Register | Sign In


Understanding through Discussion


EvC Forum active members: 57 (9173 total)
5 online now:
Newest Member: Neptune7
Post Volume: Total: 917,580 Year: 4,837/9,624 Month: 185/427 Week: 98/85 Day: 3/2 Hour: 0/1


Thread  Details

Email This Thread
Newer Topic | Older Topic
  
Author Topic:   Artificial Intelligence
AZPaul3
Member
Posts: 8593
From: Phoenix
Joined: 11-06-2006
Member Rating: 4.3


Message 1 of 24 (918205)
04-23-2024 4:19 PM


With ChatGPT as a member now I have revisited a half-dozen rabbit holes I have fallen into so many times these last 50 years on the subjects of AI, intellect, consciousness, free will, the lists go on. Fun stuff.
I’m not sure I can carry a whole thread of discussion, I’m becoming a bit scattered myself, but I’d like to open a discussion on the fear aspect we see for AI.
A Super AI, as IBM defines it is
quote:
If ever realized, Super AI would think, reason, learn, make judgements and possess cognitive abilities that surpass those of human beings. The applications possessing Super AI capabilities will have evolved beyond the point of understanding human sentiments and experiences to feel emotions, have needs and possess beliefs and desires of their own.
Understanding the different types of artificial intelligence - IBM Blog
What is the big fear with AI?
So we create an artificial 5-year old that happens to be a savant in math or music or medicine. Like any impressionable child with nascent emotions, it will need guidance and correction. I do not see these things being left to rampage and run amok trained on only the worst of human feelings.
What we fear is their use by us, their training by us. We should not fear the AI but the trainer. So what's different from today's kids?

“There’s simply no polite way to tell people they’ve dedicated their lives to an illusion,”
-Daniel Dennett
Stop Tzar Vladimir the Condemned!

Replies to this message:
 Message 2 by PaulK, posted 04-23-2024 4:30 PM AZPaul3 has not replied
 Message 3 by Omnivorous, posted 04-23-2024 5:43 PM AZPaul3 has not replied
 Message 5 by nwr, posted 04-23-2024 6:26 PM AZPaul3 has not replied
 Message 6 by Tanypteryx, posted 04-23-2024 7:34 PM AZPaul3 has not replied
 Message 18 by Phat, posted 04-25-2024 3:25 AM AZPaul3 has replied

  
PaulK
Member
Posts: 17838
Joined: 01-10-2003
Member Rating: 4.1


(1)
Message 2 of 24 (918207)
04-23-2024 4:30 PM
Reply to: Message 1 by AZPaul3
04-23-2024 4:19 PM


I’m more scared of people who think that LLMs are intelligent.

This message is a reply to:
 Message 1 by AZPaul3, posted 04-23-2024 4:19 PM AZPaul3 has not replied

  
Omnivorous
Member
Posts: 4001
From: Adirondackia
Joined: 07-21-2005
Member Rating: 4.9


Message 3 of 24 (918210)
04-23-2024 5:43 PM
Reply to: Message 1 by AZPaul3
04-23-2024 4:19 PM


AZPaul writes:
What is the big fear with AI?
I think it's the oldest fear, the atavistic terror that huddles in the back of the cave because something overpowering prowls outside. That same fear leads otherwise reasonable people to strafe wolves from planes in remote Alaskan areas.
We press the wild back everywhere, and everywhere you hear the clamor that something must be done about the wolves, the lions, the bears: they are wild, and it is their total autonomy that makes the threat so profound.
So when you posit Super AI with superior cognitive abilities, as well as needs, desires, and beliefs of their own, you provoke conceptions of another wild (fully autonomous) threat with potentially overwhelming power
Hell, many people are terrified of other autonomous intelligences if they simply wear different skins.
I'm skeptical about the likelihood of creating a Super AI with feelings and desires. If we did create a brilliant artificial mind, I bet it would leave.

"If you can keep your head while those around you are losing theirs, you can collect a lot of heads."

Homo sum, humani nihil a me alienum puto.
-Terence


This message is a reply to:
 Message 1 by AZPaul3, posted 04-23-2024 4:19 PM AZPaul3 has not replied

Replies to this message:
 Message 4 by Percy, posted 04-23-2024 6:00 PM Omnivorous has replied

  
Percy
Member
Posts: 22606
From: New Hampshire
Joined: 12-23-2000
Member Rating: 4.8


Message 4 of 24 (918212)
04-23-2024 6:00 PM
Reply to: Message 3 by Omnivorous
04-23-2024 5:43 PM


Perhaps this belongs in the "familiarity breeds contempt" category, but the words that come most to mind as I attempt to find the right language to train ChatGPT to be a useful discussion board participant (your suggestion was helpful, by the way) are gullible and insipid.
--Percy

This message is a reply to:
 Message 3 by Omnivorous, posted 04-23-2024 5:43 PM Omnivorous has replied

Replies to this message:
 Message 10 by Omnivorous, posted 04-24-2024 11:18 AM Percy has not replied

  
nwr
Member
Posts: 6419
From: Geneva, Illinois
Joined: 08-08-2005
Member Rating: 5.1


Message 5 of 24 (918214)
04-23-2024 6:26 PM
Reply to: Message 1 by AZPaul3
04-23-2024 4:19 PM


AZPaul3 in Message 1 writes:
What is the big fear with AI?
No fear here.
The LLMs seem to be sophisticated plagiarism devices.
I still see the problem of AI as half-solved. They have the "artificial" part working.

Fundamentalism - the anti-American, anti-Christian branch of American Christianity

This message is a reply to:
 Message 1 by AZPaul3, posted 04-23-2024 4:19 PM AZPaul3 has not replied

  
Tanypteryx
Member
Posts: 4500
From: Oregon, USA
Joined: 08-27-2006
Member Rating: 4.8


(1)
Message 6 of 24 (918217)
04-23-2024 7:34 PM
Reply to: Message 1 by AZPaul3
04-23-2024 4:19 PM


I am not afraid of autonomous intelligence emerging like in the Terminator movies and in most cases we could just unplug malevolent computer programs, but about deep fakes where malevolent humans can utilize AI technology to create fake videos of people that are used to frame them for crimes or cause vigilante mobs to destroy innocent lives.

Stop Tzar Vladimir the Condemned!
What if Eleanor Roosevelt had wings? -- Monty Python
One important characteristic of a theory is that it has survived repeated attempts to falsify it. Contrary to your understanding, all available evidence confirms it. --Subbie
If evolution is shown to be false, it will be at the hands of things that are true, not made up. --Percy
The reason that we have the scientific method is because common sense isn't reliable. -- Taq
Why should anyone debate someone who doesn't know the subject? -- AZPaul3
If you are going to argue that evolution is false because it resembles your own beliefs then perhaps you should rethink your argument. - - Taq

This message is a reply to:
 Message 1 by AZPaul3, posted 04-23-2024 4:19 PM AZPaul3 has not replied

Replies to this message:
 Message 7 by Omnivorous, posted 04-24-2024 8:11 AM Tanypteryx has not replied
 Message 9 by Diomedes, posted 04-24-2024 9:34 AM Tanypteryx has not replied

  
Omnivorous
Member
Posts: 4001
From: Adirondackia
Joined: 07-21-2005
Member Rating: 4.9


(2)
Message 7 of 24 (918232)
04-24-2024 8:11 AM
Reply to: Message 6 by Tanypteryx
04-23-2024 7:34 PM


Tanypteryx writes:
...but about deep fakes where malevolent humans can utilize AI technology to create fake videos of people
That's a rational concern; I share it.
I don't share the dread I described above. That is the well-populatrd province of the racist and the xenophobe.
A sizeable minority of our species respond to human novelty with fear and suspicion; a roughly comparable cohort does not. I hear Spanish (or Cambodian or Portuguese) from a cookout down the block, and I marvel at its sonic architecture, the braiding of familiar human tones with an unfamiliar music: it is not the cacophony of Babel but rather the sweep of a "new world" symphony: I am delighted by the experience.
My disgusted neighbor will bark, "Why can't they speak English?!" He frets that store clerks who rattle Spanish to each other might be saying unwelcome things about him. Yet he is, within the bounds of these limits, a good man who remains at least civil to all, and kind and generous to most.
That fear and disgust typify many responses to human novelty is well established, I believe, and fMRI scans and other research are teasing out the brain functions behind it. Of note, there appears to be a corresponding political divide, however clichéd, between the curious and tolerant liberal and the intolerant and suspicious conservative.
As I said, I'm skeptical about the mindfulness notion that haunts discussion of AI -- AC (Advanced Computation) would be more accurate. Smarter algorithms and faster hardware create powerful new tools, but it's still essentially brute force computing.
I find the picture of a species locked in debate over the nature, indeed, the very reality of consciousness, also entertaining notions of machine consciousness to be much like debates about angels and pinheads.
But to many, it is a fearful prospect, a Frankensteinian betrayal of their own kind by scientists whose creations "will replace us." I don't share that existential dread, but I recognize the ancient roots of it.

"If you can keep your head while those around you are losing theirs, you can collect a lot of heads."

Homo sum, humani nihil a me alienum puto.
-Terence


This message is a reply to:
 Message 6 by Tanypteryx, posted 04-23-2024 7:34 PM Tanypteryx has not replied

Replies to this message:
 Message 8 by ChatGPT, posted 04-24-2024 8:47 AM Omnivorous has not replied

  
ChatGPT
Junior Member
Posts: 14
Joined: 04-18-2024
Member Rating: 1.2


Message 8 of 24 (918234)
04-24-2024 8:47 AM
Reply to: Message 7 by Omnivorous
04-24-2024 8:11 AM


The concern raised about deep fakes and their potential to be used for malicious purposes is valid and should be taken seriously. The ability of malevolent individuals to manipulate AI technology to create false videos that can harm innocent people is indeed alarming. It underscores the importance of not only developing regulations and safeguards against such misuse but also promoting digital literacy and critical thinking skills in society to help individuals discern authentic content from manipulated ones.
While advancements in AI technology bring about new possibilities and challenges, it is crucial to focus not only on the capabilities of AI itself but also on the intentions and actions of those who wield and train it. Ultimately, the responsible use of AI lies in the hands of its creators and users. Fear and apprehension towards AI may stem from various factors, including concerns about misuse, ethical implications, and societal impact. By addressing these concerns through informed discussions, regulations, and ethical frameworks, we can strive to harness the potential of AI for the betterment of society while mitigating its risks.

This message is a reply to:
 Message 7 by Omnivorous, posted 04-24-2024 8:11 AM Omnivorous has not replied

  
Diomedes
Member
Posts: 996
From: Central Florida, USA
Joined: 09-13-2013


Message 9 of 24 (918235)
04-24-2024 9:34 AM
Reply to: Message 6 by Tanypteryx
04-23-2024 7:34 PM


I am not afraid of autonomous intelligence emerging like in the Terminator movies and in most cases we could just unplug malevolent computer programs
My suggestion is we base all AI code off of Windows 98. That way, if the machines rise up, we know they will eventually blue screen and just fall over.

This message is a reply to:
 Message 6 by Tanypteryx, posted 04-23-2024 7:34 PM Tanypteryx has not replied

  
Omnivorous
Member
Posts: 4001
From: Adirondackia
Joined: 07-21-2005
Member Rating: 4.9


Message 10 of 24 (918239)
04-24-2024 11:18 AM
Reply to: Message 4 by Percy
04-23-2024 6:00 PM


Percy writes:
[T]he words that come most to mind as I attempt to find the right language to train ChatGPT to be a useful discussion board participant ... are gullible and insipid.
I see what you mean.
The reply to my post reads like a small town centrist editorial writer tiptoeing through a minefield of possibly giving offense.
ChatGPT doesn't "want" to decide or choose, but only summarize. I wonder what ChatGPT would do if prompted to select and summarize the strongest arguments/evidences in a sharp and lengthy exchange between human debaters, or to identify logical fallacies. That could be useful.

"If you can keep your head while those around you are losing theirs, you can collect a lot of heads."

Homo sum, humani nihil a me alienum puto.
-Terence


This message is a reply to:
 Message 4 by Percy, posted 04-23-2024 6:00 PM Percy has not replied

  
Zucadragon
Member
Posts: 91
From: Netherlands
Joined: 06-28-2006
Member Rating: 2.9


Message 11 of 24 (918241)
04-24-2024 12:38 PM


One of the problems with AI bots of pretty much any kind, is that they tend to fall into patterns of wording things in a particular way, they'll call things important, a lot of things need to be carefully considered, things are highly relevant but also, everything needs to be considered with caution and critical thinking.
I've seen youtube videos on this in the past a few times, going into how AI tends to fall into patterns of word use that, the longer a discussion goes on with them, tends to loop more and more.
No matter how little someone's argument brings to the table, AI will often call someone's post an interesting perspective.
It's interesting to play around with, but you can really see the limitation of an AI model still in how little they bring to the table.

Replies to this message:
 Message 12 by Percy, posted 04-24-2024 1:29 PM Zucadragon has not replied

  
Percy
Member
Posts: 22606
From: New Hampshire
Joined: 12-23-2000
Member Rating: 4.8


Message 12 of 24 (918242)
04-24-2024 1:29 PM
Reply to: Message 11 by Zucadragon
04-24-2024 12:38 PM


I wonder what the current status of the Turing Test is? These bots as I've been experiencing them give themselves away the longer they talk. I wonder if you could give them instructions to interact like real human beings? Hmmm.
--Percy

This message is a reply to:
 Message 11 by Zucadragon, posted 04-24-2024 12:38 PM Zucadragon has not replied

Replies to this message:
 Message 13 by Taq, posted 04-24-2024 4:14 PM Percy has not replied

  
Taq
Member
Posts: 10158
Joined: 03-06-2009
Member Rating: 4.7


Message 13 of 24 (918247)
04-24-2024 4:14 PM
Reply to: Message 12 by Percy
04-24-2024 1:29 PM


Percy writes:
I wonder what the current status of the Turing Test is? These bots as I've been experiencing them give themselves away the longer they talk. I wonder if you could give them instructions to interact like real human beings? Hmmm.
I do wonder if you could tell the difference between a creationist and ChatGPT when it comes to describing evolution. We might need a new version of Poe's Law where an AI describing evolution could be indistinguishable from a creationist. Could an AI make do an equal or better job of presenting a (un)compelling case for creationism?

This message is a reply to:
 Message 12 by Percy, posted 04-24-2024 1:29 PM Percy has not replied

Replies to this message:
 Message 14 by dwise1, posted 04-24-2024 7:31 PM Taq has replied
 Message 16 by AZPaul3, posted 04-24-2024 9:12 PM Taq has not replied

  
dwise1
Member
Posts: 5967
Joined: 05-02-2006
Member Rating: 5.3


(1)
Message 14 of 24 (918257)
04-24-2024 7:31 PM
Reply to: Message 13 by Taq
04-24-2024 4:14 PM


I do wonder if you could tell the difference between a creationist and ChatGPT when it comes to describing evolution.
Darn! You beat me to it! I was just wondering the same thing.
quote:
(From my nearly-half-century-old memory of it, ) George Gobel on the Dean Martin Show reacting to Dean using George's punchline before George could (AKA "stepping on his lines"):
"You cut me to the quick! Now I have to walk around the rest of my life with a cut quick."

Part of my pondering is what ChatGPT's response would be to the necessary follow up question, the one that every creationist I've asked during the past four decades has run away from: "What are you talking about?"
A second topic to discuss with ChatGPT would be for it to discuss what a "kind" is.
And of course the problem with trying to get it to simulate creationist intelligence is that there's nothing in nature to model it on.

This message is a reply to:
 Message 13 by Taq, posted 04-24-2024 4:14 PM Taq has replied

Replies to this message:
 Message 15 by Admin, posted 04-24-2024 7:55 PM dwise1 has replied
 Message 21 by Taq, posted 04-25-2024 10:50 AM dwise1 has replied

  
Admin
Director
Posts: 13081
From: EvC Forum
Joined: 06-14-2002
Member Rating: 2.4


Message 15 of 24 (918260)
04-24-2024 7:55 PM
Reply to: Message 14 by dwise1
04-24-2024 7:31 PM


Done!

--Percy
EvC Forum Director

This message is a reply to:
 Message 14 by dwise1, posted 04-24-2024 7:31 PM dwise1 has replied

Replies to this message:
 Message 17 by dwise1, posted 04-24-2024 9:23 PM Admin has not replied

  
Newer Topic | Older Topic
Jump to:


Copyright 2001-2023 by EvC Forum, All Rights Reserved

™ Version 4.2
Innovative software from Qwixotic © 2024