A.I. Is Not Sentient. Why Do People Say It Is?
-
- Posts: 137
- Joined: April 10th, 2022, 4:44 pm
A.I. Is Not Sentient. Why Do People Say It Is?
https://www.nytimes.com/2022/08/05/t...nt-google.html
I do not agree with this, but the article is worthwhile.
This is a good definition of intelligence, from the article: “The ability to learn — the ability to take in new context and solve something in a new way — is intelligence.”
-
- Posts: 137
- Joined: April 10th, 2022, 4:44 pm
Re: A.I. Is Not Sentient. Why Do People Say It Is?
The reaction to AI is that it has no soul and therefore cannot think.
-
- Posts: 502
- Joined: May 11th, 2021, 11:20 am
Re: A.I. Is Not Sentient. Why Do People Say It Is?
(The link is broken. See if you can create it again please.)Sunday66 wrote: ↑August 5th, 2022, 1:57 pm Robots can’t think or feel, despite what the researchers who build them want to believe.
https://www.nytimes.com/2022/08/05/t...nt-google.html
I do not agree with this, but the article is worthwhile.
This is a good definition of intelligence, from the article: “The ability to learn — the ability to take in new context and solve something in a new way — is intelligence.”
-
- Posts: 137
- Joined: April 10th, 2022, 4:44 pm
Re: A.I. Is Not Sentient. Why Do People Say It Is?
cannotAverageBozo wrote: ↑August 5th, 2022, 4:19 pm(The link is broken. See if you can create it again please.)Sunday66 wrote: ↑August 5th, 2022, 1:57 pm Robots can’t think or feel, despite what the researchers who build them want to believe.
https://www.nytimes.com/2022/08/05/t...nt-google.html
I do not agree with this, but the article is worthwhile.
This is a good definition of intelligence, from the article: “The ability to learn — the ability to take in new context and solve something in a new way — is intelligence.”
-
- Posts: 762
- Joined: July 19th, 2021, 11:08 am
Re: A.I. Is Not Sentient. Why Do People Say It Is?
It's the same with humans: Humans can’t think or feel, despite what the human common sense wants to believe.
-
- Posts: 84
- Joined: July 22nd, 2022, 7:22 am
- Location: Chile
Re: A.I. Is Not Sentient. Why Do People Say It Is?
What? So if you are chopping onion and you slice off a finger clean off and blood splatters everywhere, you will not feel it?
-
- Posts: 762
- Joined: July 19th, 2021, 11:08 am
Re: A.I. Is Not Sentient. Why Do People Say It Is?
Read again "It's the same with humans: Humans can’t think or feel, despite what the human common sense wants to believe."
-
- Posts: 2181
- Joined: January 7th, 2015, 7:09 am
Re: A.I. Is Not Sentient. Why Do People Say It Is?
Is this about the bloke who thinks the Google chatbox program (Lamda) he was working on has phenomenal experience?Sunday66 wrote: ↑August 5th, 2022, 1:57 pm Robots can’t think or feel, despite what the researchers who build them want to believe.
https://www.nytimes.com/2022/08/05/t...nt-google.html
I do not agree with this, but the article is worthwhile.
This is a good definition of intelligence, from the article: “The ability to learn — the ability to take in new context and solve something in a new way — is intelligence.”
He believes that Lamda has conscious experience because through his convos with it Lamda responds as we'd expect it respond if it had experience, ie that it would pass the Turing test.
Unfortunately because we don't understand the necessary and sufficient conditions for conscious experience, and it's private and can't be observed and detected by others, we have to resort to something like the Turing test (which basically tests similarity to us experiencing beings) to test for it. It's not reliable, but perhaps the best we can do is say the Turing test is indicative, for now at least.
A significant issue with Lamda imo is it's a learning chatbox program, designed to interact with humans in a human way, and fed by a wealth of internet examples of human knowledge and examples of how humans think and feel. So it being able to convincingly imitate a human it's talking to another human is an aspect of what it's designed and equipped to do, as part of its ability to answer questions better and its user-friendliness.
All that said the transcripts are intriguing!
https://insiderpaper.com/transcript-int ... bot-lamda/
-
- Posts: 137
- Joined: April 10th, 2022, 4:44 pm
Re: A.I. Is Not Sentient. Why Do People Say It Is?
One of the things mentioned in the article. AI has passed the Turing test ("the imitation game") long ago.Gertie wrote: ↑August 7th, 2022, 7:54 amIs this about the bloke who thinks the Google chatbox program (Lamda) he was working on has phenomenal experience?Sunday66 wrote: ↑August 5th, 2022, 1:57 pm Robots can’t think or feel, despite what the researchers who build them want to believe.
https://www.nytimes.com/2022/08/05/t...nt-google.html
I do not agree with this, but the article is worthwhile.
This is a good definition of intelligence, from the article: “The ability to learn — the ability to take in new context and solve something in a new way — is intelligence.”
He believes that Lamda has conscious experience because through his convos with it Lamda responds as we'd expect it respond if it had experience, ie that it would pass the Turing test.
Unfortunately because we don't understand the necessary and sufficient conditions for conscious experience, and it's private and can't be observed and detected by others, we have to resort to something like the Turing test (which basically tests similarity to us experiencing beings) to test for it. It's not reliable, but perhaps the best we can do is say the Turing test is indicative, for now at least.
A significant issue with Lamda imo is it's a learning chatbox program, designed to interact with humans in a human way, and fed by a wealth of internet examples of human knowledge and examples of how humans think and feel. So it being able to convincingly imitate a human it's talking to another human is an aspect of what it's designed and equipped to do, as part of its ability to answer questions better and its user-friendliness.
All that said the transcripts are intriguing!
https://insiderpaper.com/transcript-int ... bot-lamda/
-
- Posts: 9
- Joined: August 17th, 2022, 4:29 pm
Re: A.I. Is Not Sentient. Why Do People Say It Is?
The AI thing is blowing up these days, but people miss a critical point: AI has no desire. All life is driven by the in-built desires or hungers. We need to eat to survive, otherwise we have the subjective experience of starvation, fear, and ultimately death. This is what determines all our behaviors.
AI has none of these things. It has no objective experience of fear; if you leave an AI sitting there, it will not worry about it's power draining or its body rusting way unless you program it to worry. Even if you did, its merely the appearance of worry.
Demis Hassabis, the CEO of DeepMind, is arguably the greatest AI engineer in the world. He is the creator of the AI that beat the Go and Chess world champions, achieved nuclear fusion plasma control, and solved protein folding. He conclusively states in a recent interview that no sentient AI exists; its a tool, basically a fancy hammer, nothing more.
- Meta Island
- Posts: 107
- Joined: August 31st, 2022, 5:49 pm
- Favorite Philosopher: Anyone who makes me think
- Location: USA
Re: A.I. Is Not Sentient. Why Do People Say It Is?
For a robot to feel, it would need a self-image and a program to constantly monitor the condition of that self-image. Then reinforcement/growth of that self-image (for example, additional programmed abilities) would imitate positive feelings of power as diagnosed by the monitor, deletions from that self-image would imitate negative feelings of lessened power. Sounds vaguely familiar.Sunday66 wrote: ↑August 5th, 2022, 1:57 pm Robots can’t think or feel, despite what the researchers who build them want to believe.
https://www.nytimes.com/2022/08/05/t...nt-google.html
I do not agree with this, but the article is worthwhile.
This is a good definition of intelligence, from the article: “The ability to learn — the ability to take in new context and solve something in a new way — is intelligence.”
-
- Posts: 137
- Joined: April 10th, 2022, 4:44 pm
Re: A.I. Is Not Sentient. Why Do People Say It Is?
Why would an AI system need human feeling in order to operate?Meta Island wrote: ↑September 29th, 2022, 9:20 amFor a robot to feel, it would need a self-image and a program to constantly monitor the condition of that self-image. Then reinforcement/growth of that self-image (for example, additional programmed abilities) would imitate positive feelings of power as diagnosed by the monitor, deletions from that self-image would imitate negative feelings of lessened power. Sounds vaguely familiar.Sunday66 wrote: ↑August 5th, 2022, 1:57 pm Robots can’t think or feel, despite what the researchers who build them want to believe.
https://www.nytimes.com/2022/08/05/t...nt-google.html
I do not agree with this, but the article is worthwhile.
This is a good definition of intelligence, from the article: “The ability to learn — the ability to take in new context and solve something in a new way — is intelligence.”
- Meta Island
- Posts: 107
- Joined: August 31st, 2022, 5:49 pm
- Favorite Philosopher: Anyone who makes me think
- Location: USA
Re: A.I. Is Not Sentient. Why Do People Say It Is?
Hello Sunday66, I was just offering an answer as to whether a robot could be built with the ability to feel by defining feelings as states of power. I didn't mean to imply that a typical AI system needed to have feelings in order to operate.Sunday66 wrote: ↑September 29th, 2022, 1:00 pmWhy would an AI system need human feeling in order to operate?Meta Island wrote: ↑September 29th, 2022, 9:20 amFor a robot to feel, it would need a self-image and a program to constantly monitor the condition of that self-image. Then reinforcement/growth of that self-image (for example, additional programmed abilities) would imitate positive feelings of power as diagnosed by the monitor, deletions from that self-image would imitate negative feelings of lessened power. Sounds vaguely familiar.Sunday66 wrote: ↑August 5th, 2022, 1:57 pm Robots can’t think or feel, despite what the researchers who build them want to believe.
https://www.nytimes.com/2022/08/05/t...nt-google.html
I do not agree with this, but the article is worthwhile.
This is a good definition of intelligence, from the article: “The ability to learn — the ability to take in new context and solve something in a new way — is intelligence.”
My laptop’s OS monitors the laptop’s battery. When the battery runs low it dims my screen and pops up warnings - in terms of power, it is announcing that it needs (is hungry for) more power or the laptop will shut down, but that doesn't qualify as a feeling AI as you are using the term.
2023/2024 Philosophy Books of the Month
Mark Victor Hansen, Relentless: Wisdom Behind the Incomparable Chicken Soup for the Soul
by Mitzi Perdue
February 2023
Rediscovering the Wisdom of Human Nature: How Civilization Destroys Happiness
by Chet Shupe
March 2023