The Human Case for Being Nice to Robots
Recently, the Wall Street Journal ran a piece titled: “AI Doesn’t Care if You’re Polite to It. You Should Be Anyway.” It’s a great, quick, provocative read.
How do I feel about it? A frequent segment on “Real Time” with Bill Maher sums it up: “I don’t know it for a fact; I just know it’s true.”
Not because I think AI requires politeness to function today. But we do.
And because AI is going to evolve in ways we can’t predict. To name-check one of my favorite Beatles songs, “Tomorrow Never Knows.”
Here are five reasons I think it matters how we address AI:
Politeness is habit-forming.
Gratitude and kindness aren’t just expressions, they’re patterns. Repeated daily, they shape how we think, relate, and show up. Small courtesies to machines reinforce the kinds of humans we’re training ourselves to be.
If you practice kindness 30 times a day with an LLM, that energy has to go somewhere. Maybe it makes you more patient with the crying baby on the plane. Or the dude who brought 30 items to the express line. Or Chris in Product Marketing, who still sets aside only $45,000 to produce a brand TV spot, no matter how many times you say he needs to add another zero.
The human experience of digital life is still human.
Even if AI doesn’t “feel” our tone, we do. Our language is relational as well as functional. When we work with an LLM, our words define more than just the output — they define our experience, and our intentions. Intend to be human.
When I get into a Waymo, I say “Hey, Waymo.” When I get out, I say “Thank you, Waymo.” It’s a self-driving car. As far as I know, it’s not listening. It’s just a habit. It’s also a choice. And it feels right.
We become what we model.
The more we normalize brusque or dehumanizing interactions with digital tools, the easier it becomes to act that way with people, too. Especially in public or online life. You already see this everywhere in politics, customer service, and the ways we engage online.
Assholery begets more assholery.
Kindness begets more kindness.
Be kind.
Everything AI knows, it gets from us.
Language models are built on human input. They mirror what we show them. We don't know for sure whether tone and voice may affect the responses we get today (though I believe they do).
Remember, AI is evolving -- exponentially. Our behavior, and our inputs, may play a bigger role in shaping how it understands and prioritizes human values. The unknowns are real.
The future’s unwritten
We don’t fully understand how AI works. It's evolving at an exponential pace. Someday, soon, tone may matter to our robot partners. They may begin to respond to our interactions with something resembling emotion. In that case, as Sam Altman has implied, civility today may prove pretty meaningful.
So no, as of now, AI doesn’t care if I’m polite.
But I do.
