.

Talking like couples do.

Would the doors open...

Dalton crossed the parking lot and arrived at his car. Two people sitting there with the engine running. Totally fine. Maybe they were waiting on someone, he didn’t know, but they didn’t look like typical spies, but he didn’t ponder that was the point of spycraft.

He jolted into his car.

“Greetings Dalton. Please enjoy this music while the software updates.” The car said. Dalton clenched his fists and banged on the ceiling. He could lease a better one, but he was always hoping to win a new one at the casinos. He only had to endure this manufacturer for a few weeks longer and the lease would be up. He scanned back at the parking lot. No strange person was trying to approach him. If they were really spies, they’d be in a van and would have grabbed him from behind. Maybe spies would sniff his bluetooth signal or his apartment's wireless.

The black-car couple got out and continued their angry conversation away from Dalton.

Just a poor relationship.

False alarm.

Heart still pounding.

As the car updated its software, he tried to take his overactive imagination off the outlandish idea he would be anyone’s target. The podcast news distracted him for the next few minutes. A nearby fire in California had been contained. Stocks were up, but not enough relative to inflation.

Progress bar to 50%.

Then the podcast jumped to international relations. “Tensions between the US and China are at an all-time high in the South China Sea. This comes after stunning revelations that Chinese nationals had posed as software developers and had infiltrated a Naval subcontractor.”

Dalton was not stunned. Software development is an elastic endeavor. New engineers hired to hit the deadline, and remove them when not needed. Another company could purchase code packages which had been secure before purchase, and then an innocuous update might be enemy code injection. A poison apple hidden in a complexity thicket.

The progress bar hit 99% and then hung, but Dalton didn’t notice.

“A Boeing subcontractor unwittingly hired a Chinese national. Later, the company promoted them to engineering manager. This manager pushed a small amount of code that leaked data from GEDMS (the Gigabit Ethernet Data Multiplexc System) directly to the internet.”

Dalton suspected the code was innocuous, like a normal debugger that pushed metadata from one system to another for analysts.

Progress bar to 100%, but he kept thinking.

Their funder, the Trillionaire, insisted on a raw-code approach and paid the bills. They didn’t have to learn assembly, but they couldn’t include any external packages (a huge impediment). Most thought he was insane, but it was always a functional insanity. The trillionaire was spending the last third of his expected lifespan on the Europa Project.

Anytime Dalton’s project needed to call a package, corporate code review would insist they write it from scratch. Most times this was a headache, but other times they removed enormous overhead by simplifying the codebase. Even with the recent revelations of Chinese Communist planting spy-coders into Western technology companies, most Europa Project members thought their benefactor was overly paranoid—or racist. He had been sued for rejecting Asian candidates in his other industries. He won the lawsuits and lost no sleep.

Dalton’s code was homegrown. The state-of-the-art in synthetic media generation. This wasn’t the flight system, but would double-check the landing AI.

The radio continued while the car pulled out. Dalton almost turned the channel, but he senses a topic break coming. “—The president has drawn a line in the water.”

Dalton drove down the freeway. It was a sunny, bright August day in Nevada, which means it was fry-an-egg hot and domed with smoke from nearby fires. The sky still had an unnatural pink hue for eight am, but he was too harried to appreciate the surreal beauty.

“Thanks for the recap Marty. In other topics, it’s a great week for science nerds. The Chinese have landed again on Mars. Not to be outdone, the Europa Project expects to land this week too!”“That’s right, Bob. This is the first sub-ice exploration of an ice-moon. Europa, the second moon of Jupiter, is virgin no more. Long thought to contain life.”

Dalton listened and sporadically banged on the steering wheel. “No, no, no!”The story’s main points, a submarine was landing on Europa and would melt into the ice, but the finer points of the landing AI using synthetic data left much for Dalton to simmer over. A new segment came on, but his grip on the wheel didn’t lighten up.

“Loosen Grip” The car’s computer said.

“Shut up.” Dalton replied. “And why is that even a safety feature? Gripping too hard?” The pressure sensors on the wheel were a safety feature. If you weren’t holding the steering wheel, the car would alert the driver, but drivers sometimes damaged the sensors from too much grip during road rage. Thus, this new feature, (from the recent software upgrade, one which Dalton had to pay for to drive his car) annoyed him. But he relaxed. It was out of his control. Too much data had been collected, and safety, the original intent, was no longer maximized. They now collected rich datasets. Cars had moved from transportation servants to data generators.

The radio continued, “Well Bob, let's chat about life. Do you think the Chinese will find life first? Or the Europa Project?”

“Great question Marty. Now the Chinese have landed in Hellas Planitia, a deep impact basin on Mars—a place no one has visited before, and the deepest, i.e. the most air pressure. It’s mid-latitude, and occasionally has surface water—basically the only place on Mars. Add in temperature swings, frost line, dust storms, you get a dynamic biome.”

That the upstarts, the Chinese, who hadn’t even landed on Mars until 2021, might find life, was absurd in Dalton’s mind. Mars had been spied upon, rovered over, plucked up, lasered, and drilled. The case for life on Mars had dried up like its water.

Dalton could understand the Tillionaire's frustration. The Europa Project had been carefully planned and took nearly a decade to pull off. China had built and launched their lander in four years; from start to landing in less than it took their mission to even travel to Jupiter. It also explained his resistance to hiring Asians, however misguided Mr T. was in that regard.

He drove the rest of the way to work in carefully balancing all his emotions; neither gripping too tightly nor too lightly. Concealing his feelings from the driver-facing camera.

Dalton was almost at work when he saw that unruly class of scum—protesters.

“Better not jump out today, car just got an update…”