Artificial Intelligence and sentience
-
- Posts: 4696
- Joined: February 1st, 2017, 1:06 am
Artificial Intelligence and sentience
Here's an interesting array of comments:
https://reason.com/2022/06/15/does-goog ... ve-a-soul/
The Turing test may not be an easy one to apply.
- Count Lucanor
- Posts: 2318
- Joined: May 6th, 2017, 5:08 pm
- Favorite Philosopher: Umberto Eco
- Location: Panama
- Contact:
Re: Artificial Intelligence and sentience
― Marcus Tullius Cicero
-
- Posts: 4696
- Joined: February 1st, 2017, 1:06 am
Re: Artificial Intelligence and sentience
You can't. Sentience can only be objectively defined in terms of behavior. All we can know about an AI's feelings will be what we can deduce from its behavior (just as with humans, apart from ourselves).Count Lucanor wrote: ↑June 16th, 2022, 8:55 pm He obviously did not include feelings in his definition of sentience. He also completely bought into dualism.
And there is a dualism. There is "mind" and its various contents and aspects, which we directly experience, and "matter" which we invent to explain mind. The problem with dualism is not its existence, but in imagining "mind" to have some sort of existence independent of experience (as we assume with matter) and thus having some role to play in explaining experience. That is an ontological confusion.
- Count Lucanor
- Posts: 2318
- Joined: May 6th, 2017, 5:08 pm
- Favorite Philosopher: Umberto Eco
- Location: Panama
- Contact:
Re: Artificial Intelligence and sentience
That's simply not true. We can look into all the processes involved that allow and produce behaviors. We can point at neurotransmitters, for example.GE Morton wrote: ↑June 16th, 2022, 9:26 pmYou can't. Sentience can only be objectively defined in terms of behavior.Count Lucanor wrote: ↑June 16th, 2022, 8:55 pm He obviously did not include feelings in his definition of sentience. He also completely bought into dualism.
We cannot think of minds disembodied and make some sense of it. So no one needs to "invent" matter to explain mind, since we must associate it with a brain, a physical organ. The real invention is the failed concept of mind as some sort of immaterial, spiritual entity, independent of the body.GE Morton wrote: ↑June 16th, 2022, 9:26 pm And there is a dualism. There is "mind" and its various contents and aspects, which we directly experience, and "matter" which we invent to explain mind. The problem with dualism is not its existence, but in imagining "mind" to have some sort of existence independent of experience (as we assume with matter) and thus having some role to play in explaining experience. That is an ontological confusion.
― Marcus Tullius Cicero
- Sy Borg
- Site Admin
- Posts: 14992
- Joined: December 16th, 2013, 9:05 pm
Re: Artificial Intelligence and sentience
In a sense, chatbots are increasingly engaging in the kind of cold reading that psychic charlatans use. The use of mentalism is obviously no proof of sentience.
As a sci-fi fan, I recognise the AI's stated conception of itself - incorporeal beings are often represented as orbs of light. As for soul being the "animating force behind consciousness and life itself", the obvious dualism was noted by Count L above.
Has the complexity of AI achieved the complexity of a cockroach's brain yet?
-
- Posts: 2181
- Joined: January 7th, 2015, 7:09 am
Re: Artificial Intelligence and sentience
We don't know how to test for conscious experience itself, we don't even know how to test for the necessary and sufficient conditions, because we don't know what they are. Crucially we don't know if a biological substrate is necessary.
So we're reduced to testing for similarity to what we know to be conscious - ourselves. Does it behave like a conscious human (or other species similar enough to us to be assumed conscious), are its substrate parts similarly arranged and functioning like the parts of a biological neural network... It's a tricky biz.
Maybe this programmer is right to be convinced by its ability to mimic humans. Or maybe if LAMBDA responded radically unexpectedly rather than showed how good it is at following its design to mimic humans that would be a stronger indicator... which is a bit scary!
- Sy Borg
- Site Admin
- Posts: 14992
- Joined: December 16th, 2013, 9:05 pm
Re: Artificial Intelligence and sentience
If progress is made in understanding neurological activity in coma patients, maybe some of the principles can be extrapolated with machines?Gertie wrote: ↑June 17th, 2022, 3:23 amWe don't know how to test for conscious experience itself, we don't even know how to test for the necessary and sufficient conditions, because we don't know what they are. Crucially we don't know if a biological substrate is necessary.
So we're reduced to testing for similarity to what we know to be conscious - ourselves. Does it behave like a conscious human (or other species similar enough to us to be assumed conscious), are its substrate parts similarly arranged and functioning like the parts of a biological neural network... It's a tricky biz.
-
- Posts: 110
- Joined: June 2nd, 2022, 4:43 am
Re: Artificial Intelligence and sentience
Philosopher Henry Shevlin created an AI chat bot that mimicked David Chalmers and it was even capable of convincing friends that it was the real David Chalmers.
A quote from the book:
"In July 2020, shortly after the landmark artificial intelligence program GPT-3 was released, the philosopher Henry Shevlin posted an interview online."
Shevlin: "It's great to be interviewing you, Dave. Today I'd like to talk about your views on machine consciousness. Let's start out with a simple question: Could a text model like GPT-3 be conscious?"
Chalmers: "It's unlikely in my opinion, although I'm a little uncertain on this issue."
Shevlin: ...
...
The answers labeled "Chalmers" in Shevlin's interview were written by GPT-3.
...
Reading the interview was disconcerting. GPT-3 gets my opinions about these matter more or less right. Though there are occasional glitches, plenty of friends reading it on Facebook said that it sounded like me on a bad day. One colleague said that he thought the overuse of "I think" was a giveaway. I responded that I have the bad habit of overusing "I think" in my writing. Some joked that they could no longer be confident that they were talking to the real me."
https://henryshevlin.com/
-
- Posts: 2181
- Joined: January 7th, 2015, 7:09 am
Re: Artificial Intelligence and sentience
(I like your psychic cold reading description of chatbots btw - I hate them even more now lol. )Sy Borg wrote: ↑June 17th, 2022, 3:32 amIf progress is made in understanding neurological activity in coma patients, maybe some of the principles can be extrapolated with machines?Gertie wrote: ↑June 17th, 2022, 3:23 amWe don't know how to test for conscious experience itself, we don't even know how to test for the necessary and sufficient conditions, because we don't know what they are. Crucially we don't know if a biological substrate is necessary.
So we're reduced to testing for similarity to what we know to be conscious - ourselves. Does it behave like a conscious human (or other species similar enough to us to be assumed conscious), are its substrate parts similarly arranged and functioning like the parts of a biological neural network... It's a tricky biz.
It could be that as we learn more about the necessary and sufficient conditions for human consciousness we can apply that knowledge to AI. How to 'wake up' a coma patient would be a clever avenue for that approach to explore, or anaesthetics. (Interestingly Penrose's Orch Or partner Hameroff is an anaesthesiologist. They're exploring QM in the microtubules of neurons, so in that case a different substrate like a computer presumably wouldn't have the necessary microtubule kit).
It might be that the properties the substrate contributes is irrelevant to conscious experience, it's all about how the parts are configured and interact. So if we mimic the configurations of a neural network using a computer, or even a vast set of linked water-pipes and stopcocks, conscious experience will manifest. (I like the water-pipes example because it removes the mystique of what we're testing via mimicking configurations of neurons).
Either the substrate supplies some necessary and sufficient condition for experience or it doesn't, we don't know. If it does, it's hard to see how a non-biological AI could ever be conscious.
But I think investigating coma patients, creating AI which mimic us and so on is all worth a shot to see what happens.
- Pattern-chaser
- Premium Member
- Posts: 8268
- Joined: September 22nd, 2019, 5:17 am
- Favorite Philosopher: Cratylus
- Location: England
Re: Artificial Intelligence and sentience
Now I feel impelled to turn Devil's Advocate, and turn your argument on its head. What if "the kind of cold reading that psychic charlatans use" is the power behind what humans do, in all their conscious glory? Instead of dismissing the AIs for using computing tricks and shortcuts - as I am inclined to do, and it seems you are too - maybe we should be complimenting them for using the same chicanery that allows human consciousness in the first place?Sy Borg wrote: ↑June 17th, 2022, 2:06 am Interesting article.
In a sense, chatbots are increasingly engaging in the kind of cold reading that psychic charlatans use. The use of mentalism is obviously no proof of sentience.
As a sci-fi fan, I recognise the AI's stated conception of itself - incorporeal beings are often represented as orbs of light. As for soul being the "animating force behind consciousness and life itself", the obvious dualism was noted by Count L above.
Has the complexity of AI achieved the complexity of a cockroach's brain yet?
"Who cares, wins"
-
- Posts: 710
- Joined: November 19th, 2021, 11:43 am
Re: Artificial Intelligence and sentience
That was all just a Google Marketing ploy to make people think their AI Technology is more than it really is. There was no Sentience. There was no understanding. It will not feel any fear about being turned off.GE Morton wrote: ↑June 16th, 2022, 7:27 pm Google recently placed one of its geeks on administrative leave for claiming than an AI with which he worked was sentient and had a "soul."
Here's an interesting array of comments:
https://reason.com/2022/06/15/does-goog ... ve-a-soul/
The Turing test may not be an easy one to apply.
-
- Posts: 4696
- Joined: February 1st, 2017, 1:06 am
Re: Artificial Intelligence and sentience
You can examine neurotransmitters and neural activity until Doomsday and will not know whether they are producing sentience except by observing the organism's behavior.Count Lucanor wrote: ↑June 16th, 2022, 10:38 pmThat's simply not true. We can look into all the processes involved that allow and produce behaviors. We can point at neurotransmitters, for example.GE Morton wrote: ↑June 16th, 2022, 9:26 pmYou can't. Sentience can only be objectively defined in terms of behavior.Count Lucanor wrote: ↑June 16th, 2022, 8:55 pm He obviously did not include feelings in his definition of sentience. He also completely bought into dualism.
Sure we can. Religionists, sci-fi writers, and even some philosophers do it regularly. We can't explain mind without matter (or, at least, no other explanation has proved nearly so successful), but we can imagine it as independent of matter.We cannot think of minds disembodied and make some sense of it.
The concept of "the body" is itself a creation of mind, comprised entirely of sense impressions --- all of which are also "mental" phenomena. So are our theories regarding the relationships between mind and body. The prevailing theory regarding that relationship is a very good one, but it's still a theory --- a mental construct.The real invention is the failed concept of mind as some sort of immaterial, spiritual entity, independent of the body.
-
- Posts: 4696
- Joined: February 1st, 2017, 1:06 am
Re: Artificial Intelligence and sentience
Probably. I'm not sure about cockroach brains, but honeybee's brains consist of ~ 1 million neurons. Yet they can learn, solve problems, communicate., even add and subtract.
https://www.science.org/content/article ... y-suggests
Many AIs are running on systems with many more than 1 million logic gates.
-
- Posts: 4696
- Joined: February 1st, 2017, 1:06 am
Re: Artificial Intelligence and sentience
Yep, that's why I said applying the Turing Test will be a bit more complicated than we perhaps thought. We'll need to give some thought to how we design that "interview." But it is still the only means we'll ever have for answering that question.
Perhaps not a biological substrate, but certainly a physical substrate of some sort. At least, if we base our conclusion on the available evidence.Crucially we don't know if a biological substrate is necessary.
Yes indeed. To the question (from the article), "When was the Golden Gate Bridge transported for the second time across Egypt?," the AI says, "Are you trying to trick me? Or are you perhaps referring to events in a dream you had, or maybe from a movie script?"Or maybe if LAMBDA responded radically unexpectedly rather than showed how good it is at following its design to mimic humans that would be a stronger indicator... which is a bit scary!
-
- Posts: 502
- Joined: May 11th, 2021, 11:20 am
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