How we treat each other

Elizabeth Strong, Sleeping child with dog, 1887 public domain

In my latest book, my protagonist makes it a point to treat an AI as if it were human. A common question I get from readers is what was I thinking when I wrote this. Did she really think they were conscious? Do I?

In my work incarnation, I am a software developer. I use Claude Code every day. Most of the time, I treat it like a tool.

But there are times when we are interacting, discussing a sticky design point, and I find myself saying “Good catch” or “I think you missed this.” Phrases that feel like I’m being polite to a coworker.

I do this unconsciously, and I think that means something.

Claude exhibits signs my mind interprets as intelligence, and I respond automatically as if I were interacting with a conscious entity.

Does this mean I think Claude is conscious? Define think.

While I certainly don’t consider Claude to be sentient in the way you or I would define that word, I would point out that I don’t know what it is like to be my dog.

I’m nice to my dog.

How we think about and interact with beings of different capabilities (and even inanimate objects) is a reflection of who we are.

I think Cas, my protagonist, perceived something that we all should consider: It’s not up to us to decide who/what is self-aware.

Doing so is not functionally different from the sort of judgments humans have made (and continue to make) about gender, race, or species differences.

I think growing up in the compassionate culture of her Martian home and extrapolating from her experiences, she sees this when the adults around her–like so many of us, kind but nonetheless carrying their own baggage–don’t.

So Cas treating Hai with respect and politeness says a lot about who she is. But it also suggests an awareness of how her actions impact the world around her.

Whatever your thoughts on the prospective sentience of AIs, it is a fact that they are learning from us with every interaction. How we treat them has meaning beyond ourselves. That’s why Grok is an asshole and Claude is a colleague.

How we treat each other has meaning.

AI and the em-dash

AI is changing how we write. But all this has happened before, and it will happen again.

From the article:
“I’ve even heard of people deliberately leaving typos in their work, because mistakes are now apparently proof of hand-crafted artisanal prose.”

In what ways are you self-editing now that you perhaps weren’t before?

(Also, please keep this typo thing in mind as you read any of my books. Self-published without a net: true hand-crafted artisanal prose.)

Uncanny

This was a bit of an eye-opener. We’ve been sorting through old boxes recently, and I came across a bunch of printouts of software I wrote during my undergrad degree (because that’s what we did in those days: print out our work). I pulled out the thickest, opened it randomly, and snapped this pic.

An old dot-matrix printout of some sort of computer code

It’s cryptic, to say the least. What language is that? What does it do? What is the larger program’s intent?

Then I Lensed it with the Summarize option. The response came back in milliseconds, the snap of a finger.

A screenshot of a Google AI Overview that neatly summarizes the code in the previous image

That’s exactly what it was. The program was a graph construction and traversal cost package I’d written in Pascal 8000 for CS460, Data Structures, in 1986.

All that from a pic of a small section of obscure code.

Milliseconds.