Dalton returned to his console and watched the progression of events with Marco. They celebrated with a solid, firm handshake that went into a full hug between men, celebrating an accomplishment. Other cheers and continued celebration rang out from the office.

“The next steps should be simple.” Marco said.

Dalton crossed his fingers in both hands and held them up. “Should…”

They bought up any data. It had been more than an hour since the probe had landed. Thirty-seven light-minutes to Jupiter and all systems were nominal. Dalton and Marco had access to all the internal data, which was not very much. They already had low latency to the few downloaded kilobytes. All streamed in from the Deep Space Network dish in Goldstone, California. Each bit of good data plucked out of a field of view splotched with a million false signals—constellations of satellites pushing data through low earth orbit, radio bursts from Jupiter’s monster lightning, or comic energies from the beginning of time—all noise competing for a listener. Every bit of data was error-corrected, and every packet checksummed.

The office held the compiled copies of the data. Each of them looked at the data with low latency, but it pertained only to system health. Marco checked surface radiation, while Dalton was engrossed with counting the number of dropped frames from the low-quality video feed data. The project was a public-private partnership, with private developers supplying the spaceship bus, and only the NASA funded instruments had public release requirements (and none were live at landing). They watched the soon-to-be-explosive drama, The Blowoff, in which explosive bolts severed the fan-brake and the wheels to leave only the submarine and the communications bus flat on the ice.System check was next. Dalton and Marco listened into the command center feed.

“Comms” – “Go”

“Top Reel” – “Go”

“Bottom Reel” – “Go”

“SubSystems Sound-off,”

“Reactor” – “98”

“Guidance” – “Go”

“Impellers” – “Go”

“Lights” – “Go”

“Video” – “Thirty-On”

“Sonar” – “…”

“Sonar?” Jim asked.

“No Go.”

“Run that by me again.” Jim said, standing at the lead desk and pointed to the operator at the sonar console.

“Sonar has detected eight pings returned for zero sent.” The lady said.

“Sonar, recheck the data and reconfirm.” Jim said.

Another operator came up and chatted with Jim. Then an engineer came up.

“Sonar has confirmed: eight received, zero sent.”

“Prep for blow off. Get it under ice.” Jim called out. “Sonar troubleshoot.”

Dalton and Marco looked at themselves, mouths agape. They were not part of the landing team, but as data engineers, they were familiar with the data feeds from all instruments and immediately set to work pulling the sonar data.

“Okay, we knew bad data would happen. We’re on a moon bathed in Jovian Radiation. I’m pulling up sonar…” Dalton said. “We only have the metadata: that the AI detected eight things.”

“Max amplitude and timestamp. Eight rows of data,” Marco said. “A sound emission from the ship itself wouldn’t look like this. It would be loud and quicker. And there’s amplitude.”

“Nor would it show four sets of two pings.” Dalton said and showed a quick chart he built: ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓

Jim came back on their livestream. His hand was covering his microphone as he talked to two gray-haired men. They were the lead engineers of the landing system. “Okay, Engineering has an idea that the fan-brake is picking up natural pulses and amplifying them.”

“Keep the wheels on and blow the fan-brake.” They had done their job and were no longer needed. Engineering grasped for theories and concluded the pings were from some portion of the metal chain undergoing deformation as it refroze.

The backmost device on the probe was the fan-brake, a large hydrophilic triangle designed to grab and hold any bit of ice kicked up by the landing and snowball the mass. It was the primary method to cancel out the forward velocity.

“Blowing the chain. Confirmation in seventy-four minutes.”

The braking chain was the rearmost portion of the landed vehicle. The next section was the stowed communication dish. Right now they were using the Europa Clipper to pass signals, but that wouldn’t always be in sight. The Comm module also contained a small radioisotope generator, which fed power to itself and the next second section, the Sub-Ice Disk. The SID would follow the sub down through the melt hole and onto the underside of the Europan ice-shell, however deep the ice might be. It contained few scientific instruments and was primarily responsible for sub-to-surface communications.

The SID module had a small ring that would inflate, securing it to the underside of the ice. A long fiber-optic cable ran from the Comm Module to the SID and fed both data and power to it. From there, the SID would again unspool a kilometer of line to a microphone. This would enable a longer communication with the free swimming sub via acoustic transmission using the JANUS protocol to transmit data underwater. At best, it could transmit up to twenty-eight kilometers on Earth's oceans.

Janus was the Roman god of traveling, trading, and shipping. Also, the Romans thought this god presided over the beginning and ending of conflict, but Dalton didn’t know this, nor did he think about the U.S.- China conflict.

After a few minutes, Jim came over to Marco, Dalton, and Emmanual the Nigerian, who was not there, but had been messaged to come in.

“Hello! How’s my fake data team? Need you to check the signals the Sonar team detected. Test what the Submersible AI would do if we turned it autonomous.” Jim said. “We’ve blown the chain and Sonar’s running more tests.” He knocked goodbye on their cubicle wall and then fast-walked to his office with his assistant. Dalton watched them leave. Right before the assistant closed the blinds, he saw Jim pick up his phone.

Dalton hadn’t been expecting they’d be helpful until the sub melted through. His team began by taking the local copies of the SubAI and simulating that it had melted through and deployed Comms and SubIce modules. They locked it in the SubIce module’s location as 0,0,0 for the SubAI's inertial frame of reference.

Emmanuel, the Nigerian, came in and they brought him up to speed. “Emman. Sonar detected a signal. We’re going to feed data to our local AIs and see what happens.”

They played back the same sound (which was heavily distorted having traveled through the lander’s wheels or brake-chain.)

“When did it hear this sound?” Emman asked.

“Soon after stopping.” Marco said.

“How? There’s no atmosphere.”

“Through the wheels or chain. They turned sonar on for melt-prep.” Dalton replied. The lander had eight titanium wheels similar to the original moon rover, except far larger and stronger.

The team got to work and followed what their AI copies did with the strange signal.

In autonomous mode, the SubAI had a reward function to maximize ‘science.’ This, of course, was a vague term and no two scientists could agree on priorities. All were attentive in advancing their own goals. Most of the junior researchers were interested in writing papers on anything about Europan geology with the goal of getting those limited tenure spots. Most of the older ones that were already tenured were more interested in preserving their legacy and defending their past papers. They had enabled all scientists to quadratically vote on the ultimate reward, the ‘sci-points,’ that the AI could score.

Each scientist got about ten votes (some principal scientists got 16, while other juniors 5 since some scientists were more equal than others). They could increase the reward for each goal by one-for-one vote. So a geologist might only want the sub to retrieve a rock from the seafloor and bring it back up. If he voted on this once, he would have 9 votes left to assign on other topics. But what is a geologist without their rocks? One vote would cost one point, but two votes would cost four total points, then three votes on the same target would cost nine points, and finally four votes would cost sixteen, the scientific equivalent of putting all your eggs into one payload. Nearly everyone spread their votes out for a mixture of goals. They had already uploaded these to SubAI. On the return journey, they could also vote to move each piece of data through the upload queue, either with the same method, but could trade votes for bandwidth (this was all net of system status data, which had first priority).

Dalton and team local AI classify the sound. It could not be explained by natural phenomena such as sea-ice or tectonic sources. “Multiple-Asteroid-Strike-On-Ice was the highest probability.” Dalton said.

The AI could do combinatorial classification. The Trillionaire’s retailer developed this method to decompose someone’s picture into multiple fashion products and sell the entire ‘look.’ Their retail AI could perturbate the elements of style into a new ensemble or customize it for the customer. Later it was scaled up to coordinate home furnishings, and even a custom-built house designed interior-out.

“Makes sense. A mini-Shoemaker-Levy hitting Europa.” Marco said.

“—But the probe is not confident. It only gave 37%." Dalton said. But every confidence score by an expert system is irrelevant without a baseline, so they ran it through another set of sounds to classify. They sent whale songs, dolphin pings, sub sounds and each one came back confidently. Then they sent noise, and again the sub was ‘confident’ about its classification.

“Alright, so now let’s give it the original data and change the labels. Give it SciFi-FantasyNet for labels. What does it score now for classification?”

“Merman—high probability.” Marco said.

A chuckle came from everyone. Trying to suppress his smile, Emmanuel scolded Dalton, “What a sexist dataset you have there, Dalton. You didn’t de-bias the data?” Again, everyone laughed, and they all breathed deeply. It was a respite from the stress of the morning.

But then Dalton took a second and said, “Wait a minute… why didn’t it say ‘mermaid’ vs ‘merman?’ I’d have thought mermaids would be overrepresented in the fantasy data.”

“It’s just a baseline. Why does it matter?” Emmanuel asked.

“We don’t have the time to figure that out.” Marco cut in.

After a few minutes, the copy-AI on Earth gave a small amount of attention to the sonar data, with the skeptic-coAI classifying it as bad data and ignoring the sound. The Europan AI determined an easy low-effort science goal to accomplish was to investigate the asteroid impact that it presumed was nearby. If a tepid search produced no more evidence, then it would ignore the data and carry on.

“The Panspermist are going to be happy.” Dalton said as they both watched the simulated AI float upwards from the SubIce Module and investigate the ice it had just left. It repeatedly sent sonar pings into the ice, both for orientation but also to detect masses of rock within the ice flows. A high concentration of rock might allude to the quantity of impacts that Europa had experienced. Estimates for inter-system asteroid exchanges. A high number of asteroids (from which Europa regularly wiped off the surface) would mean that if there was life anywhere in the solar system, then life would be anywhere hospitable—Panspermia.

This was a long-standing idea with hard facts supporting it. Asteroids from other planets fall on other worlds, and extremophile bacteria can survive long trips hibernating inside the rocks, but so far no asteroid could definitely be said to have had Earth and hit Mars, or Martian life and hit Earth. But those who wanted to study this phenomenon had put enough votes that the subAI investigated.

“Okay, so are we good?” Marco asked.“Yup, just a blip. It’ll move on.” Dalton said.

Jim left his office, took a central position in the room and called out for all to hear: “Everyone that was our sponsor.” (The Trillionaire) “I can’t give specifics, but Tory here is asking you to do an emergency pack-up. She’ll give you details.”

His assistant, a well dressed black-haired woman in a long blue skirt and white blouse, stood on a chair and yelled. Her ID lanyard flopped on her side as she moved around.

Dalton looked at his teammates in disbelief. The landing was over, the melt in process. Now was not a time for rush.

Jim came over to the AI group and consulted with a few other people before coming to the trio of the synthetic data team. They did not wipe their faces of surprise when he came over.

“So, what’ll the AI do?” Jim said, drowning out the sound of his assistant behind him, who was dutifully calling out instructions. The team had a hard time tracking what they should concentrate on.

“So?” He asked again. Three heads snapped to his attention.

“Yeah, after the melt through, it follows the sounds and then (if needed) returns to the sub-ice module to communicate. Afterwards, it will continue the mission and ignore the sound.”

“Okay, that sounds good. Pack up.”

“Pack up for what?” Marco said. This was beyond odd, all thought.

“The Trillionaire has a hunch and wants us prepared.” Jim said. He left with his customary quick-walk to another team.

The trio now tried to jump back to what Tory was telling them, while looking around at what their office-mates were doing to piece together the missing instructions.

No one lands a probe and sits there. Dalton thought. He was incorrect, though. After the landing on Apollo 11, it took about six and a half hours before they started lunar operations. The Mars landing also took nearly five hours between landing and the first step. Is Mr. Trillion even interested in the science? But Dalton didn’t have time to ponder.

“Okay. The bus will arrive shortly. Fifteen minutes we’ll leave with or without you. If you rode a classic bike, bring that. Bring any extra clothes.” Tory’s sharp and focused voice called out. She got down, answered no questions publicly, but continued to talk individually to others. “If you have children, or others to care for. Please leave immediately and handle the affairs. Take this flier and meet us there if you can. Data may be affected, so work-from-home might not be an option.”

Most people crowded around Tory. A few left to call their neighbors or family to take care of their people or animals for a few days longer. Most of the operators on the landing mission were dedicated to the mission and its strange on-demand. Everyone already had plans for the grandparents to watch the children, or neighbors to walk and feed the pooch. The calls were indeed strange, but the supportive friends and relatives nonetheless agreed.

Dalton thought nothing of his cat, which would be alone for several days, but the enterprising feline would be fine.

“What is going on?” Marco asked.

“And why the bike…?” Emman asked.

Dalton pulled up the news on his phone to check on US-Sino relations. Nothing new. More beating on the war drums by those who had never seen combat. “Impending cyberwarfare strike?” He said. “I guess it’d be better to take us offline than allow China to hijack the probe.”

“Well, that’d make sense… They could also hack our cars, keeping us from getting in.” Marco said.

“But why the batteries…” Eman added.

###

The probe sat on the surface of Europa for less than an hour. It had objectives to accomplish. It had a long list of aspirations to investigate and rewards to pursue. Designed to operate with no direct human intervention, it had ‘science’ to do, though it did not know what that entailed. It was there to record data and a series of exploration / exploitation trade-offs coded. The probe could do very little new science on the surface. It needed to get under the ice, to the undersea world that had never seen light.