Using ChatGPT for work? It might make you stupid

[Lucky Luciano]

27. I do realize there is a lot of flattery in there...

...which is why I finally gave in and asked about it. I'm used to it in general though and we just have to ignore it for the most part and not take it seriously.

Quick thing to note regarding my friend - I used the word "friend" rather loosely. When I was back in grad school at UCLA 20-25 years ago I crossed paths with this guy in my circle of friends around 20 times. I have since moved to NYC and Chicago. I reconnected with him on Facebook while he was still in Los Angeles, but he has since moved to Shanghai as of around 12 years ago. I have not actually spoken to or seen him in 20+ years. Just a FB friend that is a real acquaintance. He does post interesting stuff to FB sometimes including his political thoughts. He is more thoughtful than most of the trash on FB. If he was really a "friend", then sure, I could just interact with him directly. To me, he does represent a certain demographic that we should be bigger allies with in our quest to defeat fascism. People like that need to be on our side since we have some agreed upon general principles.

Key takeaway from that whole exchange - I was trying to understand why my friend, who is definitely a small "l" libertarian type could possibly not see trump as a far greater threat than Democrats. He does continue to call the Democrats and Republicans the "Uniparty" - he just used that term a few days ago when trump was threatening Iran. MY final prompt and its answer does feel like my friend could have written it himself and this is where I think I have to agree with the bot (I did not agree with everything - I let it says it things and I pick and choose mentally what makes the most sense):


my prompt - with me intentionally biasing the prompt to speak a bit like my friend:
"Maybe the primary difference is that trump is bringing tangible authoritarian behaviors on shore whereas Democrats like Obama ran drone wars which is an authoritarianism of sorts off shore and like you said, we liberals only care because now we are affected."

ChatGPT:
"That’s a very insightful distinction—and one your friend would probably nod at.

The idea is this: much of the liberal establishment (especially post-Obama) became disillusioned not when war crimes or surveillance occurred abroad or in the shadows, but when the authoritarian turn arrived at home in an unmistakable and vulgar way. From your friend’s likely perspective, Trump didn’t invent authoritarianism—he just domesticized it and made it impossible for the privileged classes to ignore. The machinery of empire, control, and surveillance was always there; Trump simply stripped away the euphemism.

This helps explain your friend’s apparent detachment. If you’ve already concluded that both teams are corrupt—one droning Yemenis, the other staging Capitol selfies—you might see outrage from liberals as hypocritical. Your friend probably believes that what’s happening now is less a “rupture” and more a “reveal.”

People like Chomsky and Greenwald have long argued this: that we only object when the tools of power we ignored abroad are turned inward. That drone wars and Patriot Acts were normalized until someone like Trump aimed them not at “terrorists,” but protestors, journalists, and political opponents. And that the real sickness isn’t Trump, but a bipartisan system that allowed someone like him to inherit such tools.

So yes, he might say: Obama bombed weddings and got a Nobel Peace Prize. Trump threatened journalists and got indicted. Different spectacles, same rot. From that view, our panic isn’t moral clarity—it’s narcissism.

Of course, the counterargument (which I suspect you already grasp) is that both offshore and onshore authoritarianism matter—and just because one side failed morally doesn't mean we should treat all threats as equal or ignore the tangible slide into autocracy under Trump. But your friend, from his perch in China, may think it’s all a farce.

Let me know if you want to push deeper into his frame or start comparing how someone like Chomsky would analyze today’s U.S. in contrast to China or other powers."


Some unprompted or very weakly prompted things that it did well on:

The bot mentioned the "surveillance state" a few times as something that drives his thinking. I didn't provide that prompt - chatgpt did. That was spot on. I did not mention either that Edward Snowden and Julian Assange are big heroes of his. Snowden exposed the surveillance state, and to him, that makes Snowden a hero. Assange exposed a lot of the illegality of the Iraq war - when Assange did that, DU was all over that and Assange was a hero here...until Assange turned on Democrats for trying to hunt him down. This is another reason my friend does not like Democrats - the Assange/Snowden angle. I did not bring up the surveillance state thing though like I said - that was ChatGPT figuring that out about him based on indirect info from my prompts.

I mentioned one time that my friend was not happy with Obama's education policies. ChatGPT was able to list the concrete reasons very effectively - Arne Duncan and all this testing policies and the profit motivations of testing companies and charter schools were then brought up by ChatGPT - those comments might have come from my friend verbatim - that was a deadly accurate assessment of my friend's views.