Humanist Discussion Group, Vol. 36, No. 281.
Department of Digital Humanities, University of Cologne
Hosted by DH-Cologne
www.dhhumanist.org
Submit to: humanist@dhhumanist.org
[1] From: scholar-at-large@bell.net <scholar-at-large@bell.net>
Subject: Investigative Tool: Algorhithmic Failure (17)
[2] From: Bill Benzon <wlbenzon@gmail.com>
Subject: ChatGPT is more fun than a barrel of monkeys (57)
--[1]------------------------------------------------------------------------
Date: 2022-12-07 00:13:45+00:00
From: scholar-at-large@bell.net <scholar-at-large@bell.net>
Subject: Investigative Tool: Algorhithmic Failure
Willard
Have two minutes or so? Jill Rettberg has provided a video abstract to introduce
her paper "Algorithmic failure as a humanities methodology: Machine learning's
mispredictions identify rich cases for qualitative analysis” whose title
describes the process. The video ends with a lovely call out to scholars to
carry the research on, test on other data sets, and informs researchers that the
code is made available for repurposing.
<https://bigdatasoc.blogspot.com/2022/10/jill-rettberg-introduces-new-
paper-on.html>
New to me was the genre of video abstract: the scholarship shines.
François Lachance, Ph.d.
Life cannot be all told.
It is lived in the telling.
--[2]------------------------------------------------------------------------
Date: 2022-12-05 17:15:59+00:00
From: Bill Benzon <wlbenzon@gmail.com>
Subject: ChatGPT is more fun than a barrel of monkeys
Hi Willard & fellow humanists,
You’ve no doubt heard that OpenAI has released a new chatbot to the web,
ChatGPT. It’s based on its GPT-3 engine, first released in 2020, and it is lots
of fun. Over a million users have signed up for a free account. If you haven’t
done so already, you should. I’ve spent several hours playing with it and have
accrued a 30 page MSWord doc containing transcripts of my interactions. While a
lot of folks have been going at it adversarially, trying to get it to screw up
in one way or another, I’ve not engaged in the way, except in one minor
instance. My interactions have been cooperative. I’ve come to think of them as
piloting ChatGPT toward a goal.
I’ve written up some of these interactions in blog posts. Here’s one where I ask
it to explain a Jerry Seinfeld bit, which it does straight off:
Screaming on the flat part of the roller coaster ride: From GPT-3 to ChatGPT,
<https://new-savanna.blogspot.com/2022/12/screaming-on-flat-part-of-roller.html>
I chose that particular bit because I’d posted about it with the original GPT-3,
which was unable to cough up a reasonable account to the joke.
Here’s a post where I asked it to parody “Kubla Khan.” The first was so funny –
though defective as a parody in various ways, that I asked it for another.
How ChatGPT parodied “Kubla Khan” and pwned DJT45 at the same time,
<https://new-savanna.blogspot.com/2022/12/how-chatgpt-parodied-kubla-
khan-and.html>
Then I got ambitious, with a long post in which I guided it through a Girardian
reading of Steven Spielberg’s Jaws:
Conversing with ChatGPT about Jaws, Mimetic Desire, and Sacrifice,
<https://3quarksdaily.com/3quarksdaily/2022/12/conversing-with-chatgpt-about-
jaws-mimetic-desire-and-sacrifice.html>
Here’s a short coda:
ChatGPT goes to the movies, Jaws and Girard @3QD,
<https://new-savanna.blogspot.com/2022/12/chatgpt-goes-to-movies-
jaws-and-girard.html>
I know that for several years now you’ve been calling for ways to think about
AIs as, you know, not as approximations to people. They’re new kinds of beings
and we need to come to terms with such. I think that’s more important than ever
now that millions of people will be playing with very powerful engines with
little or now guidance and sophisticated reflection.
There are two extreme responses to this technology, both of which make the same
mistake. Both are trying to treat these systems as things we’re already familiar
with. One group we’re seeing the coming a super intelligent machines, perhaps
evil ones that will destroy humanity. We’re familiar with such devices. We’ve
seen them in the movies. The other group says there’s nothing there; it’s just
statistics on steroids. That’s a line Chomsky’s been pushing since the 1950s.
Nothing there either. So we’re faced with a failure of intellectual imagination.
This is NOT the way to sail the uncharted seas to lands unknown.
Best,
Bill B
_______________________________________________
Unsubscribe at: http://dhhumanist.org/Restricted
List posts to: humanist@dhhumanist.org
List info and archives at at: http://dhhumanist.org
Listmember interface at: http://dhhumanist.org/Restricted/
Subscribe at: http://dhhumanist.org/membership_form.php
