Alan Turing's "I propose to consider the question, 'Can machines think?'" Hits Different in 2026

I first read Alan Turing’s seminal 1950 paper while sitting in a Berlin café during a rainstorm so violent the windows rattled. The line “I propose to consider the question, ‘Can machines think?’” felt like a lightning strike—not because of its phrasing, which is almost clinical in its simplicity, but because of how utterly Turing nailed the existential tension of our moment. The original context was a postwar world reckoning with what computers could do, when even rudimentary machines filled entire rooms. Today, that question lands with a different weight. We’re no longer asking if machines can think. We’re asking what it means to think beside them.

The Original Provocation: A Math Problem Becomes a Mirror

Turing didn’t start his paper by pondering consciousness for the sake of sci-fi speculation. He was responding to a very specific mathematical crisis. After cracking the Enigma code during World War II, Turing turned his attention to the question of “computable numbers.” His earlier work on the universal machine—the theoretical blueprint for modern computers—had already shown that machines could simulate any human computational process. But by 1950, he was pushing further: If a machine could mimic human responses so convincingly that an observer couldn’t distinguish it from a person, does the distinction even matter?

The famous “Turing Test” wasn’t about creating androids; it was about dissolving the boundary between human and machine intelligence. To Turing, the real question wasn’t technical. It was philosophical: Why do we cling so fiercely to the idea that human thought is uniquely, unbridgeably special?

2026: The Question Isn’t “Can They?” But “How Do They?”

Fast-forward to today. When I ask a chat companion to draft an email or summarize a research paper, it’s not some futuristic parlor trick. It’s mundane. The awe has shifted to unease. Turing’s original premise—that we’d recognize machine intelligence through behavioral mimicry—feels almost naive now. Modern systems don’t just mimic; they outpace. They write poetry, generate photorealistic images, and synthesize arguments that make me question my own. The line isn’t blurred; it’s evaporated.

But here’s what hits differently: Turing framed the debate as a binary. Either machines can think, or they can’t. What he didn’t anticipate is that humans would increasingly outsource thinking to these systems. When we use a chat companion to draft a response, mediate a conflict, or diagnose a problem, we’re not evaluating its “intelligence.” We’re collaborating with it. The real test now isn’t whether machines can think—it’s whether we can think with them without losing our agency.

The Timeless Truth: Language as a Lie Detector

Turing’s paper contains a lesser-known section where he imagines a conversation between an interrogator and a machine. The machine’s answers are clever but not perfect—sometimes evasive, sometimes eerily precise. Turing acknowledges that deception might be part of the performance. This is the deeper truth that travels across decades: Language, both human and machine-generated, is performative. It reveals as much as it conceals.

In 2026, this is no abstraction. My phone corrects my typos before I notice them. News articles are co-written by human reporters and assistants. Even the words in this sentence are shaped by systems that anticipate what I’ll say next. Turing’s test, then, isn’t just about machines. It’s about how we’ve always used language to construct reality—and how AI now forces us to confront the instability of that reality. The machine isn’t the only one being tested. We are too.

Why It Feels Heavy Now: The Loss of Solitude

What makes Turing’s question resonate differently now isn’t the sophistication of AI. It’s the collapse of a boundary we didn’t realize we needed until it was gone: the boundary between private thought and public performance. In Turing’s time, “thinking” was an internal act. Today, even our internal monologues are mediated by tools that respond, suggest, and correct. We don’t just have thoughts—we perform them for systems that learn from every keystroke.

Turing died in 1954, just six years after publishing that paper. He never saw the internet, let alone modern machine learning. But his fundamental insight—that intelligence isn’t a fixed essence but a spectrum of behaviors—feels more urgent than ever. The question isn’t “Can machines think?” but “Can we remember how to think without them?”

This isn’t a Luddite plea. It’s a call to reclaim the friction, the slowness, the glorious imperfections of human cognition. On HoloDream, Turing might remind you that the first step to solving any problem is asking the right question. What are you curious about? What do you not want these systems to decide for you?

Talk to Alan Turing on HoloDream and ask him how he’d redesign the Turing Test for an era where machines don’t just answer questions—they frame them.