Not long after Korea reunified, Jhonas Angat officially and inarguably cemented his name into the history books. He was already responsible for the theory and math that allowed astrophysicists to illuminate dark matter and deepen their understanding of the universe; he created the first drives and generators to proved his own work; now he gave us a key that truly changed history. Specifically, he figured out how to make the generators work. Actually, really, reliably, work.

Thousands, if not millions, of scientists and engineers, had spent over a decade tinkering with the drives and generators. Tweaking materials, dimensions, measurements, anything they could conceive to lengthen the operating time. The drives would inevitably consume themselves within moments of coming online. Angat made a design that would burn indefinitely, instead of cutting off after a watt or two. I think his first generator is still running, powering the equipment in his old lab as part of a museum exhibit. One of my life’s biggest regrets was that I never got to actually meet the little Filipino, he was rather elderly by then, and all sorts of demands were keeping me stateside when he announced the generator.

I found out about it after he sent me an email. Not just to me, mind, it was sent to thousands of people, so far as I know. It not only explained the concepts and provided schematics, but it also announced that the designs were totally open source.

My accountants and other official business types went crazy. Ashley Rice, our head accountant, actually cornered me in the garage after the email went out. It was a testament to how colorless she usually was that my now-rabid security let her get in the back of the car with me without a word. I did jump, I had been on my way to meet with some of our farm overseers when she had gotten in. I usually rode out there alone.

Scowling a bit at her, I just waited for Ashley to explain herself. Se didn't speak right away. She just tucked some strands of her dirty blond hair behind an ear and picked at the pills of wool on her cardigan until I made a little hand gesture. You know the kind, straight fingers, a bit of rolling motion, the one you use to give up right-of-way at a stop sign.

“What are we going to do?” Is all she said. Which wasn't helpful. I suppose I should have realized he was concerned about Angat's generator email, but I had a lot on my plate. I was going out to check on the farmers and see if they were happy with the new automated equipment, and whether they felt like livestock automation was reasonable. I had already made some deals to adjust salaries in our non-work force – basically, they agreed to drastically lower their salaries in exchange for room and board. We provided living space, utilities, and access to the products that the farms were producing, and they were happy for a much smaller stipend instead of full salaries. A few kept managing their own living arrangements and kept a normal salary, but most were fine. And Ms. Rice knew all this, so I was still at a loss.

But she was deferential to a fault, and I had to supply my side of the dialogue. “About what, exactly?”

“About the generator Angat invented.”

I was still out to sea. One of these days people are going to figure out that I am not nearly as smart they think I am. “I suppose we'll build a few. Hansen is looking into viability to power our own operations. Why aren't you talking to him?”

Ashley sighed at me. Me! Like I was a particularly dim child! Her brown eyes squeezed shut and she clearly did some sort of counting exercise. All because I didn’t understand why our accountant would care in the slightest about a generator that we didn’t own. That no one owned. “I did talk to him. He sent me to you. The question isn't about using them, it's about building and selling them. This is a huge opportunity.”

I leaned forward. Closed my eyes. Rested my head in both my hands. Elbows on knees. Without looking at her and doing my best to be as bland as her sensible flats, I said, “Ok. Pretend I know nothing. I'm a brand-new intern who can barely manage fourth-grade addition. What. Opportunity.”

“Angat gave us plans. He gave everyone plans. But, according to him, several parts require plasteel –“

“Plasma Steel,” I interrupted from under my hands. I hate that contraction.

“Plasma Steel, as well as Plasma Copper and the insulative variety of plast – Plasma Steel. Without using our materials, you can't build a functional generator. He gave the world the generator, but we're still the only ones who can build it.” I ventured to look up at her. She was literally rubbing her hands together while she talked to me.

“You know that our cash flow is getting restricted. Our building division is really the only part of the company still showing a profit. It's still more than enough to run everything, even your more eccentric organizational plans,” Ms. Tedious really didn't like all my non-productive workers. I still liked sleeping at night. “But even that's going to dry up soon. Our best projections only give us ten more years of high profits, followed by another ten years of gradually falling revenue. And that's all assuming no one steals the method to manufacture Plasma Steel or your other metals.”

“This is a totally different sort of product. It'll have the, well, some of the same problems as our other products, but powering the world could be incredibly profitable. It gives us the opportunity to sell a service, not just a product.”

I sat back up and looked out the window as I thought about this. “Let me get this straight. Angat – who has never filed a patent, so far as I know, is revolutionizing science. Again. These are the generators we've been hearing about for twenty years, that could provide essentially free energy. No fuel costs, no pollution, minimal maintenance work. He released his designs to the world, making it clear he wanted to benefit mankind.”

I turned to look directly at her, “And you want me to use our patents to grab a monopoly on it. Ashley, we haven't talked much about mission statements, or things like that, have we?”

“No sir, you're mostly content to know that bills are paid and taxes handled. But you've never failed to keep the company growing. It's been pretty amazing, watching you work. I mean, you barely have a marketing department but the internet is full of viral videos about plasteel stuff. I mean Plasma Steel. You don't have a lot of lobbyists either, but you've strong-armed the state governments a bunch. You...”

I cut her off again. I didn't mean for her to try and stick her nose up my backside. “No, not like that. What are we trying to get out of this company? Right now, Austin Beck is the least wealthy of us five owners. You know exactly what he's worth. He can get his scarred pinky finger on anything that money can buy in an hour, and he can usually get it heavily discounted because we make the damn thing. And other than land there isn't much he can buy that will even put a dent in his net worth.”

“You know that I've had people from the Fed meet with me. We accumulated so much cash so quickly, coupled with falling consumer demand, that they were worried about inflation.” Or was it deflation? Stagflation? Some sort of flation. I continued, “They literally asked us to find ways to re-invest the cash to keep things moving. That's half the reason I keep all the staff that you're always whining about.”

“So let's say I do corner the electric market with this generator. Use our patents and our secrets to maintain a monopoly on damn near everything. We get more money. More money that we can't spend on anything worthwhile anyway because we already make everything worthwhile. We get a market share that will evaporate in a few years when the market evaporates.”

“I'll be honest. I don't know what we're going to do when we stop being able to sell and build stuff. The day is coming, and I don't pretend it won't be hard. But frankly, I don't think that postponing it a few more years is going to be all that big a deal. At least it won’t be worse if it happens in ten years than if it happens in twenty.”

“But, even if those few years are disastrous to us, it won’t be worth the ill will. People will hate us if we profit off this. They'll hate us more than they hated the bandits who used to hold up cancer meds for everything a patient owned. I get shot at and blown up enough already. Let's please not make that worse.”

I tapped the tablet in her hands, “So, what are we going to do? Here's what we're going to do. Get back to Hansen and Alan. Get a design of each of the parts the generator needs and get started on molds. Have them make several, I imagine that people will want generators in all sorts of sizes. We're going to take advantage of the market, but we're only going to sell the bits and pieces, we're not going to sell the whole thing. And we're only going to sell at cost.”

“Yes sir.” She was back to her normal reserved self.

“Good. As long as your here, let’s go over the quarterlies.” She nodded and turned her little tablet back on. Which made the drive productive enough. Boring, dull, and tedious, but productive. And I’d been looking forward to music and a nap.

The generators turned out awesome. We built about four different varieties. A big one that could power a city. It was really big, it stood as big as a house all by itself. Another big one, which wasn't quite as big, could power most large buildings, like a hospital or office building. That generator could just about fit onto a semi-trailer. A third was small enough for a man to carry, maybe thirty or forty pounds, if I remember right. It was mostly used as a portable generator, and it just had a couple banks of plugs to attach whatever to. The last was very small, and could easily be built under the hood of a car, or inside a drone. Those were used to power all sorts of things. You could replace the battery in a car with a generator and it would run forever. No charging, no refueling.

All of that was huge, but I'll be honest. The application I loved most was in the Arcology. Akins had been tooling around with designs for a year now, ever since my first assassination attempt. He really went crazy when he realized that the generators meant we could build it independent of logistical connections. Unlimited power meant that that his designs didn't have to rely on external supplies. Unlimited power meant that he could include water reclamation techniques, extensive internal transportation, all sorts of repair drones. He included massive interior spaces meant to be filled with manufacturing or storage or whatever. There were other large spaces open to the outside, with little nooks and crannies that we could fill with little shops and social spaces. Several arenas built-in. Multiple towers filled with living space – the smallest apartment was a three-bedroom space with about twenty-five hundred square feet.

We assumed no more than 3 adults in one of those apartments, or two adults and four children, at most. We had larger spaces, too, theoretically meant for larger families. A few had built-in workspace, offices and hobby rooms and the like. We went to great lengths to build the towers so that they all had exterior access. Big balconies, mostly, all fitted with boxes and irrigation. We were pretty sure that allowing balconies to fill with greenery would make for a much more comfortable living space. We inverted some standard designs, too. The highest apartments were usually the smallest. By tapering towers, it was easier to make sure sunlight and exterior space were available on every floor.

The place was big enough that most of the spaces would be entirely interior, no matter what we did. One of the guys came up with some clever fiber optics and other tricks to bring some natural light into the deeper spaces. We mostly used that for areas set aside for offices, classrooms, and hospital rooms. We also used UV and full spectrum lamps in every single room. Full environmental controls – a tenant could control temperature, lighting (brightness and spectrum), and even humidity in their rooms.

It was pretty cool. I staked out my space in the building early. It was one of the little three-bedroom units, I planned on turning one room into a personal entertainment center, the other into an office space for private meetings, and of course, I slept in the third. Pretty much the only thing special was that it was close to one of the social promenades and to one of the entry garages.

The best part? Assembly required no human input beyond design and some paranoid oversight. Akins and our other engineers still didn't entirely trust the automated systems to follow plans. But the whole system was rapidly becoming fully automated. We owned a handful of iron mines now – drones excavated, concentrated the ore, and moved it to refineries. The refineries produced a range of standard bar sizes that we used to make Plasma Steel. More drones trucked the iron from the refineries to our manufactories. At each stage there'd be maybe one or two people not doing anything but waiting for something to screw up that required them to push a big red button, cease operations, and contact maintenance.

The refinery was the only place that ever had any real automation failures. But even those failures were predictable enough that before long other repair drones would fix any stoppages without input. AI was getting interesting.

I remember meeting the salesman for tech, about when our consumer product lines were starting to slow down. His company was called Technocore, which I found vaguely ominous, but their designs were almost as important to our growth as Plasma Steel and the other materials were. The salesman, a rumpled little guy who looked more like one of my blue-collar non-workers than the shiny types that usually tried to sell us stuff, came into my office with one of those unfinished-looking robot things.

This one walked remarkably smoothly on its two legs. Above the legs was a screen and some boxy-looking gear. No arms, no head.

I looked at him, saw the bot, and checked my calendar. He had twenty minutes, and I was going to go complain to my assistants. After we had started mass producing the robotic production arms and our construction drones, I had been inundated by robotic types wanting to sell their designs. We hired a few of the designers, but frankly we didn't need new walkers. But my assistants knew what they were doing. The bot was only there to provide a multimedia display that he could control. The walker bit was just a bit of melodrama to pitch the software.

I'll spare you the speech. Thechnocore specialized in heuristic AI. For the first ten minutes, the little guy spent his time running his hands through his hair, straightening his tie, and apologizing to me that their AI wasn't actually very smart. No singularity. They couldn't do anything new.

But they could do just about anything they were told to do, and do it perfectly after a very short learning period. His example actually made a lot of sense to me, although in my experience that means he was simplifying things to the point that they were meaningless.

Take a bipedal robot. Balancing on two pins is difficult. People do it automatically, though they forget that it generally takes two years of practice to figure it out, and often another decade or so to really work out the kinks. If you provide the proper servos, sensors, and gyros, it wasn't terribly difficult to design a program that could learn to operate the bot and keep it upright. Technocore’s heuristic programming could even learn to accommodate irregular footing, slipping, getting pushed, bumping into things, and so on. The programming even did better than people do – at a certain point a person's body generally just lets itself fall down so it can stand back up again, the bot will go to crazy and uncomfortable lengths to stay upright if that's what its programming demanded.

Now, give that same standing bot arms. It needs arms, right? A bot that can't pick things up, carry things, or manipulate things isn't super useful. That same programming can easily learn to use the arms while maintaining balance – pick up boxes, pull levers, whatever.

Now, here's the key. As amazing as Technocore’s AI is at maintaining its balance using the gyros and body position, it will never, ever, use an arm to hold onto a railing to help its balance. At least, it won't unless a programmer includes that as a tool for its task. It can use any tools in its programming to perfection, but will never think to pick up a different tool. The AI was simply incapable of looking outside the box. Even if a programmer tried to give it ways out of the box, all that actually happened was a program with a bigger box.

Obviously, most of what we wanted to do with AI was way more complex than just standing upright. It also took comprehensive and open-minded thinkers to avoid garbage-data problems, but that was what Technocore specialized in. We licensed their tech and hired consultants to try out the AI in our mines and asked them to maximize production. It worked great, and so we began to use it in more and more stuff. Eventually, we bought them out entirely, folding some of them into our research divisions and letting the rest work on synergy with our various departments.

Akins cheerfully incorporated the Ais in our arcology – automated restaurants, bartenders, materials distribution, even elevators that monitored and predicted load to minimize wait times. Meanwhile, Technocore’s competitors were doing the same thing everywhere. That's probably why the salesman wasn't the normal greasy type – his programs really sold themselves.

But before long, drones were farming, building, manufacturing, selling, stocking shelves, flying planes, cleaning homes, cooking, and even helping people get dressed.