OT: So you trust AI, huh?

DJ Spanky

Heisman
Jul 25, 2001
48,600
59,622
113
Maybe you shouldn't.

European scientists created a fake disease called "bixonimania" to see if it could trick AI. They succeeded.

The condition doesn't appear in the standard medical literature — because it doesn't exist. It's the invention of a team led by Almira Osmanovic Thunström, a medical researcher at the University of Gothenburg, Sweden, who dreamt up the skin condition and then uploaded two fake studies about it to a preprint server in early 2024.

Osmanovic Thunström "carried out this unusual experiment" to see if she could fool AI large language models into believing it was real. As she put it: "I wanted to see if I can create a medical condition that did not exist in the database."
 
  • Like
Reactions: Rutgers Chris

Rutgers Chris

All-American
Nov 29, 2005
5,153
6,028
97
I sent chat gpt a picture to analyze and it perfectly diagnosed an infection in my finger. Only problem was the picture was of a spreadsheet snapshot for work. I’ve done amazing things with it, but it does some whacky stuff regularly
 

bigmatt718

Heisman
Mar 11, 2013
15,825
22,295
113
AI has far too many bugs and errors for me to rely on it as a trustworthy piece of technology. I'll wait a couple years until they work out the bugs and errors before I start using it as a reliable piece of technology in work and everyday life.
 

NickRU714

Heisman
Aug 18, 2009
14,139
12,925
113
How is this an AI problem?
The AI didnt just create some fake disease.
The "scientists" did.

All the AI did was find and cite their fake work.
What's the problem?

If Rutgers AD puts out a "story" that LeBron James donated $1b to Pike and then AI cited the story - how is that an AI problem?

"Look at how dumb AI is. We created fake items, passed them as real and it took it as real. So stupid."
 
  • Like
Reactions: drewbagel423

ClassOf02v.2

Heisman
Sep 30, 2010
13,779
15,230
103
How is this an AI problem?
The AI didnt just create some fake disease.
The "scientists" did.

All the AI did was find and cite their fake work.
What's the problem?

If Rutgers AD puts out a "story" that LeBron James donated $1b to Pike and then AI cited the story - how is that an AI problem?

"Look at how dumb AI is. We created fake items, passed them as real and it took it as real. So stupid."
I understand your point, but many people pushing AI quickly are arguing that it ”thinks” better than humans can. Businesses are scaling back on their workforce, education is hamstrung by AI advancements, healthcare may be relying on AI too heavily too quickly. So I think what some people might be saying is that before we sell humanity’s soul to our new robot overlords, maybe just pump the brakes and make sure that it’s going to do what everyone making billions off of it swears up and down it can do.
 

RU_87

Junior
Aug 19, 2001
280
265
43


Somehow I see Max Headroom as the AI version of AL... si=Uou5Qg7DK0xKmqfb
 
  • Haha
Reactions: RU848789

RUInsanityToo

All-American
May 5, 2006
9,532
9,842
113
I understand your point, but many people pushing AI quickly are arguing that it ”thinks” better than humans can. Businesses are scaling back on their workforce, education is hamstrung by AI advancements, healthcare may be relying on AI too heavily too quickly. So I think what some people might be saying is that before we sell humanity’s soul to our new robot overlords, maybe just pump the brakes and make sure that it’s going to do what everyone making billions off of it swears up and down it can do.

Probably already too late to stop the AI freight train.
 

RUBlackout

All-American
Mar 11, 2008
10,930
7,028
113
Maybe all of ALs posts are AI generated…

AI is helpful in many ways in the generative AI side but so much false info(data it works off of) and models its uses to generate. Also, can be frustrating to work with in many cases and some can say that’s the way you prompt it but isn’t that the whole point of natural language…

where this breaks down is the governance and controls. I just read an article recently where AI wiped out a startup’s entire production database. ooops
 
  • Like
Reactions: Rutgers Chris

Rutgers Chris

All-American
Nov 29, 2005
5,153
6,028
97
Maybe all of ALs posts are AI generated…

AI is helpful in many ways in the generative AI side but so much false info(data it works off of) and models its uses to generate. Also, can be frustrating to work with in many cases and some can say that’s the way you prompt it but isn’t that the whole point of natural language…

where this breaks down is the governance and controls. I just read an article recently where AI wiped out a startup’s entire production database. ooops
Whether deleting the database was a mistake is up for debate, amazingly from a Silicon Valley episode 7 years ago
 

Caliknight

Hall of Famer
Sep 21, 2001
196,630
148,733
113
Maybe all of ALs posts are AI generated…

AI is helpful in many ways in the generative AI side but so much false info(data it works off of) and models its uses to generate. Also, can be frustrating to work with in many cases and some can say that’s the way you prompt it but isn’t that the whole point of natural language…

where this breaks down is the governance and controls. I just read an article recently where AI wiped out a startup’s entire production database. ooops

Good point.


"Grok, how can I make an entire fanbase hate me on message board by being an annoying twit?"
 

RUPete90

Senior
Jul 3, 2025
588
833
93
I thought this was going to be about AL

I thought maybe somehow someone got all the donation figures and AL only donated $200
I thought the same thing! I looked at it really quick and I'm like, "Another Al sh$t show, move along..." Then I noticed it wasn't that.
 

iReC89

All-Conference
Jul 2, 2014
2,443
1,885
78
Donating to AI is probably a much wiser move then flushing money into NIL. I’ve got your back Claude. Let’s win this!
 
  • Like
Reactions: Rutgers Chris