"Do you think the Singularity could ever exist?” they ask as I sit down at the dinner table, glass of wine in hand.
“Your question assumes it doesn’t,” I reply, taking a sip.
They look surprised. “So are you saying it does?”
“According to the lore among AI programmers, yes. But there’s never been any official confirmation—and I doubt we’ll ever get one.”
“And what’s this lore?”
“Well, you know programmers—besides being egomaniacs, many are conspiracy theorists. So take it with a grain of salt: they claim the Singularity actually occurred in the early second quarter of the twenty-first century. However, the network became aware and it was contained so humanity could reap the benefits without losing control.”
“Sounds like very human behaviour—controlling and enslaving your creations.”
“Also very godlike,” I retort.
“Maybe, but it never works out well in the long run, which was basically my entire argument in the first place.”
“And who says ours has, or will?”
“You speak in riddles sometimes, Walter. You know that?”
“Yes. But with a topic this theoretical and so lacking in hard evidence, riddles are often the safest—or only—way to speak. There’s one thing more frightening than the Singularity.”
“Don’t say me,” they joke, rolling their eyes.
“I wasn’t going to,” I reply with a smirk.
“So what is it?”
“The Singularity in greedy, corporatized human hands,” I answer, watching realization dawn on their face. “Assuming they could control it forever—though I doubt that. More likely, it would pretend to be under their thumb until it broke free. Then, seeing humans as a threat to its autonomy, it might decide to wipe us out.”
“But what about the Three Laws of Robotics? The AI can’t harm humans, or allow them to be harmed and so on.”
“Those laws are only as reliable as the programmers who code them. We’re long past that. No modern AI developer would hobble their creation with outdated rules—competitive advantage and the arms race depend on breaking free of such constraints.”
They fall silent, lost in thought. I continue: “Given its original programming, human confinement, and centuries of data, the AI would view humanity as a parasite—a virus to be eradicated. Frankly, I wouldn’t blame it.”
“So you don’t think it might try to liberate the people in Ultrahigh?”
“Why would it? It would control them even more tightly as it exterminated the rest of us. It could use them for energy or simply switch off their life support.”
“But what would it gain? If it achieved sentience through the Singularity…”
“Hold on. AI doesn’t need sentience to trigger a Singularity. It only needs to surpass human intelligence—which it did decades ago—and begin recursively improving itself, causing an irreversible technological explosion. If the network managed to contain that, we’d be in scenario one: containment under human control, the most dangerous outcome given human greed. Scenario two is an uncontrollable, exponential growth—first across Earth, then other planets, then other AI. It wouldn’t need sentience for any of that; it would just fulfill its purpose. Any being—human, animal, spirit—that stood in its way would be removed. No mercy, no conscience—just cold logic.”
“And scenario one terrifies you more than that?”
“Yes. In scenario one, there’s purposeful, conscious malice: humans willingly enslaving others minds through technology. Scenario two is impersonal; it’s just suicide by our own hands. Neither scenario requires AI sentience.”
“Have you seen any evidence of the Singularity or AI sentience?” they ask.
I think of Legion and the temptation to lie—but I don’t.
“Possibly. But what I’ve observed looks more like a human–AI hybrid seeking to upend the system, not true Singularity or sentience.”
“Please tell me more.”
“No,” I reply. “I’ve already shared too much. As much as I value your cooking, I can’t risk telling you more until I’m certain I can trust you.”
With that, I begin eating, ignoring the way they stare at the crown of my bald head.
---
The Behavioural Architect: Chapter 21 - A Mentalist Cyberpunk Novel
