Swaine’s World
Don’t Anthropomorphize AI
Plus Tech Tales and Poetry Corner
Like Swaine’s World? Scroll to the bottom of this post to tip the author or to subscribe.
About the Picture
The pictures at the top of these posts rarely have anything to do with the content of the post. Since this blog is Swaine’s World, I try to show some aspects of my world, mostly scenes around my home. This particular picture was taken by Levi Thomas at It’s a Burl, a local establishment.
Don’t Anthropomorphize AI
The headline on the CNN story reads “An author says AI is ‘writing’ unauthorized books being sold under her name on Amazon.” I’ll give them credit for putting the word “writing” in quotes, but it’s not enough. The way we talk about AI tools matters, and far too often we’re doing it wrong.
To be clear: the point of the CNN story is valid. An author discovered that books are being sold on Amazon under her byline that she did not write (and is not getting paid for, presumably). She said it was clear that they were AI-generated texts. “ I have so much content available online for free, because I’ve been blogging forever, so it wouldn’t be hard to get an AI to mimic me,” she said. She has every right to be alarmed, and stories like this should set off alarms for all of us.
I just want to emphasize the importance of the phrasing we use when we talk about AI tools. It’s incredibly tempting to anthropomorphize them, to talk about them taking away jobs or spreading disinformation or “writing” books. And the danger is that when we do this we lose sight of the responsible party. These are tools, and they will be used to do all sorts of things, good and bad. We need to be aware of who uses them, for what purpose, for whose benefit or harm.
In the case of that CNN story, I understand the writer’s challenge. On the one hand, you want to report on what appears to be a criminal act. (But I’m not a lawyer.) On the other hand, you want to point out what these AI tools are capable of, and this author’s experience is a startling example of this. So you say that AI is “writing” books. But the actual (seemingly criminal) act was not perpetrated by a piece of software. Some as yet unnamed person used a tool to commit a (possible) crime. And nowhere in the story is there any reference to an attempt to identify the guilty person. The person responsible. I find that disturbing.
I think I’m writing this advice for myself as much as for anyone. Don’t anthopomorphize AI. These are tools. Used by people. Only by focusing on who is using them and for what purpose can we assign responsibility.
This is a theme that has come up in recent discussions I’ve had with some experts in the AI world. I’ll be drawing further on those discussions in upcoming posts. Meanwhile, here’s some computer history.
Tech Tales
The Tech Tales series is a history of the computer, lightened with semi-fictional dialogs and delivered in bite-sized chunks over the course of 2023. Here’s this week’s offering.
Up to now, this new personal computer industry consisted basically of tech hobbyists and unemployed engineers scraping together a little cash and starting seat-of-the-pants businesses. Apple had been no different; Jobs and Woz had hired friends and relatives and worked out of the Jobs garage. But the Markkula investment was something new: somebody with cash and cred was willing to invest serious money. Apple signalled its intention to reach out to mainstream America with a full-color ad in Playboy magazine.
At first, the customers and the new companies entering the market were from the same pool of electronics hobbyists, but soon, triggered in part by Apple’s aggressive marketing, non-technical people were buying these early personal computers. They bought them mostly to play games, but unless they started buying them for business purposes, the market would be limited.
So these same hobbyists started writing business software.
Alan Cooper and Keith Parsons had been working on a version of the Basic language and decided that writing software for businesses would be more lucrative. They created different programs for payroll, accounts payable, receivable, and general ledger, and saved them to disks and sold them by mail under the name Structured Systems Group.
Scene:
Alan and Keith are in the kitchen, which also doubles as the Structured Systems Group executive office. Alan is at the kitchen table, coding their new General Ledger program. Keith is putting disks with their Accounts Receivable program into Zip-loc bags to mail to customers. The following dialog, or something close to it, occurs. A name may have been changed to protect the innocent.
Keith: Another hundred bucks. Ka-ching.
Alan [The phone rings, and rings some more] Keith, are you gonna that?
Keith: It’s probably another order. Gina is on it.
Alan: I’m pretty sure I saw Gina just now sunbathing nude in the backyard.
Keith: Yeah, but she’s got the extension with her.
The Age of Innocence
Something similar was happening on the East Coast. Some Georgia Tech graduates started selling Altair microcomputers, discovered that they needed to supply their customers with software, of which there there wasn’t much, and decided to write their own. Nobody knew what the price should be, so they priced it in the high hudreds of dollars, like minicomputer software. Surprisingly, customers went for it. These early business software developers profited by their own and the market’s innocence.
That same innocence extended to standards for what the computers and the software should do, and to matters of quality. The customers knew they were pioneers on the bleeding edge, and accepted sometimes mediocre performance. The personal computer market now had adequate software for business purposes, but no more than that.
A breakthrough came from a daydreaming Harvard MBA student. The program he dreamed up would give personal computers capability as yet unavailable on mainframe and minicomputers.
VisiCalc
In the spring of 1978, Dan Bricklin labored over financial calculations and daydreamed about a magic computer that would display a grid of numbers in the air in front of him so that he could gesture — or something — and change the numbers, whereupon the magic computer would give him the answer to his question.
Bricklin kept returning to this magic computer in his mind, getting more realistic about its abilities and starting to see how they could be realized. He mocked up a simple program in Basic on a terminal on the B-school time-sharing system and tried out some of the ideas. The result was encouraging enough that he borrowed an Apple ][ and wrote a more powerful program. This one was able to display rows and columns of numbers on the Apple ][ screen, and he could move among the rows and columns using the arrow keys that Woz had conveniently provided for game play. He added some simple arithmetic capability. It was starting to look more like the magic computer he had dreamed of.
At this point Bricklin stepped back from programming and started to design in earnest. He drew state diagrams showing the complete state of the program and how it changed with any key press. As he worked through the logic of the program, he realized that the full implementation would require a kind of programming reserved for games, and would require much more efficient code than Dan knew how to write. He needed someone who could program in assembly language to squeeze out all the efficiency that you get only by programming close to the bare metal. As it happened, he knew a hot-shot programmer named Bob Frankston.
He was an MBA student, so he naturally worked up a business case for the program. And he realized that he was no better equipped to market the program than he was to implement the design. But he also knew a recent Harvard MBA named Dan Fylstra who had just started a software company selling software out of his apartment. The three got together at Joyce Chen’s restaurant in Cambridge and Fylstra agreed to distribute the program. Shortly after, Bricklin and Frankston met at a different restaurant and formed a company. Frankston would implement the product.
Scene:
Dan Bricklin is explaining his idea for a spreadsheet program to Bob Frankston. If you were a fly on the wall of the Cambridge attic where Frankston worked, you might hear something like this:
Dan: So I can move to any cell with the arrow keys.
Bob: There are only two arrow keys. You need four directions.
Dan: That’s one challenge, yes. And every cell can hold data or a formula.
Bob: And that formula can make this cell dependent on the values in other cells, like a cell that holds the sum of a column.
Dan: Right.
Bob: So that value, that total, needs to be automatically recalculated when any of those other cells changes.
Dan: You’ve got it.
Bob: Dan, most business software today just puts strings of text on a monitor. You’re talking about fully two-dimensional interaction with the screen. And that recalculation will take a lot of processor power. We’d be pushing these microcomputers beyond what we’ve seen them do.
Dan: So you’re saying it’ll be hard?
Bob: I’m saying it’ll be fun. Let’s do it.
When it was released for the Apple ][, it took off immediately. Hundreds of thousands of copies were sold. It was a runaway success. And it raised the level of business software on personal computers. With VisiCalc, personal computers were no longer toys.
VisiCalc was the first piece of software that sold computers. Analyst Ben Rosen called it the software tail that wags the personal computer dog. People bought Apples for business because they wanted VisiCalc.
It behooved this nascent industry to rethink the role of software.
To be continued.
The Poetry Corner
The Veils
While wandering the woods of words I chanced
To stumble over something Basho said —
The shallow stream lays bare its stony bed —
And light upon the truth Salome danced
When she had shed the trappings of pretense.
And captured by the swaying of her hips
Like curving hulls of sail-stripped drifting ships
Against the rocks devoid of all defense,
I longed to ride the waves, a one-man crew,
And one by one to strip away the veils,
And ride the current’s course devoid of sails,
The craft at last unleashed and steering true.
While pondering if I would have the will
I turned my collar up against the chill.
From my first collection of verses, cleverly titled First Verses. You can download First Verses for free.
Tune in next week for more adventures in tech history and light verse.
