The Future Machine
Some people who should have known better tried to say that the machine must have brainwashed us. That’s nonsense. It was only long afterwards we got our first true glimpses of what would be that people started worrying about the morality of it all. By then it didn’t matter. We were fanatics and the future was our scripture.
“It won’t tell the future so much as it will tell a future,” I said, feeling absurd before the words had left my mouth.
“How many billions of dollars did that answer cost to get to?” It was one of the w. executives who hadn’t seemed to be paying attention, but I certainly had her attention now.
“Sorr-.” I said and must have looked as terrified as I felt, because she shook her head and cut me off.
“Relax. I’d have to be an asshole to be upset with you for accurately pointing out that your time machine still has some kinks. We’re obviously going to pass though.” She didn’t spare me a second glance and returned her attention to a handheld display.
I shut up for the little that was left of the meeting.
The team were chatting in our private feed about how it was typical of these corporate types to use terms like ‘time machine’. I ignored them. Yes, a watch is a technically a kind of time machine and we were doing something more complicated, but also who cares? They might not use proper terminology, but being technically right doesn’t come with funding. We needed to convince them, not the other way around.
When I got back from the meeting, waiting in my office was the executive who had objected to my correction. It felt like a power move and made me uneasy before she even said anything.
“Your time machine will never be viable. It’s going to require growing a new tree for every glance into the future and nobody is willing to wait for that. Eventually your investors will catch on and your funding will be be cut before the year is up. It should have been done months ago,” she said without any trace of sympathy. “When that happens, you should come work for us. Even if it doesn’t happen, you should anyway.”
I didn’t hear what she said next and stood there kind of dazed. What an insane way to treat another human being. She continued to talk about the worthlessness of my life’s work as if it were a fact that we had agreed upon. I struggled to process what she was saying. Eventually the rest of my brain caught up.
“Even if this project is misguided, your team will hate to lose you. You’re a talented engineer and a likable enough colleague.”
It felt like my lungs had popped and so I didn’t try to find something witty to say in response. I started to object before I knew what I wanted to say, but she just spoke over me.
“No. You don’t get it. This…” She gestured vaguely towards my desk. “None of this matters. We’ve already got a machine that can see into the future.”
“I know we’re not supposed to say it, but the future machine is essentially an algorithm,” my liaison said with all the reverence of someone who has no idea what they’re talking about. He was right that there was an incredibly complex code repository involved, but there were more people involved in the mushroom farm than there were coders. Hell, there might be more baristas on site than there were coders. If any one of those pieces wasn’t operating at peak productivity, the machine wouldn’t work. It wasn’t essentially anything.
It might seem too fine a point to be worth getting hung up on, I admit, but I was overwhelmed by the number of people who casually accepted the reality of a machine which could see into possible futures. They treated all this like it was a normal Tuesday.
It’s hard to appreciate the technological achievement that the future machine represents because we can’t really do anything with the slice of the future that we’ve got access to. The machine’s massive physical and digital infrastructure spits out a series of numbers that we have only barely began to understand.
My liaison even had clearance to access the future machine’s data, but he didn’t even bother looking at it because he couldn’t do anything with that information. They were numbers without context. The present isn’t connected with the future in sort of linear way. Neither is the past for that matter. When the machine looks elsewhen, there is no way to control what it is aimed towards.
You have to realize that it is merely convenient to us that there are predictable patterns to the entire universe. There is nothing necessary about it. Not just things like the fluid mechanics or orbital decay, but also basic rules of existence such as object permanence. It is fairly useful to us that unobserved things continue to exist.
But when the future machine opens its eyes, so to speak, there isn’t any sort of pattern as to when it is looking. To say nothing of where, what, or who it is looking at.
They’re pretty sure that the number is connect to whoever activates the machine, but they haven’t settled on what counts as ‘activates’. It’s not like there’s a wall plaque that says ‘The Future Machine’ and a brightly colored button. There’s no retinal scan and nobody has to turn their keys in unison. It could have those things, I suppose, if they wanted it to. There are a lot of moving parts and it’s at least hypothetically possible for those things to be added, but that would still only be a small part of the much grander whole.
The point being that if everyone does their jobs, the future machine has all the resources it needs to activate the display which then lights up with some number between zero and one hundred. That it’s. That’s all that happens.
There might be a handful of significant digits or it might be a nice round number, but always less than one hundred.
Obviously they’re keeping track of all the numbers and subjecting them to all kinds of complex analysis. People keep talking about ‘machine learning’ in hushed tones, but the only thinks they know about the numbers is that they’ve been trending upwards and have been for a while.
When the future machine was first activated, the project was essentially being run out of someone’s garage and the numbers it gave tended to be closer to zero, overall. It wasn’t that they could predictable say what any result from the machine would be, but the bigger the project has gotten, the bigger the numbers tended to be when looked at in the aggregate. That’s it. It obviously not just a coincidence. It also isn’t Pi, Euler’s number, the Feigenbaum constants, or the golden ratio.
It also isn’t a duck.
There are a lot of things that the future machine is not, but it’s not altogether clear what it actually is.
After two weeks, I was already bored by the future too.
There was, ever briefly, a novelty to each new number. ‘37’ came up twice and I embarrassed myself by being excited at the reoccurrence. Ravi had the decency to stop me before I got carried away. “To be expected,” he said, as if there was anything expected about any of this.
It wasn’t any one thing that disabused me of that notion. Pulling numbers out the future really wasn’t that interesting until you knew what the numbers were for. Inevitably there were ways to sort and organize the numbers in all kinds of different ways. Given sufficient time and a diverse enough set of numbers, patterns and repetitions will inevitably emerge. They were useless without rules to decide which to pay attention to.
Every night I am dreaming about the Ancient Greeks weaving the tapestry of fate, with numbers rather than thread. Some nights they are beautiful, some nights they are hideous. They never look the same, but I always recognize them for what they are: the machine.
There are other dreams.
I am floating in a nutrient bath which turns into a sea under a starry skyscrape. I am effortlessly riding a bicycle through a field. I am having a lot of sex. There is one where I am walking through the streets of Marrakesh, fondly remembering the fresh mint tea at a cafe where I once spent an afternoon watching the tourists. Waking from that dream was like saying goodbye to an old friend. But I have never been to Marrakesh.
I am not myself in all of them, but I almost always recognize myself in them. That is to say, I am often in a stranger’s skin but never find it strange. At the time, anyway.
“I think maybe I’m losing my mind,” I said to Ravi over lunch. He was technically my boss, but was also assigned to keep an eye on me. Not that it was as nefarious as it sounded. The machine had a tendency to pose some messy questions about the nature of existence, so it was nice to get advice from someone who had been there already.
That made it easier to confess that, for three nights running, I had walked on the surface of Mars as someone else.
“They’re the dreams of other people on the project,” Ravi said with a gesture around the room. “It’s how we know that you’re actually involved in the machine. We’ve been having your dreams since you started, but you’ve taken a lot longer than usual to catch the change in your dreamscape.”
I parse enough of his comment to be confused, but my confusion gives way to concern as I tried to remember whether I had dreamt anything that merited embarrassment. It must have shown on my face.
“If it’s any consolation, your subconscious isn’t any worse than anyone else’s.” As he spoke, I looked around the cafeteria to see if anyone was looking at us. If anyone was looking at me. “In fact, those who have shared your dreams kind of appreciate it. That forest of yours with all those different trees. Not everyone on the project has stepped foot into a proper forest, let alone seen all the varieties that sit at the edges of your subconscious.”
“I still don’t understand. How are you getting inside my head? How am I getting in yours?” I said, feeling frustrated at how ridiculous the questions sounded.
“You’re not going to like the answer, but we don’t know.” He gestured for me to keep eating. “Not really. We know that the mushrooms are involved somehow. That can’t be the whole story though, because the dreams happen to people who have never been to the facility here. We share a dreamscape with anyone who is part of the supply chain that keeps the machine running — whether they’re the executives at the head office or the people who make deliveries to the facility. We learned that one the hard way.”
He chuckled at that last comment and again tried to get me to eat.
“That doesn’t even make sense,” I said.
“You’re going to have to be more specific. The dreaming each other’s dreams or who gets included in the process?” He grinned as he spoke.
“I don’t know anything about fungi. I don’t even know if that’s the right pluralization of fungus. But I can wrap my head around the idea that the farms are doing something to us. Maybe microscopic spores shooting up into the air when the mushrooms bloom?” I had altogether given up on my lunch at this point. “Whatever it is, sure, I can make some sense of that. But how could that possibly go further afield? Is it only people who have interacted with people on the farm? Can mushrooms even do that?”
“It’s a good thing you weren’t hired as a mycologist then, eh?” Ravi had finished his lunch but seemed in no rush to get back to work. Then again, explaining to me how the machine worked was part of his job, so this was as good a place as any. “There was a brief period where we thought that it might be something along those lines. Spores breathed in here and then breathed out on various parts of the supply chain, but…” He left it hanging and it was clear he wanted me to work out the obvious problem.
“But…” I said, less stalling for time and more trying to force out an obvious detail that I must have missed. “But if it were that straightforward, then why wouldn’t it spread more widely? Why only people in the supply chain and not their families and friends and casual contacts? Even if someone only carried a handful of these spores, surely eventually they would spread to someone who wasn’t directly connected and you-.” I caught the error almost as quickly as I’d said it. “And we would start having their dreams, and us theirs.”
“Ding ding ding,” Ravi said. I couldn’t tell if he was happy that I had gotten there on my own or that I was seeing myself as a part of the machine.
“Then what the hell is going on here?” I said.
“If we knew the answer to that, we wouldn’t have hired you to figure it out.”
Not that Ravi meant me specifically. It would have been a nice ego boost if that were the case, but nobody had suggested as much. Once they had figured out that the future machine actually worked, they started recruiting anyone who was available. That turned out to be almost everyone doing serious work on time travel, because a functioning temporal device was worth any number of old ladies. We were all here to work on the future machine.
“I still don’t get the mushrooms,” I said after a lecture that was supposed to have explained what the farm actually did for the future machine. “And if your answer involves the universe as some sort of breakfast metaphor, I quit.”
Ravi groaned and sipped at his coffee before answering.
“The mycology teams really need to tone it down a couple of notches,” he said. “Meetings with them are feeling less like a science fair and more like a religious experience.”
“Given that I’m neither a mycologist nor a priest, I’m not sure the perspective is the issue here.” I took a sharp inhale and continued. “I would believe that you can get electricity out of fungi somehow, but how could it ever be enough to power a computer? Is it just the CPU that needs to be ‘plugged into’ the mushrooms or do the hard drives too? How about the monitors? And why does the farm need to expand as more people are brought onto the project?”
“You’ve got to think of it as poaching an eg-.”
I swore loud enough to draw a couple glances.
He laughed.
“If we’ve developed mushroom-powered displays then why haven’t we solved the energy crisis?” He took a moment to collect his thoughts. “The computer requires the same amount of electricity as any other computer, but it only spits out a number if there are enough mushrooms for all the people who have been looped into the machine. They’re not fuel, because they’re not used up. Their presence is required, that’s all. In fact, the future machine sits in a fairly small patch of temporal mushrooms. The rest are connected through mycelium in the soil.”
“Right. Mycelium. Which is the invisible mushroom root system?” I said, somewhere between sarcasm and a legitimate guess.
“The biological parts of roots are visible. They look kind of like moldy thread,” he said with shrug, as if to apologize for the next part, “but all mushrooms have that. The temporal farm also has an energy field underneath it, around the mycelial threads.”
“That’s a layer overtop of the visible part?”
“Exactly.”
“That’s different from an invisible mushroom root system?”
“Because it’s not there all the time. It only shows up when the future machine spits out a number.” He gave another apologetic shrug.
“So we activate the machine and the threads pulse briefly. Okay, but that doesn’t explain why I’ve got my own personal mushroom.”
“They don’t really keep track like that on the farm though. It’s not as if we which mushrooms were set aside for you. It’s just that sometimes after new hires the future machine stops spitting out numbers until they plant more fungus spores.”
“Why don’t they keep track?”
“We’re not really talking about a one-to-one relationship here.”
“Sure, but let’s pretend the machine stopped working after I got hired. The mushroom priests plant one mushroom at a time until the machine starts working again. Are we tagging those as my mushrooms?”
He shook his head slowly and it looked like my thought was dawning on him.
“I’m not sure I’m ready for your inevitable mad descent into the becoming one of the mushroom people,” he said, “but I guess we need to introduce you to the mycologists.”
I’ll take credit for the initial thought, but there’s an entire team looking at the relationship between the numbers and the fluctuations in the energy field. I sent them a memo with my proposal and sat in on a couple of meetings with them, but it was clear that I was in over my head on what they were talking about. For starters, what I kept calling ‘mushroom energy’ they insisted was ‘mycelial temporal radiation’.
Not only did they start keeping track of which mushrooms belong to which new hires, but they also worked backwards to provide their best guess about earlier growth. Given how methodical the farm teams were in their work, it was easier than anyone had expected. The result was that most people involved with the machine knew exactly which mushrooms were theirs.
Aside from sating curiosity, this proved important because the mushroom energy didn’t appear across the entire farm at once. Instead, right before the future machine activated, a small handful of mushrooms started giving off barely measurable amounts of energy. It had actually been noticed once earlier in the project, but the technician decided it was a fault in the sensor. She replaced the sensor and went on with her day without reporting it.
Now that we knew what to look for, it became clear what was happening: when the future machine activates, a specific patch of mushrooms lit up the instrumental panels like a meteor shower. Because each patch was tagged according to employees on the future machine, we also started tagging the future machine’s numbers to people working on the project. We didn’t still know what the numbers were for, but suddenly everyone was excited about them again. After all, people were starting to have their own slice of the future.
My first number was ‘37’. Ravi tried to assure me that it was a coincidence, nothing more, but I still needed to take a week off after learning that.
“… but what if the radiation doesn’t just activate the future machine but instead explains it?”
Most of the project involves throwing highly specialized experts at very specific problems with the hope that it’ll lead to new discoveries. In an attempt to share ideas and let everyone know what else is being worked on, there’s a mandatory lunch session every week and each team has to choose someone to present. Most teams make it grunt work for new recruits because nobody takes it seriously, but nobody complains either because of how well catered it always is.
Are you hearing this? What team is she on? Ravi messaged me through the feed. He was on the other side of the room and he was staring at the current presenter who was in the middle of speaking.
Through a mouthful of bánh mi, I told my feed to replay the last few seconds of audio at double speed so that I could catch up.
“We’ve assumed that the only purpose of the farm is to power the activation process of the machine, but what if the radiation doesn’t just activate the future machine but instead explains it? It would make a mess of our understanding of temporal mechanics, but what if the mycelial radiation isn’t merely flowing into the computer to give us num-,” she said through the recording and then continued live, “-bers? What if it’s simultaneously flowing backwards into the farm? We-. I think we need to stop focusing on the numbers that the future machine gives us and instead focus on listening to the mycelial radiation.”
I quickly messaged Ravi nack. She’s on one of the programming teams, but she sounds like one of the mushroom people. I also think she’s right. Even from here, I could see him nodding while he read my message. He looked up at me and tilted is head towards the woman who was now leaving the lunchroom stage.
I headed through the crowd towards him and grabbed another sandwich on the way.
Cassia had been brought on shortly after I had, but wasn’t even being listened to by her own team. We had approached her before she had been able to find a table and she had obviously hadn’t expected anyone to follow up.
After quick introductions, the three of us stood in the hall outside the lunchroom.
“The obvious answer has always been probabilities,” I said while Cassie ate my bánh mi. She looked up at my comment and made a noise that was difficult to parse with her mouth full. She took a moment before trying again.
“Didn’t the guy who first proposed that get fired after starting a fistfight in a meeting?”
“Technically it was after the meeting, but yes,” Ravi said with a laugh. “At one point, I thought he was going to stab me with a pen. Still, being a prick doesn’t mean it was a bad idea.”
“The problem has always been what kind of probabilities. Without that added context, the raw numbers are utterly useless. That’s why so few people have gone along with it,” I added, getting back to the matter at hand. “But what if we can find that context by listening to the mushroom energy?”
“You get that I have no clue how to do that, right? My team lead told me to spitball for a couple minutes. That’s all I was doing.” She seemed worried that we were expecting more out of her than she was prepared for.
“You mean we don’t have a way to plug your feed into mycelia?” Ravi said in mock disappointment. “Or a set of speakers so we could vibe to the rhythm of the fungi?”
A relieved expression crossed her face as she recognized the joke.
“With your permission, I’m going to request to transfer you to my team,” Ravi said. “We’re not quite an island of misfit toys yet, but we’re close enough that we get left alone to solve weird problems.”
“Like how to understand what mushrooms saying when they whisper sweet nothings into the soil,” she added, not quite a question and not entirely a joke.
“Yeah, you were right, Ravi. She’s definitely one of us.”
While the entire farm was being fitted with almost every sensor that we could think to point at the mushrooms, we busied ourselves with trying to translate past energy data into something understandable. I spent almost a full week comparing the radiation patterns to weather data while Ravi delved into betting and stock markets. Neither of us got anywhere.
Cassia really did try something we ended up calling ‘mushroom music’ and found a way to push it through a synthesizer. It wasn’t useful, but it was pretty cool to listen to, so it was a more productive use of time than anything else the team was working on.
“We should stop calling it the future machine,” Cassia said to herself, but loud enough that it interrupted a debate Ravi and I were having about which mushrooms made the best sounds to fall asleep to.
“Why?” We both said in near unison.
“If we’re right, at best it’s a probable future machine.”
“You’ve the soul of a poet, you know?” I said.
“Fine, call it the probability machine. It doesn’t offer a guaranteed future is my point.”
I glanced at the display on Ravi’s wall to see future machine feed.
“It won’t offer any kind of future if we can’t figure out what 68.44 is a probability of. At this point, we might as well call it a random number generator.” Before I finished speaking, the machine had updated to 1.914667, 12, and then hung for a few seconds on 23.1. It was running fast today.
“Admittedly, that’s a bigger issue, but we can work on multiple problems at once. We’re talented people,” she said, setting her handheld display down, “and what would we lose by being more accurate with what we call it?
“Well, probably your paycheck, for starters.” Ravi said completely serious and Cassia looked confused. “Even if the overall details are shaky, most of us here have a pretty good sense of the limits of the future machine. It’s almost certainly never going to turn into a streaming feed of tomorrow. We know that. Who doesn’t?”
“Oh,” she said understanding, “right. We still need investors.”
“Don’t get me wrong. There’s probably funding to be had for a probability machine too, but we’re burning too much money to start being realistic now.” Ravi said. Given that part of his job was managing the team’s budget, it wasn’t a surprising comment from him. “ The end goal of all this is still a functioning time machine. One day. Until then, call it whatever you want as long as you’re putting ‘the future machine’ on anything that might make it in front of the money. The only reason I’ve got all this leeway is because the investors like me. I don’t make them feel like idiots and I don’t confuse them. If that changes, then it’s off to the fungus mines with you two.”
Why don’t we just build an app? I messaged privately to Cassia, thinking she’d respond in the morning and that I’d have time to work out what I meant. Instead she replied immediately.
I can’t tell if you’re still mocking my old day job.
No. Not right now, anyway. I mean it. I had a screen full of data from the mushrooms. Our new sensors had been live for a week and it was interesting to learn that light was bending, ever so slightly, around the mycelial radiation. I had even forwarded the results onto another team that might be able to find a practical use for that.
But we were no closer to being able to prove that it was spitting out probabilities.
I’ll bite. What do you want this app to do?
That was a good question and I wished I’d thought of an answer before messaging her.
No, I meant we should turn the machine into an app.
She didn’t respond, either because she’d fallen asleep or was waiting for me to elaborate.
For all the technological knowledge that went into making the future machine function, there were still parts that were essentially magic. Those of us who work on site are at least occasionally breathing the same air as the mushroom farm and each other, so I can wrap my head around us all sharing a dreamscape. But why do I share dreams with people on the other side of the world?
Plus the whole sharing dreams at all detail. It’s become so normal for us all that we fail to appreciate that we know absolutely nothing how that is happening. Or what is happening for that matter. Are our subconscious minds being linked together? Are our incorporeal forms gathering on some shared metaphysical plane of existence we can’t prove exist? We might we be racing towards solving the least interesting problem that the future machine had presented.
But there were people better equipped and more interested in that work. We sometimes joked about the mushroom people, but they were pushing up against the rules of the universe too. Do that for long enough and it would be indistinguishable from madness.
The difference was that our team’s madness was making sense of bits of future that we received.
We should flood ourselves with probabilities. Give ourselves so many probabilities that patterns become more obvious when we compare it to real world data. Maybe the machine predicts the likelihood you’ll commit a crime. Maybe it’s whether you’ll check your feed in the next two minutes. If nothing else, we need the ability to run blind experiments on people who aren’t working in the building.
I considered whether I needed to add anything else to my wall of text and added one more sentence before letting myself head to the dreamscape.
Besides, this’ll be a step up from building porn games.
———
Before I had woken up the next morning, Cassia was ready a quick mock-up of the app. It took her a couple more days to have something that could run on our handhelds. It didn’t need to be polished, we just wanted to test the concept. At that point, it was essentially just a button on your display that you tapped and it ran a system diagnostic on a non-critical server in the building. We’d figure out something more elegant if we had to, but this was sufficient for now.
When the next batch of new recruits were being looped into the machine, Ravi got special permission to mess with the process. Normally they get moved on site and do minor busywork related to their field of expertise until they show up in the dreamscape. Instead they stayed at home and were told to use the app a differing number of times a day.
We started having their dreams almost immediately, so there wasn’t an obvious relationship between app usage and before they were looped in. That didn’t matter to us, because the future machine app was viable. Now we had to figure out what to do with it.
———
In the first six months, the future machine couldn’t keep up with the increase in scale. We gained users faster than our most ambitious projections. The farm had to be rapidly expanded and that presented all kinds of technical difficulties for those teams. Apparently it isn’t as straightforward as simply planting more mushrooms, but the app kept us too busy to pay attention to the farm.
Initially, we had been keeping an eye on a couple promising indie game developers who seemed likely to go viral, hoping we could pay them to let us slip some background code into a patch. It would have been the easiest solution, but the future machine won’t let us passively loop anyone into using it. So, Cassia went ahead and designed The Future Machine. We mostly just call it the app though, so as to not confuse it with the actual future machine.
Given that all horoscope apps are run by charlatans, we didn’t feel too bad getting into the market. Unlike them, we actually have access to the future, we just don’t know what it means. Instead of hiding that, we embraced the absurdity of it. We can easily push probabilities to their feeds, but nobody wants an app that just occasionally throws numbers at you. Which is why we gamified the whole thing. It’s a journal, horoscope, and life coach, but with addictive gameplay loops that incentivize regular usage.
While we were beta-testing, Ravi’s boss stopped by and suggested he was going to cut the app project altogether because it was just a distraction from more fruitful avenues of inquiry. That’s all Ravi needed to finish his proposal to management for the app to be its own department. Up to that point, we really had been making it up as we went along. We paid for that mistake with back to back all-nighters preparing spreadsheets and slide decks.
Ravi came to that meeting with a full progress report to-date, and a detailed roadmap for the next three years. We put a lot of work into looking like we knew what we were doing, but most of it turned out not to matter at all. Management saw that we’d have a modest revenue stream and signed off.
It probably helped that they could play with The Future Machine. They had to dedicate a non-zero amount of their days to teams proposing ideas like shooting mushrooms into black holes. Compared to that, app development was always going to seem downright straightforward.
Insofar as anything do with the machine was straightforward, that is.
———
“This might be stupid, but are we supposed to only ask the lawyers hypothetical questions?” I said legitimately unsure
“How can you possibly be this naive?” Cassia said without looking up at her screen, but Ravi was ready to finish the thoughtful her.
“We’re not passing coded messages back and forth. If this were a legal thriller, we’d be the villains who talk about their plans over an expensive scotch with a crackling fireplace in the background.” He paused for a moment and nodded thoughtfully. “Can we be in a legal thriller instead?”
“Maybe if you’d gone to law school, none of us would be in this mess,” Cassia said.
“I’m not clear why any of this is my fault. I think that’s why it’s call a team effort.”
“That’s only until something goes wrong. Then you’re management,” she pointed and me continued, “and we’ve just done what you told us to.”
“Oh I see how it is now,” he said in mock horror. “I’ve been betrayed.”
“You really should have seen it coming. You do run the number six horoscope app in North America.”
Ravi shrugged at that and then turned his attention back to me.
“The lawyers are pretty adamant that we avoid any dream-related content on the app,” Ravi said apologetically. “The liability concerns are just too much hassle.”
“As much as I’d love to talk in this circle yet again, I actually have a different point here,” I and paused. Cassia caught my tone voice and set aside her display to pay attention.
“Well this should be good,” she said.
“I’m just thinking that if we’re breaking a bunch of laws already, why not get some better data out of it?”
———
Opening up the dreamscape to anyone who used the app wasn’t exactly an oversight. There were countless meetings about the logistics of scaling up the mushroom farm. The legal team nixed the dream tracker features before we had gone live. We knew that anyone looped into the machine would eventually start sharing their dreams with us.
We just didn’t care. What problem couldn’t be solved by just a little bit of an edge on the universe? We wanted the future more than we cared how we got there.
The entire app was built on the assumption that the future machine is giving us probabilities. Those probabilities relate, in some way, to the mushroom energy surges on the farm. And those particular mushrooms relate, in some way, to particular individuals. Obviously we were hazy on the details, but that we knew these things were all connected. How was to be seen.
When the machine generates a number for an individual, we identified who it was associated with and then the app pushed a seemingly personalized journey of self-discovery. Each journey was a different selection of mental health advice stacked on top of Barnum statements stacked on top of whatever pseudoscience questionnaire we were pushing that week. Then the app gave you your number and a vaguely quirky but still plausible outcome.
Maybe you had a 96% chance of missing a stranger smiling at you. Or 48.52% chance you’d discover something new that you’ve always believed. Maybe it was as simple as a 13% chance the rain will follow you today. The numbers were the only part that weren’t a lie, but the machine doesn’t seem.
We also did virtual star charts explorations, for whatever those are worth.
It wasn’t enough for people to casually use the app. We needed them to let us into their lives so that we get looped into their lives. Paired with time stamps and location data, we made some serious inroads on identifying when the machine was giving weather probabilities. We learned that the mushrooms responded differently depending on the type of probability they were generating. The mycelial radiation had the appearance of randomness that matched the randomness of the machine’s numbers, but there were incredibly subtle patterns in the radiation.
The patterns only really became apparent when you already knew what you were looking for. After comparing our data to highly localized weather and temperature information, we had a collection of almost two-thousand probabilities that could plausibly be related to the weather. We started sorting and tagging each number based on how confident we were that the weather was involved. It was an absurd exercise, like playing darts but not knowing if you’re aimed towards the dartboard. Or if there’s a dartboard at all. Or darts, for that matter, but you still have to hit the same spot twice.
Then Ravi did it.
“You can’t think of the radiation data as numbers,” Ravi said. said. He paced across the room in excitement and agitation. Excitement because he was certain he had figured it out. Agitation because he hadn’t figure out how to communicate it.
“What do you do when you’ve woken up from a dream that you can’t remember but want to?” He seemed torn between wanting us to guess and wanting to get his thought out. The latter won. “Instead of focusing on the dream, you let it go. You calm your thoughts and give the thought permission to leave. When your mind is that empty, a single word might be all it takes for an idea to come flooding back.” Ravi cycled through different data radiation sets, pointed at the office display, and paired each point with an emphatic “See!” that wasn’t immediately helpful.
There was a strong temptation to mock him when he got like this, but it was better to let him work ideas out in this way. If nothing else, it wore him out. “Once you look at it like a memory of a dream, then you’ll remember that these are predictions about weather.”
As absurd as it sounded, that wasn’t an unfair description of the process. The shape of the mycelial radiation made sense in the same way that dreams do. Fortunately, we were sharing dreams which made all this significantly more straightforward once we knew what we were doing and Ravi’s ramblings were translated into something the rest of us could understand. It was a method of identifying when the machine was giving a weather probability. That’s not particularly useful without more context. What are you going to do with a 33% chance of some kind of weather event? The correct answer, by the way, is compare it to weather data and see if you can’t repeat the whole process with the risk of rain. Other teams took notice, adapted Ravi’s approach, and built on his breakthrough to make their own.
Finally, the machine was beginning to live up to its name.
“Can we start calling it a probability machine yet?” Cassia said.
“Still no,” Ravi said without turning away from the presentation.
We were in the middle of a lunchtime lecture where yet another team was advocating a mushroom farm on Mars. Variations on the idea came up every couple of months, but this team wanted to see whether the machine could operate without people looped in at all. It was a genuinely interesting research question, but I couldn’t see why they needed to go to Mars when they could have gone to Antarctica. The romance of space might have something to do with it, but there wasn’t much space for romance anymore. Not now that we had a pinhole into the future.
There were a lot of people working on the machine who didn’t understand that. We got away with idle research before, but only because the project had stalled out. For most of the team, the future machine was the most interesting work we had ever done. For the investors, it was the fountain of youth. Or it was the closest they had seen in the risk management strategy that was their lives. There were other ways to make money than by gaming the future, but there weren’t other ways to unmake a tragic accident.
Cassia’s insistence on calling it a probability machine wasn’t really rooted in the grammar of time travel. It was her quiet rebellion at the smallness of the whole project. She wasn’t particularly interested in getting fired so, unlike the aspiring Martian colonists, she kept her rebellion quiet.
“We have access to the future and the best we can think to do with it is to with it is horoscopes and spreadsheets.” The presenter was avoiding looking at Ravi and our table, but enough people were trying to gauge our reaction that it felt like the spotlight was on us. “The machine could do some good if it weren’t being run by a bunch of assholes.” For most of his presentation, he had been venting these kind of vague frustrations about the app, but now there was an expectant silence across the lunchroom. He still wasn’t looking at us, but he wasn’t not looking at us.
Ravi was clearly irritated, but gave no indication he was going to respond. Cassia was nonchalantly looking at her feed and might not have cared about the tension. I probably looked like I was wincing in pain, because Ravi had kicked me when I opened my mouth.
He was right, but only because we knew something nobody else in the room knew: last week, the machine had spit out a 98% chance our antagonistic Mars enthusiast would be fired.
Looping people out of the machine proved to be as easy as firing them and uprooting their patch on the farm. No more mushrooms meant no more dreams.
A couple of folks were disgruntled enough to break their non-disclosure agreements, but they made the mistake of telling too much of the truth. Things might have gone differently, had they focused on sheer amount of personal data we were collecting. Instead they tried to convince people that three wellness influencers and their horoscope app were a front for a multibillion dollar machine that could kind of predict the future. That it was true didn’t make it any easier a sell.
Yes, some people believed it, because some people always do. Never enough that it mattered to us, anyway. I imagine the security budget grew accordingly, but that was someone else’s problem.
Our problem was getting enough sleep.
Most of us think about the future as something that belongs to us, as moments that will be ours. What that misses is that we’ve more claim to the past than we do the future. Each moment is an oil painting, brushed over the last but the present layer is always dry. No strokes can be tweaked. When the machine activates, it doesn’t change that. The present is.
But the near future isn’t likewise dry. It’s close though, because the physical world has already been set in motion. As it turns out, however, the subconscious mind isn’t governed by the same rules as the rest of the universe, so that’s where we went to watch the brush strokes. If the machine was showing us the future, the dreamscape was how we made sense of what we were shown.
Apparently some people can directly control what happens in their dreams. That sounds a lot like witchcraft to me, but so is soothsaying, so who am I to judge. Those of us whose dreams are essentially a random word generator have a harder time with the process, because we have to focus our subconscious on mycelial radiation patterns. That focus usually involved a pre-sleep hypnotherapy session and a light barbiturate cocktail. All to end up dreaming about the work as precisely as possible.
Even though we were often passengers in each other’s dreams, there was nothing about the dreams themselves that were different. You knew you were privy to someone else’s subconscious, but that was all in the normal range of weird, wild, and work. People dreamt about the machine because it was the most significant thing going on in their lives and that was inevitably going to show up in the subconscious on occasion, but no more than was to be expected.
Ravi’s discovery was that dreams about the machine were different. They made sense when you woke up and, equally importantly, they didn’t fade away like dreams had a habit of doing. You could spend weeks trying to read and translate a burst of mycelial radiation into a useful probability, and still fail, only to have a dream about it where the pattern finally came together.
The dreamscape couldn’t do our job for us, but it was an ideal testing site for different theories. Our nights were spent dream-working out the syntax of the future and our days were spent coding out what we’d learned. There was no point wasting time by manually translating individual probabilities. They would usually be long out of date by the time they’d been translated. In order to make use of the information, we were always going to have to automate the process.
Translated probabilities came in for Ravi and Cassia long before my first one did, but its weirdness was worth the wait: 0.0001492% chance that I would think about dragon fruit. Cassia rechecked the code while Ravi took a nap to translate the data in the dreamscape, because obviously I immediately thought about dragon fruit. After a thorough review, there were no technical problems found with the machine. Unlikely, after all, doesn’t mean impossible.
For the most part, though, it had a boring streak. 5% chance that you’d receive a feed alert. 82.01% you’d catch a nail on fabric. These kinds of probabilities were hated by investors, the military, and the curious alike. It wasn’t what they wanted it to be or it couldn’t be relied on in a crisis or it wasn’t inclined to be interesting.
At the same time, the app was more popular than ever, but that had nothing to do with the machine. We hadn’t updated it to include the translated probabilities, because it seemed unlikely that anyone would appreciate their wellness content veering into absurdist poetry. It’s hard to center yourself when there’s a 7.7% chance of turning on classical music that drowns out the sound of barking dogs. Finding a way to integrate the future into our horoscopes might have been interesting, but it wasn’t a high priority. Whatever edge on the universe our investors had been expecting, it surely wasn’t giving random people a heads up that they had a 41.71% chance seeing two chairs that they briefly thought was one chair.
It wasn’t that they were all useless probabilities. There’s a team testing the machine as an early warning system for natural disasters. It’s incredibly rare to get a probability along the lines of an 80% chance of dying in a tornado, so they’re trying to find disaster indicators in seemingly unrelated probabilities. The math looks pretty complicated, but Ravi is pretty sure they’ve onto something.
That’s how a lot of the work ended up going after we first started to translate probabilities. It only takes a couple of seconds after the machine activates and we’re not likely to get it much faster than that. The technical limitations of the sensors mean that readings take more time to be processed than the translation code does to run. Given how few probabilities are actionable, let alone useful, research is focused on a lot of sideways thinking from highly specialized experts. Most of it goes above my head. Which is how it should be. Between maintaining the app and tweaking the translation software, Cassia and I had work of our own.
“They’re looking for you to translate these probabilities by hand,” Ravi said. He had forwarded the request to my feed and was poorly hiding an urge to laugh.
“Do they think that I’m wanting for work?” I opened the files and glanced through the data on my portable display. Nothing odd jumped out at me immediately, but I was groggy and only halfway through my first coffee. I was going to have to sit down with them at my desk to make sense of them. It wasn’t something you could eyeball. If I were dreaming, it would have taken a couple seconds, but I wasn’t about to waste REM cycles on other people’s projects unless there was good reason. “Contrary to popular belief, I still have stuff of my own to do.”
“I mentioned that. Ali says it’s not a high priority, but his team thinks the software is consistently mistranslating a set of probabilities as related to weather when they’re…” he cleared his throat and then paused. “He thinks they’re about sex.”
There wasn’t anything titillating about that. Once you’ve shared dreams with your colleagues, your definition of risqué shifts pretty dramatically.
“I’m not following.”
“He looped his parents into the machine a while back. He would call them up to give them a heads up about thunderstorms,” Ravi said, “and last week his mother said that she always loved to hear from him, but she would disown him if he interrupted date night again. He’d been doing it for six months…”
I started laughing in the middle of Ravi’s retelling. The machine had a strange sense of humor at times. Like most parts of the job, it was troubling if you thought about it for too long.
Accidentally keeping one’s parents from getting laid aside, a lot of my work with the machine looked like that, because, for most people working on the machine, the dreamscape was more effort than it was worth. People used it in different ways, but nobody as much as Cassia and I. It was a difficult art and everyone was thankful that we had largely automated the process, but that meant we had accidentally become the foremost experts on translating the future.
So, when the machine gave out a 100% chance event, we were the ones who got the call.
“I know the number appeared as ‘00’ on your feeds. That’s a bug that the original programmers missed. They assumed that the machine couldn’t guarantee anything. It appears that they—we were mistaken.” Cassia spoke quickly but confidently.
“So, it’s genuine?” It was one of the executives on the shared call, but it was hard to tell which because they all started talking at once.
“It’s genuine, but hold the questions, please.” She had more important points to make, but wanted everyone on the same page first. “Thirteen and some change minutes ago the machine gave us a ‘100’ result. Diagnostics are being run, but we have no reason to assume an error. That is because there was an accompanying new event on the farm too.”
She paused to take a sip of water.
“Our sensors need to be precise to determine the origin point of mycelial radiation. As you know, we have gotten quite good at determining whose mushrooms it originates from. For this event, every mushroom lit up simultaneously. Subject to a bit of negligible variance for sensor malfunctions, every mushroom gave off the same radiation pattern for exactly 37 seconds.”
I could feel Ravi’s eyes on me, but I didn’t look up. I was focused intently on memorizing the raw mycelial data we had captured.
“That means we’re dealing with a probability for not only every person on this call, but every person looped into the machine. The translation software buckled under the processing strain, but that didn’t matter anyway. The complexity here is unlike any probability we’ve seen thus far. The software will have to be rewritten for the next time this happens, but that’s a problem for tomorrow. Today we’re trying to translate this probability.
“Okay, I’m going to pass it off to Rav now.” She walked off screen before he was ready to replace her, but we’d be forgiven for a bit of sloppiness today.
“Given that we learned how to translate all prior probabilities with the help of dreaming, that’s what we’re going to do now. Our team all just woke up, but we won’t be waiting until this evening to get to work. Once we’ve memorized the data, we can induce sleep.” He waved towards me, although I wasn’t in the camera’s sight. Cassia had joined me at the table and was back to the radiation readings. She caught my glance and gave a tight-lipped smile. No matter how banal the translation ended up, this was a big deal.
There was an eternity between a probability and a guarantee.
Chemistry meant that sleep came immediately, but dreams did not. They flitted around at the edges, almost cohering. Almost.
There was no way to know how much time had passed. It wasn’t as if our watches dreamt too. Just the growing frustration of dreamless sleep. Insofar as I could grab hold of thoughts at all, I blamed it on overeagerness to see the future.
The farm appeared eventually — or immediately after I fell asleep, who can say. The mushrooms and their intricate mycelial network were such a regular occurrence in my dreams that this felt like going to the office, even if the commute was a weird one.
Still, it felt wrong when the dream came together. Mushrooms thrived in the dark, but the farm was always lit by dusk in my dreams. This was neither. It was cold and clinical and bright enough that I thought I was waking up. Instead I was struggling to breathe. The air was dry in a way that would have been impossible on the real world farm. Without any smell of life to the place, this would never be mistaken for any kind of farm.
I wanted to explore the contours of the wrongness and, as dreams are wont to do, the dream started to shift. I was almost tired enough to let it, but there was already more work than my subconscious was inclined to do. The reminder of work was all I needed for the farm snap back into focus. The eeriness didn’t go away, but I set it off to the side so that I could start translating.
The sleeping mind can be guided, but only if you’ve vigilance enough to sidestep distractions without letting consciousness interfere with inspiration. Take a left at the white rabbit and if you come upon a mad hatter, you’ve gone too far.
Even with the uneasiness, the data I had been memorizing was waiting in the dream. The sensors measured every sense available to us and many that were not. Those readings could be graphed and charted, put into spreadsheets and databases, or ignored altogether in favor of audiovisual records. The trick to translating the future was that all of this mattered, but it was a puzzle that you only knew how to solve once you had solved it.
Outside of dreams, that process looked a lot like unraveling a conspiracy. A subtle rise in temperature might cause you to reassess a frequency caught by microphones that listened beyond human hearing. That, in turn, might remind you a childhood memory of stumbling under your own feet and scraping your hands on concrete. Suddenly, after days or weeks of examining these connections, it would become clear: you were working on a probability about falling in love. No straight lines and the subconscious got there faster. It just had to be tricked into getting started.
I wanted to get to work, but wanting doesn’t work in dreams.
There were an infinity of mushrooms stretching off into the distance in every direction, so I jumped at the unexpected sight of Cassia and Ravi, because there had been nowhere for them to come from. It wasn’t really them, though. They were off having dreams of their own and this was just my subconscious trying to takeover. That was the whole point.
The artificial trappings of the farm — the walls and walkways, the electronics and all their unnatural noises, and all manner of debris that we leave behind when we leave a place — all that began to grow more distant. It didn’t go away exactly. The farm was still the farm, but there was infinitely more space between things. If I had wanted to, I could have looked off towards the horizon and seen the office in which I was currently sleeping.
Cassia and Ravi were still walking towards me, but never getting closer. I tried not to wonder at whether their actual sleeping selves were having an easier time with the translation than I. That was enough to stop them from becoming the focus of the dreams, because my subconscious was more interested in the soil vibrating underfoot. There was a long, deep hum as the entire mycelial network pulsed. Everything that was connected to it pulsed in unison and everything else got even more distant. My office became a star, flickering away the light years between us.
Certainty flooded through me. I didn’t need to see a number on a screen to tell me that this was no mere probability. It was more than that and I knew that it was the future in the same way that I knew how to breathe.
To know what will be is a powerful intoxication. I didn’t know what to do with it at first and that indecision almost dropped me out of the dream altogether. Instead I let my knees buckle and fell, not worrying if I hit the ground, not worrying if I was caught. The dream versions of Cassia and Ravi never came any closer, but their arms supported my weight and kept me from being overwhelmed.
Now that I was certain that the future was coming, I let it.
The mycelial network was gossamer strands, stretching out underneath the soil and connecting all the disparate parts of the machine. These intricate threads were a living net and when the machine activated, the mushrooms exhaled and the entire network relaxed. This was how it readied itself to catch the future.
Everyone dreams the data differently. Ravi’s translations come from conversations had while dreaming. Cassia sees code and debugs the future. There was even a team lead who swore that all his translations came in the form of intricate Gantt charts. The easiest way to make sense of the future is through what we already know.
For me, the process looks a bit more natural. The farm always becomes a forest, albeit an impossible one. Each mushroom grows into a different kind of tree. Weeping willows next to ginkgo biloba. Douglas firs surrounded by tropical fruit trees. If I were so inclined, I could wander the forest until I found an ancient Norway spruce perched on a rocky outcropping, but to see trees that I already knew was not why I was here. I was looking for a newcomer to the forest and watch the future bloom.
I don’t know how long I wandered, but eventually a patch of cherry blossoms led me to a vast, ugly glade where there I spotted what I was looking for.
I knew that the tree was sick before I ever saw it. I could see that the grass beneath my feet was dying and looked worse the closer I came. It was as if something resembling growth rings extended beyond the tree’s bark, marking the spreading wrongness that could be felt through the dream but found its center here. I wanted to want to turn back, but the machine had told this future and I was here to translate it. Wrong or not, there was something to be learned from this.
The tree must have looked impressive in its prime, but now it was collapsing inward on itself. Branches that once might have held the weight of eternity had snapped under their own weight. There were no leaves, whether they rotted on those branches or the world beneath my feet. Even the deadest trees are filled with life — animals that repurpose it for their homes, insects that live out entire generations on decaying wood, and new growth of some sort will inevitably be found sprouting out of the old. But there were no such signs of renewal here.
Not knowing what to do, I reached out to put my hand on its bark.
I saw the tree as a seed growing on the branches of some long forgotten tree that only existed now as its ancestor. I felt the seed fall and get picked up by the wind long before it ever connected with the ground. It flew for what seemed to be forever before touching down here, in a spot exactly like this one. These were the earliest memories of life that the tree had and it was telling me its story. Stories. Eager to learn them all, I reached out to place another hand on it.
Instead the tree fell apart at my touch. From its great heights down to the remaining roots deep in the soil, it turned to something less than ash. The tree may have been a shadow of what it was once, but there was still so much of it that I was engulfed and overcome as it disintegrated.
Then the remnants spread. I watched the already infected glade become a lifeless husk. I watched the cherry blossoms lose first those blossoms, then their color, and finally the trunks themselves. A mango tree had its fruit rot on the branch before becoming unrecognizable. An arbutus peeled away its layers until the tree was impossibly thin and then it was nothing. Each tree reacted differently, but the end result was the same: death, disintegration, and dispersal. It spread to every tree in sight, then on the horizon and, eventually, even the distant stars faded from the sky.
Hours might have passed while I sat there, watching the dream play out. It felt like longer. I wanted to wake up. I didn’t want to be alone with the emptiness. Even the dream versions of Cassia and Ravi had stopped their futile journey towards me, although I didn’t notice when and I didn’t know where they were now. Had they shadowed me as I wandered by the jacaranda and the whistling thorn? Had I lost them somewhere in the maze of quickbeams and rowan trees and dogberries? Or had they followed as far as the glade only to get swept away with everything else?
The answer didn’t matter. They had been nothing more than idle thoughts bouncing off my subconscious and not needed to translate the future, but their absence was a reminder that I didn’t know how far along Cassia and Ravi had gotten in their own subconsciouses. They would come to the same conclusion. Maybe they already had.
“The end of everything,” I said out loud, giving voice to both the obvious and the future. To be precise, there was a 100% chance of the end of everything.
It was an exciting result. It meant the machine could reach further into the future than our wildest dreams. I know enough cosmology to know that, some hundreds of billion of years hence, we will eventually run out of tomorrows. That will be somebody else’s problem.
For us, the machine had touched a moment hundreds of billions of years into the future. We couldn’t do much with such a far afield future, so tweaks would be needed to find a nearer one. I’ve even come up with some ideas in the hours since I’d finished my initial translation. In fact, enough time has passed that I’ve reached the limits of idle speculation. There was nothing more to be done on that front until I had more meaningful data.
I’ve never run into difficulties with chemically induced sleep before. The work was too important to rely on the body’s natural rhythms. The future is one hell of a drug. There were obviously risks, but we considered them unlikely as long as we were safe.
And yet.
Something must have gone wrong. Otherwise I would have woken already. Even accounting for the strangeness of dreams, it has been a long time since the last specks of far off light faded away. A dream in which nothing is happening is strange. Stranger still, I feel tired. That has certainly never happened before and strikes me as an undeniably bad sign. If the dosage has been mismanaged, I may just be struggling to wake. Or it could worse than that. This could be what dying feels like without the benefits of a body to mediate the experience.
I don’t want those to be my last thoughts, especially if someone else has the misfortunate of sharing this dream. Thoughts themselves are hard to hold onto here, but strong emotions can linger much longer. When those are someone else’s emotions, it can mess with your whole day.
I keep finding myself almost letting exhaustion take me, but never quite ready to give up that fight. For a while, I tried to make things happen. I tried to urge myself into running on a sandy beach with the hot day being smoothed by diving into the refreshingly cool water. I tried to find myself in any number of a range of bodies, familiar and alien. I even tried to dream the distorted reflection of the farm back into existence, given that it had already been part of the dream.
When that didn’t work, I tried to stop trying. I hoped that might kickstart my subconscious, but this has truly become nothing at all. Whatever has caused it, I am trapped in whatever remains after a dream has ended.
Even if there were any more future left to be found in this aspect of my subconscious, I don’t have the energy to look for it. My eyes have stopped searching the emptiness. Figures sit just beyond the edges of my vision, always forming and always in retreat. I cannot see them, but I know that I have seen them before. They used to weave-paint-plant-craft-code the future. Now they are a chorus without a song. The silence fills the emptiness or the other way around. At the end of everything, they won’t even do me the kindness of a whimper.