Skip to main content

Hoosier Huddle

Trump: The Fluent F...
 
Notifications
Clear all

Trump: The Fluent Fool Problem

HHLurker's avatar
(@hhlurker)
Noble Member
“Summary: The "Fluent Fool" Problem
 
The fundamental danger of an LLM is that it separates fluency from intelligence. In human history, if someone spoke or wrote beautifully and confidently, it usually meant they knew what they were talking about. LLMs break this rule entirely: they can write beautifully while being completely wrong, tricking human brains into trusting them blindly.”
 
Sound familiar?

ReplyQuote
Topic starter Posted : 09/12/2026 9:07 am
CO. Hoosier
(@co-hoosier)
Famed Member

@hhlurker 

The opposite is also true.  We dismiss the wisdom of those who are not well- spoken. 

You are right about LLM.  This is a problem in law where LLM’s produce beautifully written briefs, but it’s all made- up bullshit. 


ReplyQuote
Posted : 09/12/2026 9:59 am
👍
1
Arthur Dent's avatar
(@arthur-dent)
Noble Member

It is how con men operate. Humans are terrible at spotting lies, there have been studies. By nature we assume a certain amount of truth.  Add confidence in, and we are just suckers waiting to be plucked. 


ReplyQuote
Posted : 09/12/2026 10:00 am
👍
1
CO. Hoosier
(@co-hoosier)
Famed Member

@arthur-dent 

Is AI scaremongering an example?  The Chinese are certainly capable of fomenting US fear for their benefit.  


ReplyQuote
Posted : 09/12/2026 10:39 am
Arthur Dent's avatar
(@arthur-dent)
Noble Member

@co-hoosier They are. I don't disagree; that is part of the problem with data centers. But as I suggest in the other thread, listen to people who work in the industry and don't attribute perfect brilliance to billionaires like Musk and Thiel. The people in the industry told the Titanic submersible guy what would happen, and it did. Being a billionaire didn't save him. Being confident didn't save him.

The risk is small, but it is there. The fact you argue there is no risk is incredible. 


ReplyQuote
Posted : 09/12/2026 10:59 am
HHLurker's avatar
(@hhlurker)
Noble Member

@co-hoosier 

Both interesting points. AI should empower less eloquent people to articulate themselves better. 

And I can imagine you have a perfect view into AI legal nonsense from your high perch. 


ReplyQuote
Topic starter Posted : 09/12/2026 11:05 am
CO. Hoosier
(@co-hoosier)
Famed Member

Posted by: @arthur-dent

The risk is small, but it is there. The fact you argue there is no risk is incredible. 

That’s not what I said.  I said the risk is centered on human beings, not in independent machines with zero ability to form intent.  


ReplyQuote
Posted : 09/12/2026 11:06 am
HHLurker's avatar
(@hhlurker)
Noble Member

@co-hoosier 

It’s clear the dangers  trace back to human factors. AI serves to amplify them in terms of volume, intensity and rapidity. 

That the US is having the discussion is reason for confidence. We have nimbleness to adapt quickly and confront AI threats. Probably the most dangerous threat is Russia or China outpacing us in AI advancement. They inherently have a chip on their shoulders about being less powerful than us. 


ReplyQuote
Topic starter Posted : 09/12/2026 11:13 am
HHLurker's avatar
(@hhlurker)
Noble Member

@arthur-dent 

I think the risk of AI being used for malevolent reasons is huge. And inevitable. I assume we’ll have catastrophic events before we take the risk seriously enough. 

I’m just with COH—it’s not AI run amok, it’s bad human actors. 

What happens when some random hacker blows our basic digital finance system to smithereens?


ReplyQuote
Topic starter Posted : 09/12/2026 11:19 am
Arthur Dent's avatar
(@arthur-dent)
Noble Member

@co-hoosier did the Titanic intend to sink? Did the 777s intend to crash? Did the Titan submersible intend to implode?


ReplyQuote
Posted : 09/12/2026 1:37 pm
Arthur Dent's avatar
(@arthur-dent)
Noble Member

Or a better example, July 2024 Crowd strike released a patch that brought down entire industries. The humans involved didn't see it, their software that evaluated patches didn't see it. Hospitals had systems down, airlines, everyone. Was there intent by anyone?

https://en.wikipedia.org/wiki/2024_CrowdStrike-related_IT_outages


ReplyQuote
Posted : 09/12/2026 1:49 pm
👍
1
Share: