#1 Wave Navigation 101
Moebius is serialized fiction about an indie private detective hired to solve crimes related to technology. This is the first edition.
Published fortnightly, Subscribe to get it in your inbox
“If artificial intelligence can figure out wave navigation, then I would quit my job” were the first words that I heard from Riz Moebius as I stepped aboard his 45-foot double-hulled canoe No Gravitas off the coast of Marshall Islands. He stood at about a couple of inches above 6 foot, his wavy jet black hair giving in to the whims of the wind but also giving his words more magnetism somehow. He often pushed his hair away from his face when he spoke, which made you focus on it more, as if he knew the effect it was having on people. I had quit my job six months ago with the hope that I could work 50% of the time on odd contract gigs and focus the rest of it on increasing my revenue from my sub stack newsletter on urbanism. It had been four months to the dot when I met Riz, and so far, I had 20 new paying subscribers. Safe to say that I was not going anywhere fast, so when a gig to work on a self-navigating, self-driving double-hulled canoe in the Marshall Islands came in, I did not hesitate. When I googled the name of the contractor Riz Moebius, I did not expect him to be a brown man - he did not sound like one, and I was generally envious of brown immigrants who seemingly had it better than me. He had a standard LinkedIn account which, thank god, showed no activity in the recent past apart from him leaving his last job as a senior product manager for Amazon. He had no Twitter, no Instagram, no Substack or think pieces on product management, all of which should have suggested that he’s probably running a scam for someone else in the Marshall Islands. I have worked on products that could have been a scam or were scams, and they usually paid well, so when I sent a response with my hourly rate + 75%, I did not expect the reply to be that he could only pay me $25 less than my hourly rate. He then gave me the same spiel he was giving me now about how figuring out wave navigation is the most exciting new thing that AI could do. Was there something here?
“Vijayan, so when can we begin testing this out? I’m not sure if I have a lot of days left”, Riz spoke out loud to me from the other end of No Gravitas as he glanced at the third person on the canoe. I thought the man must be from the capital at first glance because his question mark like figure sitting at the edge of the canoe and his striped blue polo shirt that was big at the shoulders indicated that he is some sort of canoe bureaucrat.
“Let me check out if the last commit and bug fixes got pushed,” I said with hesitance in my voice as I walked below the deck. The palpable tension between Riz and the supposed bureaucrat made me feel like a kid stuck in a passive-aggressive household, which triggered my flight response.
Here I was in the Marshall Islands after an 8-hour flight with a layover in Hawaii, trying to teach a double-hulled canoe wave navigation. The goal of No Gravitas was to reach Aur, an atoll in the Marshall Islands. The Marshalls provide a crucible for navigation: 70 square miles of land, total, comprising five islands and 29 atolls, rings of coral islets that grew up around the rims of underwater volcanoes millions of years ago and now encircle gentle lagoons. These green dots and doughnuts make up two parallel north-south chains, separated from their nearest neighbors by a hundred miles on average. Swells generated by distant storms near Alaska, Antarctica, California, and Indonesia travel thousands of miles to these low-lying spits of sand. When they hit, part of their energy is reflected out to sea in arcs, like sound waves emanating from a speaker; another part curls around the atoll or island and creates a confused chop in its wake. Wave-piloting is the art of reading — by feel and by sight — these and other patterns. The canoe had nine different lidar and radar sensors attached around it gathering data on these tiny ripples, which then the navigation system used to figure out where land was. I had no idea why an ex-senior product manager at Amazon would waste his time and money on this project. Sure it seemed challenging, but it had no real-world application except well that it was interesting.
As I finished my last commit, I heard the sound of something crackling and coming alive above me, and it took me a moment to realize that No Gravitas had begun moving. I had never been on one before. The closest I had come to being in a canoe was training the algorithm on datasets and watching it navigate choppy waters of the Pacific Ocean in a simulator. I headed back up the deck, the perplexed look likely visible on my face. As I turned to say something, Riz interjected and spoke with a sly smile, “Let’s see how it does, shall we?”
“I was hoping to run a couple of tests before we did this. I don’t think this is safe,” I replied, my voice trailing off towards the end of the sentence because Riz did not seem like the kind of person who would heed.
No Gravitas spent the first couple of minutes setting up a reference course based on waves and the sun’s position. The algorithm optimized the reference course to reduce tacking (zig-zag motion of the canoe when moving in the wind direction). These types of canoes moved against the direction of the wind. Keeping to the reference course would be tricky with changing conditions of the sea, wind, and position of the sun. The goal was to stick to the reference path as closely as possible, use it as a proverbial North Star.
No Gravitas moved ahead at a gradual pace. The first couple hundred meters were fine; then, as the conditions changed, the canoe drifted off the reference line; by the time the canoe calibrated and changed the direction of the sails to get back on course, the conditions would change again. This lead to a lot of tacking. The tacking continued for what felt like a day with my parents. While I wanted to have a technical conversation about what was happening, the idea of seasickness which I only attributed as a plot point in cartoons and movies before, quickly became real.
Riz looked at me from a distance with concern in his eyes and asked, “are you going to be alright?”. I nodded to say yes, and then he replied, “Good, we are not turning back until this crashes or we reach Aur.”
There have not been many moments where I hoped my haphazard work ethic would fail me, but this would be one of them. The algo heeded my prayers before I started seeing the sirens of the sea.
“Well, it was not bad for a first attempt. You did better than most of the others,” Riz said as he approached my stooped-over figure in the corner of the canoe. The conflicting feels of disappointment and relief washed away to be replaced by confusion. “There were others?” I asked.
We had halted our journey mid-way through when we realized it would probably take us couple more hours than the estimate to reach Aur with all the tacking.
Riz took a sharp breath to reply, but before he could say anything, the canoe bureaucrat who had not said a word until now approached Riz from behind and spoke, “So is he the man for the job?”. He had a strong Chinese accent; this was another blow to my perception of reality after learning other people had attempted this same job. Riz pushed his hair back as he turned towards the now mysterious third person. “Well, no one’s gone this far yet,” came his reply, rather nonchalantly. At that moment, I felt a sense of pride, like I had made my parents proud. “Well, but then there was that one guy who was good, really fucking good, I tell you,” Riz continued, gesturing with his hands and smiling in my direction, “but unfortunately, he didn’t want the job, so yeah.” I didn’t know if this was some kind of pick up artist trick. Still, my first instinctive response left my mouth before giving it a second thought “Well, I could try again.” Both men had turned halfway around by the time I said this and were walking back. “Let’s head back to the coast, to the one we came from,” Riz yelled, more to himself than to me.
“Well, what job is he talking about? I thought this was the job,” I yelled, the confusion probably visible on my face, but no one to acknowledge it.
“We got to figure out how a sentient plane on full auto-pilot crashed,” Riz yelled back without any change in his tone. He paused for a moment, turned around, and said, “I’ll explain after you sign a bunch of NDAs if you ‘accept the mission’ that is Not for the first time that day. I was puzzled. .”
Riz could not have been talking about what I thought when he said sentient aircraft. I laughed at myself at the improbability of it. A Halos Jets plane manufactured by Octave Aircrafts had crashed six months ago, killing all 100 passengers aboard it. Octave Aircrafts had pioneered an autopilot system that operated with a single human pilot who merely acted as an observer, while their autopilot named Sophie handled take off, flight and landing. It was remarkable technology pioneered over many years of training and test flights. By the time Halos Jets became the first public airline to adopt it, Sophie had put in a staggering 35 million miles in test flights. The flight from Denver to Austin that crashed was its 21st commercial operation. Jules Friedman, the CEO of Octave, vowed to prove that it was not Sophie that was the cause of the accident while Halos and the majority of the venture investors pulled back. A subsequent investigation by Octave and independent observers had proven Jules right. It concluded that a supply chain hack gave someone backdoor access to the silicon that Sophie ran on, and it was this backdoor hack led to the plane crash. The investigation also pinned the supply chain hack on a Chinese company that contracted with prominent chip manufacturers. Both the US and Chinese governments had pulled their embassy officials after the investigation went public. All this to say that I do not think that this ex-Amazon product manager had any role in this Cold War-Esque spy drama that was probably unfolding while he fucked around trying to figure out how to teach a canoe Polynesian wave navigation. Was he referring to any other incident that I was not familiar with?
As I approached Riz, I discovered that I had not fully regained motor controls after throwing up back there. I wobbled up and asked, “What’s this aircraft you are talking about.”
“You know the one,” Riz replied calmly without looking up from the iPad he was doodling on.
“The Octave Aircraft? I mean, what’s there to investigate?” I asked
Riz looked at the Chinese man and then back at me. “You’ll see,” he replied and continued, “I guess that means that you are taking the job?” that sly smile visible again.
I scoffed inadvertently at that question. I wondered what new kind of scam I had found myself in and why Riz seemed too charming to be an ex-Amazon product manager.