At a Calgary technology conference this week, Margaret Atwood offered nearly three thousand gathered leaders a deliberately humble metaphor for artificial intelligence: a paring knife. In doing so, the 86-year-old author of The Handmaid's Tale placed the oldest of human questions at the center of a very modern conversation — not whether a tool is dangerous, but what it reveals about the people who wield it. Her message was neither alarm nor reassurance, but something rarer: an invitation to pay careful attention.
Margaret Atwood: AI Is a Paring Knife With Good, Bad, and Stupid Sides
We made it. We're responsible for it.
Mark
So Atwood is saying AI is just a tool. Doesn't that let the people who build it off the hook?
Mimi
Not quite. She's saying we made it, which means we're responsible for it. But she's also saying the consequences depend on use—and on accidents we didn't foresee. The paring knife cuts your finger sometimes no matter how careful you are.
Luke
Right, but that's where the metaphor breaks down a little. A paring knife is simple. AI is opaque. You can see how a knife works. You can't always see why Claude gives you the wrong answer.
Mimi
That's exactly why she brought up the Claude example. It's not just about misuse. It's about the technology producing confident-sounding wrong answers. That's a different problem than a knife.
Mark
She mentioned storytelling as the most powerful tool humans have. Is she saying AI is just the latest chapter in that story?
Mimi
More or less. Humans have always shaped information and authority through narrative. AI is a new amplifier for that. But the underlying human capacity to tell stories—to persuade, to influence—that's ancient.
Luke
But again, scale matters. A story told by one person reaches dozens. AI-generated content reaches millions instantly. That's not the same as a paring knife.
Mimi
No. But her point is that the responsibility still lands on us. We decide how to use it. We decide whether to check the facts. We decide whether to challenge false information.
Mark
So critical thinking is the real tool?
Mimi
It's the thing that has to accompany the tool. Otherwise you're just holding a knife in the dark.
Luke
And we don't know yet if people will actually do that work. That's still an open question.
The Pulse
- Atwood arrived at a high-stakes energy and technology conference not with predictions, but with a paring knife — deliberately deflating the mythology surrounding AI before the conversation could begin.
- A failed attempt to identify a murderer in a Father Brown episode exposed something unsettling: an AI that answered with full confidence while being entirely wrong, a gap between the appearance of knowledge and its substance.
- Beneath the anecdote runs a deeper tension — as AI-generated content grows indistinguishable from human communication, the cost of uncritical acceptance rises sharply.
- Atwood's response to this tension is not technological but humanistic: trace the origins of ideas, examine the evidence, and be willing to name what is false.
- The story is landing not as a warning against AI, but as a call to recover something older — the habit of asking who is holding the knife, and why.
At a Calgary technology conference this week, Margaret Atwood offered nearly three thousand gathered leaders a deliberately humble metaphor for artificial intelligence: a paring knife. In doing so, the 86-year-old author of The Handmaid's Tale placed the oldest of human questions at the center of a very modern conversation — not whether a tool is dangerous, but what it reveals about the people who wield it. Her message was neither alarm nor reassurance, but something rarer: an invitation to pay careful attention.
Margaret Atwood took the stage at Calgary's BMO Centre this week before nearly three thousand business leaders and investors, and immediately reached for the most ordinary object she could find. AI, she told them, is like a paring knife — useful, dangerous if mishandled, and capable of surprising you in ways you never anticipated. "We made it," she said simply, placing responsibility where she believed it belonged: with human beings, not with the technology itself.
Atwood has spent more than five decades writing speculative fiction — stories that use history and present-day patterns to map possible futures. That long view made her an unconventional but pointed presence at a conference devoted to the intersection of energy and technology. She brought with her a recent exchange that was both funny and instructive. Trying to recall the actual murderer in a Father Brown episode, she turned to Claude, Anthropic's AI assistant, for help. Claude responded with confidence — and named the wrong character entirely, the obvious red herring Atwood already knew was innocent. The exchange revealed a structural problem: the AI could not watch the episode, could only draw on its training data, and in this case that data was not enough. Authoritative tone, insufficient knowledge.
For Atwood, this sits within a much longer human story about information and power. Storytelling, imagery, and symbols have always been used to shape how people understand the world — AI simply extends that capacity to an unprecedented scale. Speech itself, she suggested, may be the most powerful tool humans have ever developed, and like AI, it is neither good nor bad by nature. What matters is intent, use, and whether anyone is paying close enough attention to notice the difference.
Her advice to the room was pragmatic rather than alarmed: examine the evidence, trace where ideas come from, and be willing to challenge what is demonstrably false. The paring knife cuts both ways. The question she left hanging in the air was not whether to pick it up, but whether the person holding it is actually watching what they are doing.
Margaret Atwood stood before nearly three thousand business leaders, investors, and entrepreneurs gathered at the BMO Centre in Calgary this week, and offered them a tool for thinking about artificial intelligence that was refreshingly unglamorous. She compared it to a paring knife.
Not a revolutionary technology. Not a threat to civilization. A paring knife. "It's got a good side. It's got a bad side. And it's got a stupid side you didn't anticipate," the 86-year-old author said during a fireside chat with Holly Ransom, co-founder of the Energy Disruptors: UNITE 2026 conference. The knife can slice apples for applesauce. It can become a weapon. It can slip and draw blood. AI, by her logic, follows the same pattern—a human invention whose consequences flow directly from human choice and human error, not from the tool itself. "We made it," Atwood said, a simple statement that placed responsibility squarely where she believed it belonged.
Atwood has spent more than five decades writing about the relationship between technological change, human behavior, and power. She has authored more than fifty books, including The Handmaid's Tale and its Booker Prize-winning sequel The Testaments. Her work is often described as speculative fiction—stories that draw on history and present developments to explore what futures might actually arrive. That perspective made her an unconventional but pointed choice for a conference focused on the intersection of energy and technology.
She illustrated her point about AI's limitations with a recent exchange that was equal parts instructive and funny. Atwood had been trying to remember who the actual murderer was in an episode of the television series Father Brown. She turned to Claude, an AI assistant made by Anthropic, for help. Claude could not provide the answer. Instead, it confidently named the character who had been presented as the obvious suspect—the very red herring Atwood already knew was innocent. The conversation that followed revealed something important: Claude, in the version she was using, could not watch the television episode directly. It could only work with information it had been trained on, and in this case, that information was insufficient. The exchange became a window into a serious problem. AI can produce responses that sound authoritative and complete without actually possessing the knowledge needed to answer a question correctly. Atwood even asked Claude about the origins of its own name, leading to playful speculation between them about whether it might have been named after the title character in I, Claudius. (It turns out Claude is named after Claude Shannon, the mathematician credited as the architect of the Information Age.)
For Atwood, AI sits within a much longer human history of shaping information and controlling how people understand the world. Societies have always used storytelling, imagery, and symbols to communicate ideas and establish authority. AI simply adds another powerful tool to that process, particularly through its capacity to generate and distribute information at scale. But she argued that the underlying human capacity to tell stories—the ability to speak, to create music, to make art—remains fundamentally important. These capabilities emerge remarkably early in childhood. Speech, she suggested, is probably the most powerful tool humans have ever developed. Like AI, however, storytelling is neither inherently good nor bad. Its consequences depend entirely on how it is used.
Atwood's advice to the assembled business and technology leaders was grounded in pragmatism rather than alarm. She encouraged people to examine the evidence behind what they are told, to trace where ideas originate, and to assess whether claims can actually be trusted. She also emphasized the importance of being willing to challenge information that is demonstrably false. This message carries particular weight in a world where AI-generated material is becoming increasingly difficult to distinguish from authentic human communication. Yet despite identifying real risks, Atwood did not frame technology or the future as something to fear. The paring knife cuts both ways. What matters is who holds it, what they intend to do, and whether anyone is paying attention to what actually happens.
Notable Quotes
It's got a good side. It's got a bad side. And it's got a stupid side you didn't anticipate.— Margaret Atwood, describing AI at the Energy Disruptors conference
We made it. AI is a human tool rather than something beyond our responsibility.— Margaret Atwood
Want the full story?
Read the original at Calgary.Tech ↗
Technology