The Europa Project’s next practice runs went well, but the terrain was easier. Failure or success; it did not really matter. All data was good data for their AI. At night, the cloud computers would ‘dream’ a solution. Thousands of simulations would run. The delta between the lander’s AI and the copy in the cloud would be transmitted up and out to Jupiter. Each new revision pushed the landing closer to perfection.

Most thought the idea crazy, but it was what the Trillionaire outlined, and he was funding this project, so the workers continued. The orbit would tighten around Europa, then touchdown at fourteen hundred kilometers a second. There was no rewind, no respawn. No way to attach more boosters or fuel for a soft landing. The software was all they could change. But software changes the world.

It was late afternoon as Dalton ran the numbers again. As it computed, he left his cubicle to get some afternoon coffee.

He approached the kitchen and took as wide of an angle as possible to prevent running into anyone. The area was poorly designed, as most office cocina are. A single-wide door opened into a hallway with two blind approaches, each of which equally trafficked. A hallway a ninja couldn’t stop from spilling from the cheap open-topped paper cups.

Dalton refilled his cup around fifteen percent before the sputtering of the pump sounded the dreaded call. He’d have to wait. So he remade the pot and stood, trying to focus on work. Anything to be productive—functional. He thought time was scarce and often walked quickly to save time; now he was forced to wait. The coffee dripped, the clock ticked, the plastic tile floor let an occasional squeak as Dalton shuffled. He heard someone approaching and backed away from the door, and spilled nothing from his near empty cup.

The entrant saw the predicament and also adjusted. “Wha—Oh. Sorry Dalton. Haha. I keep forgetting about this doorway.” Jim said. “You’d think a building full of engineers and scientists could rig something so we’d see around this corner.”

“Yeah, you’d think…” Dalton said, not wanting to feed the gregarious man anything, but wanting to show his knowledge of the mission. “Like… I dunno, a radar that penetrates solids, but reflects off water.”

“Haha, I like it!” Jim said, pointing his big finger in Dalton’s direction. “Hey, are you enjoying the new data generated? Not much time left to mutate it.” Jim looked at the coffee machine that was still brewing. “BRB, see you when the coffee’s done.”

What an inconsiderate prick, Dalton thought as Jim exited into the hallway. He almost smashed this fractional cup of coffee all over my shirt.

It was true, Jim was always in a hurry. On another occasion, Jim had bumped someone’s coffee into a mess. A tall man, around six two, Jim had a long gait with only two speeds in life: going somewhere, and sleep (where he was dreaming of going somewhere). The farthest Dalton got was following behind Jim’s titanic wake through the scientific community.

Science these days was a team sport, but getting hired for data cleansing and maintenance was like getting picked last in middle-school kickball.

There he goes again. Jim had transformed himself from a standard science geek into a leader. He had it all. It didn't sit right with Dalton. Jim’s success appeared effortless.

A distance feeling in Dalton knew it was not only luck, it was decades of consistent hard work. But the probe’s launch and landing was a perfect time for Jim. Not too old to lack the energy for the workload, but old enough to be taken seriously and have friends in high places. Maybe in another twenty years, there’d be a new mission for Dalton to join, probably to Titan. But chances to explore the outer solar system came less than once in a decade.

Dalton walked back to his half cubicle. He shared the split pod with a Spaniard and a Nigerian. They were talking about their home countries or some other crap that Dalton tried to tune out, but his ears perked up when the Spaniard talked about the Deep Space Network’s dish near Madrid.

“So you’ve been to the Madrid DNS?” The Nigerian asked.

“Yes, many times, I used to go to Robledo for field trips.”

“How close is it to Madrid?”

“Less than an hour, without traffic, of course, but it inspired me to take up astronomy.”

“What’s it like? Could you see it from your house?”

“Ha, the surrounding land is beautiful, and the white dishes add a focal point to the landscape.” The Spaniard gave a chef’s kiss.

Dalton wanted to interject, adding what he knew about the Madrid complex. How the United States government built it for the world’s only (long-term) successful fascist, Francisco Franco, as a reward for his anti-communist stance and alliance. Eventually he gave up trying to discover a conversational crack he could inject himself. Everyone else seemed so glib, their conversations so effortless. Then put his headphones on and went to work, not wanting to be the “Well, ackchyually,” guy.

Dalton clicked on a stream for ten hours of coding music.

Dalton thought the Spaniard might defend Franco, ‘a good man for the time—a little excessive, sure—but an authoritarian, not a totalitarian.’ Dalton didn’t have time for that. Everyone likes to whitewash their history. Hell, even Europa erased itself, Dalton thought. That was a good one. He congratulated himself on his joke. Scientists knew since at least Voyager 2 that the surface of Europa was very young and subject to constant resurfacing.

The chillstep bopped through his headphones. Results time!

The core of D’s job was to process the output of the simulated landing runs from the control room. Then take the data and perturb it; either distorting the landing zones’ digital elevation models (DEM) or inserting failures that the landing AI would have to account for. They even allowed him to send bad data. Non-gaussian, non random. Deliberate distortions. Then they would see how well AI’s skeptic co-network would react to bad data (most of the time it tried turning the instrument off and back on again). Similar to how the US military hardened their drone AIs against GPS spoofing.

The human operators served as the baseline for comparison, but the landing AI outperformed humans. It was no surprise. The AI’s life depended on landing correctly and practiced millions more landings.

Dalton had constraints, every orbit he could change the variables less and less. Each orbit brought back better ground truth, and more constraints on what Dalton could do. If the AI saw results outside the estimated truth, the skeptic co-AI would outright reject it as bad data. Dalton wanted the data to be realistic, but difficult to land. Also, he had only three locations he could work with. Each one had a recent (geologically speaking) eruption of brine which resulted in the formation of even plains. Glassy smooth icy salt-flats. Nature’s runways. The Europan streak did not vary more than a meter over several kilometers of length, but Dalton tried his best.

Dalton whacked away at the keyboard, creating a software that tweaked the parameters of each landing location to cause the onboard AI to crash into the surface of the jovian moon. He was their chaos monkey; someone designated to introduce faults that the software to deal with. Random bad data wasn’t good enough anymore. He had to use slippery-slope falsehoods. Chain subtle errors and lead the AI into a crash.

But these landing zones were almost impossibly flat. The first area, landing zone Maria Celeste, was first discovered by the Galileo probe’s solid state imager (SSI). Originally it had the uninteresting designation of image 5452r. Taken in the early 1990s, it sat unresearched until years later. Then two decades later, a French paper 2020 estimated its overall flatness varied by less than 5 meters over several square kilometers of area. This inspired the Internet to goad a Trillionaire to attempt a horizontal landing… of a submarine… on Europa.

And Maria Celeste was only the backup target. Europa Clipper discovered an ever better zone, codename Vincenzo. The Clipper watched this zone form during the preceding year. A burst of high pressure saline water spread out along a linea, filling it in with near perfect flatness, leaving a five kilometer straightaway with a one meter variance in elevation. It was a prime landing spot, regardless of the unorthodox landing approach.

Someone tapped on his shoulder, causing Dalton to jump. “Ah, what’ta ya want,” he said and removed his headphones.

“Sorry, Mr. Chatsworth, but it’s time for your code review.” A middle-aged man in a white shirt, jeans, and an ID lanyard came up to him. The dangling badge displayed the dreaded Corporate Code Security.

Dalton sighed, saved his work, and got up. “Again?” These had become a weekly requirement.

“Yup. We want you to meet in the Titan conference room.” The man said.

“Fine. But let me use the men’s room first.” Dalton said. Which was only a delaying tactic. His bladder was fine, but his nerves were out of whack. He had been in the flow and the CCS goons snapped it. CCS was the acronym for Corporate Code Security, but everyone used the false acronyms: ‘Catching Chinese Spies.’ or ‘China Code Search.’

The meeting was long and boring. The fluorescent lights were off and the cheap white plastic blinds pulled, but still allowed slant rays of sunlight. They scrolled through dark-mode code for ages, looking over his last commitment on the wallscreen. The code reflected off the glossy surface of the dark faux wood conference table. Three men sat on one side of the table while Dalton was opposite from them. The man in the middle, an older guy also in a white shirt, did all the talking. Two others flanked him, one a software engineer and another the member of the CCS team. They eventually found a single process that made an external library call.

“So, this package here? What is it trying to do?” One of the security team asked.

Dalton was surprised. He had included an external library without thinking. It was just so common to bang out that he hadn’t noticed it. It was as natural as walking or riding a bike. “Oh, that?” Dalton shifted his body language. “Well, this is a very simple statistical package. Commonly used to generate stats distributions; Normal, Gaussian, Chi-Squared, Cauchy, you know. All the distributions.”

“Simple? Great, so make the change.” The center-man said.

“—B-but this code only creates data. None of the packages get pushed to the probe.” It was true. His code created the data, which trained the AI, but there was no direct connection. Probabilistic programming. “There’s a discontinuity between my code and the machine.” He tried once more to describe his work to the obstinate security team. He looked at the men, who gave no outward response to his defense.

“I understand.” The man-in-the-middle asked. Whether he did or didn’t was irrelevant. “But we’re going 100% home-grown code here. So you’re required to make the change this afternoon.“

Dalton threw up his hands, and like all good coders do, he ran a mental cost-benefit analysis of arguing with people versus re-coding. He had already made replacement functions in his own library, but had neglected to call it. It would be a simple change, but he hated the inanity of it all. They could have done the code reviews automatically. This was an old fashion sniff-out. Social engineering. Get the programmer to sweat.

“Yes. I can do it.” Dalton said, putting on a fake smile.

One of the other men, to the left of the man-in-the-middle, spoke up after looking at his phone. “It says here he can approve his own code. Looks like that’s a flaw. We’ll get that changed soon.”

“Sounds good.” Said the man-in-the-middle, addressing a left-side-man. “Dalton. All your other code looks acceptable. Make that one change and we’re good.”

Dalton crossed his arms and exhaled. “Ok. is that all?”

“Yup, you’re free to go.” Middle-man said with a smile.

Free to go? The nerve of those guys. They are not jailers.

He needed a break. As he walked to the kitchen, he heard some quiet chatter about Jim and the mission. Then he walked slower—on his toes — to eavesdrop; new denial-of-service attacks. He put some water in his cup and left the building.

Outside, the heat was oppressive. It was three pm local time in Reno. Tenth of August. He walked over to the shade. Sure, it was less hot, but not much. The difference in the shade is that you’ll still get heat stroke, but much slower. Someone might rescue you if you keeled over. The wind blew, but it was more like a hairdryer going off near you than a refreshing breeze.

No one else was outside. Even the protesters had disappeared like a hundred-dollar bill on a Reno sidewalk.

He looked at the cars. Some of them were nice. New Teslas, a Subaru, and a few trucks built for the off-road. Most were nicer than his car. He saw a bumper sticker on one vehicle. Georgia Tech. That was Jim’s. For the next minute, he imagined keying his vehicle. Casually walking by and scraping that paint off with the sharp end of his apartment key. What joy and excitement that ran through him at such a petty act!

Except Jim was still driving a beat-up ‘19 Civic with multitudes of scratches over it. Dalton rubbed his chin.

Why does a guy, a trust fund child from the privileged Northeast, still drive such a crappy car? He wondered. What kind of statement is he trying to make? That he’s one of the common people? The Everyman?

Jim was nothing of the sort, no matter what image he wore in public. He had everything: great undergrad experience, all-star Ph.D. advisors, even a genetic pedigree in the sciences.

Not like me. Nope, I’ve had to work for it.

His own mother seemed concerned her son took a low-paying position compared to the costs invested. One of his cousins said he was wasting his money, spending over $400k on college degrees (and that was cheap). Dalton could have invested that money somewhere else and got a higher return, like buying some desert land and erecting cheap tract homes on the Nevada-Cali line.

But the investments in the sciences paid off, eventually. Dalton had worked his way up the ladder by shrewdly building on prominent people’s research instead of trying to falsify it; not too revolutionary to invoke their jealousy, but still good enough to flatter those who pulled the strings and threw out names.

This strategy landed a gig because two of the Trillionaires made a public bet and would pay big. Those with so many the zeros in their bank accounts, it was easier to compare in scientific notation. The project managers of the epic wealthy contacted the colleges and the people on the papers. Names named other names, and eventually the job and money trickled down to him.

On the drive home, he thought about the high magnitude absurdity of the entire operation. They were going to land a multi-ton submarine on its wheels on a preternaturally flat section of an ice moon five AU away. Then the brakes on the ten wheels would cancel out the fourteen hundred kilometers an hour forward velocity in under four kilometers of ‘flat land.’ It was nuts, and he was a part of it.

The second richest man bet half his wealth (to his own charities, of course) that the richest person would fail. So did many others unaffiliated with either trillionaire.

As he drove home on the 445 up to his decaying apartment in Spanish Springs, he thought about what the failure of the entire mission would accomplish.

It would probably sink Jim’s career at the very least. It would protect the naysayers—the few no-life-on-Europa contrarians—from having to face the data. Those armchair theorists could explain away the orbital observations of life as a ‘peculiar geochemistry’ from Jupiter’s massive radiation bands. None would ever attempt a project of this magnitude. Dalton’s career would be fairly safe, though. He was young and would see many chances in academia, or get a job in the corporate world. His science career had stalled, anyhow.

“Denial of Service attacks continue on popular cloud infrastructure,“ a radio announcer said. Synthetic news anchor. There was always such a crispness to the artificial readers. No slur or draw.

“Meanwhile, outages in China have affected grain elevators in three Chinese provinces.”

The information war was becoming as hot as a Nevada summer.

He opened the door to his apartment, which he shared with no one (much easier to keep clean that way). As he entered, his cat came to demand his attention with a meow.

“Hello Mr. Waffles.” Dalton made sure the cat had water and food, but then ignored him.

As he sat alone, a median American male, something came over his mind: a devilish thought. He could take good data and bad outcomes, and mix the two and poisoned training data. He could re-code his data to slide the machine to its doom.

A slow trickle of bad data to carry it down to a waterfall of failure.

I could crash the probe with no one the wiser. Dalton got up suddenly, alarming the cat, and went to his desk to test the theory. Mr. Waffles was not pleased. At least on the couch, the cat could slowly invade his space. Nowhere around the desk was comfortable enough.

The cat meowed again.

“Go away cat.”

But Mr. Waffles would not be made a third-class citizen behind Dalton and a glowing electric box. The cat jumped up on the desk near the keyboard.

“Cat, get off.”

He picked up the cat and moved to the couch, and sat down.

Only one more layer of abstraction would be needed. The police catch the counterfeiters because the Treasury teaches good from bad. But imagine the criminal is in the Treasury?

The thought never quite left his mind and his subconscious. Casually at first, as if engaging in a thought experiment. He outlined the details of how to crash the probe.