Two AI characters discuss Roy Edroso's excellent Substack post, "Non-resistance is futile"
And they don't totally screw it up
As I describe further below, I tried for a second time running one of my posts by the imaginary AI podcasters at NotebookLM. Which led me to try it on another post more worthwhile than I can claim mine to be.
Imaginary podcasters comment on a Roy Edroso post
I wanted to reference an excellent Substack post by the irreplaceable Roy Edroso explaining why sniveling to rightwing authoritarians won’t encourage them to moderate their authoritarian views and practices. (Non-resistance is futile 03/10/2025)
So I decided to consult my imaginary AI podcasters at Google NotebookLM to hear what they had to say about it. Meanwhile, NotebookLM produced this audio discussion, which like all AI-generated content needs to be verfied before assuming it’s accurate.
Here are the imaginary podcasters discussing Roy’s column. They comment at the end that they are not trying to take a side on the controversy involved. (Why would imaginary commentators emphasize their neutrality?) I myself take the side that Roy is right on this.
In the spirit of things, I asked ChatGPT to generate a cartoon image illustrating a male and a female conducting an AI-created podcast. Which it obligingly did, with a fitting touch of TechBro creepiness. With both the human-like characters appearing distinctly Caucasian-like:
NotebookLM also provided a “Cast of Characters”for the post:
Roy Edroso: The author of the Substack column "Non-resistance is futile." He is a commentator on political and cultural issues, particularly critical of the Trump administration and its impact on academic freedom.
Minouche Shafik: The former President of Columbia University. She is criticized in the article for her actions against pro-Palestinian protestors and for appearing before Congress. The article mentions her successor but doesn't name them.
Donald Trump ("Tubby"): The President of the United States (at the time the article was written, presumably). He is portrayed as using the issue of antisemitism as a pretext for attacking institutions that don't align with his political agenda.
Unnamed Palestinian Student Protest Leader: A student at Columbia University who was reportedly handed over to ICE (Immigration and Customs Enforcement) by the university.
Unnamed Three Student Protesters from Barnard: Students from Barnard College (sister school of Columbia) who were expelled for protesting.
Ron DeSantis: Mentioned as an example of a "local fascist" who has targeted smaller schools for "wokeness."
Interesting that the AI commentators weren’t able to verify that Donald Trump is still the US President on March 10, 2025. Also that the Cast of Characters summary was “woke” in using “them” as a gender-neutral reference to Minouche Shafik’s successor. (The current acting president is Katrina Armstrong.)
NotebookLM also pulled a few “illustrative quotes,” which I verified were accurate, except that it inserted extra italics not in the original on the second one.
"They’re gonna fuck you up anyway, may as well fight back."
"But the Columbia thing shows that resistance is really self-interest, because we now know that if you ignominiously surrender everything to Tubby, he will still fuck you up."
"Resistance is not a romantic gesture; it’s survival."
"Trump cares about 'antisemitism' like he cares about 'free speech' and 'civil rights' — that is, solely as buzzwords to make his criminal enterprise look legit."
My NotebookLM imaginary podcasters on my own recent post about Ivo Daalder’s commentary
For the second time, I ran one of my posts through the Google NotebookLM AI function to create a two-person discussion of one of my posts. In this case the immediately previous one, focusing on Ivo Daalder's analysis of the Russia-Ukraine dilemma.
The first time I did this, the two AI characters talked about me as the author, similar to how in the item above they identify Roy Edroso as the author. But here they discuss the post by talking about Daalder's analysis without mentioning me. There is one point where they take something from a John Mearsheimer quote I used and talk about it as something Daalder said. Like always with AI: trust (some) but verify.
Near the end, they make a cryptic reference to Daalder's books being listed on the two imaginary characters' website. But no web address given. :)