Hacker News Reader

Comment by burnte | Hacker News Reader

Except it WAS intended as a chatbot... as a tool "for research into human-machine conversation and the important cognitive processes of interpretation and misinterpretation".

It's like saying the guy who invented the screwdriver didn't invent it because he was REALLY just trying to build a cabinet. The reason a tool was invented does not mean the tool isn't what the tool is.

View on HN

But .... it sounds like it was was? The original MIT SLIP version (yes SLIP not LISP) had a framework for embedding dynamic modules - so could do an early version of "learning." It feels like it was indeed more a framework for testing out different kinds of chatbot logic - maybe a meta-chatbot?

The famous DOCTOR script was just one example to demonstrate the tools's capabilities, and not the main point of ELIZA.

[Edit: the paper suggests the DOCTOR script is "ELIZA" not ELIZA, i.e. the limited version that became popular, not the full version that was built)

It at least establishes that it wasn't intended as a public chatbot. To extend your analogy: it's as if the inventor of the screwdriver threw away his tool afterwards and someone found it in his trash and mass manufactured it. The cabinet maker wasn't trying to make a [new standard tool for how to drive screws] (screwdriver) but was making a tool [to drive screws for just one project] (screwdriver).

"the world first chatbot was not intended to be public."

More accurate, less interesting.

The “public” distinction feels anachronistic when applied to early 1960s software.

This was a time before networks and software distributions. The idea that software could be standardized and shared instead of written for a particular computer was not widespread.

So it’s a bit disingenuous to say that Eliza wasn’t intended as a public chatbot, when practically no software was intended to be public.

Yeah. That's actually one of the things that makes this story so interesting: It took place exactly at the dawn of the inter(arap)net, and physically close to BBN (which was the arpanet implementor), so there was lots of sharing between MIT and BBN, in particular, McCarthy put a lisp on BBN's PDP-1, and Bernie Cosell's Lisp ELIZA, which he did from the algorithm in Wiezenbaum's CACM paper, was the one that became public, so everyone thought that ELIZA had originally been implemented in Lisp, which was wrong!

It was intended as a chatbot in the sense that it provides an appearance of conversation when it’s basically a simple parlor trick. The technology of ELIZA wasn’t the interesting part, it was how people convinced themselves it actually had intelligence.

POSIWID.

Reply on news.ycombinator.com