My search for the mysterious missing secretary who shaped chatbot history | Hacker News Reader

My search for the mysterious missing secretary who shaped chatbot history

theconversation.com

60 pointsseverine11 comments

Another small tidbit of lost computer history: The first person who lost a game to a computer was an unnamed Los Alamos laboratory assistant who lost a game of simplified chess against a computer program in 1956 [1]. The assistant's name seems to be lost to time.

[1]. https://en.wikipedia.org/wiki/Los_Alamos_chess#Los_Alamos_tr...

Already in 1950 in the Canadian National Exhibition, the visitors lost (it seems) most of their games of tic-tac-toe against Bertie the Brain, which was a custom computer: https://en.wikipedia.org/wiki/Bertie_the_Brain#Gameplay

Excellent! Upon further digging, I found out that a chess grandmaster, one Savielly Tartakower, was defeated by El Ajedrecista [1], a machine that could play chess endgames. This was 1951, so before the Los Alamos lab assistant, but a bit after the defeat of customers by Bertie the Brain.

The Ajedrecista had been around since 1912, so I would expect somebody must have been defeated by it in the years before 1951.

[1]. https://www.chessprogramming.org/El_Ajedrecista

In case anyone is wondering, the author hasn't found her yet; the article ends with:

>The blizzard is worsening. The announcement rings out that the campus is closing early due to the weather. The missing secretary’s voice still eludes me. For now, the history of talking machines remains one sided. It’s a silence that haunts me as I trudge home through the muffled, snowbound streets.

I feel the piece would have worked better backwards. Lead with the outcome that she wasn't found. Move through the effacement of the minor players (and potentially major contributors - the most interesting part). Close on the asking of how and why this happens. (Even the elimination of footsteps by falling snow could have been worked through it as a stronger metaphor)

Thanks for saving me from the clickbait. I think this sort of social archaeology is overrated; this was a person who was explicitly protective of her privacy.

I enjoyed the article though feel it kind of fell on its face with the abrupt ending after a reasonably decent middle bit.

I think about the truck driver who drove the first batch of Covid vaccines, or the driver of the first batch of h100's or whatever you deem significant. If they wanted 15 minutes in the spotlight and were denied, or if it was never offered.

The article is annoyingly written. Basically the author takes this little story from Weizenbaum's published accounts of "Eliza"—

> My secretary watched me work on this program over a long period of time. One day she asked to be permitted to talk with the system. Of course, she knew she was talking to a machine. Yet, after I watched her type in a few sentences she turned to me and said: ‘Would you mind leaving the room, please?’

—and wonders whether this story is true, and if so, what was Weizenbaum's secretary's name. It's not immediately clear to me that we should assume there was a secretary at all; Weizenbaum might have anonymized not only the participant's name but also her profession (and maybe her sex).

The author says she made an effort to find out who was Weizenbaum's secretary circa 1966, but was (completely?) unsuccessful:

> I work my way through Weizenbaum’s yellowed papers. Surely, among the transcripts, code print outs, letters and notebooks there will be evidence? There are some clues, reference to a secretary in letters to and from Weizenbaum. But no name.

> I broaden my hunt to administrative records. I look in department papers and the collections of Weizenbaum’s workplace, Project MAC – the hallowed centre of computing innovation at MIT. No luck. I contact the HR office and MIT’s alumni group. I stretch the patience of the ever-generous archivists. As my last day arrives, I still hear only silence.

The Weizenbaum archives are partially online. On page 149 of this 150-page collection labeled "SLIP, 1963 - 1967" ( https://dome.mit.edu/handle/1721.3/201706 ), it's indicated that on November 5, 1963, someone with the initials "jep" was taking dictation from JW.

Now, a single set of initials isn't remotely "identification" of JW's secretary (let alone identifying the participant from Weizenbaum's story, year unknown). But I feel like as a reward for reading all that, at least the author could have mentioned that she'd found those initials, and worked that into the tale she wanted to spin. As it is, it feels like she cared strictly more about spinning the tale than about finding the identity of the secretary. And if she doesn't care, why should the reader?

There is a whole project devoted to ELIZA here

http://findingeliza.org/

Also see

https://sites.google.com/view/elizaarchaeology/blog/weizenba...

Also see a giant archive of ELIZAs

https://sites.google.com/view/elizagen-org/original-eliza

Reply on news.ycombinator.com