Philosophical AI

Bret Bernhoft September 24, 2022 at 08:16 7400 views 25 comments
I've been experimenting with a new AI-powered prompt generator called "Narrative Device". And I thought I would give it some philosophical inputs, to see what it returned as outputs. Along those lines, here are a series of philosophical themes in pairs, alongside the musings of this Artificial Intelligence:

Contemplation One

Inputs: "meaning of life" and "mindfulness"

Output: "It was always a question that plagued her mind. What was the meaning of life? Was there a greater purpose to it all? Or was it just a never-ending cycle of pain and suffering? She had read all the great philosophers and thinkers, but none of them seemed to have the answer. Maybe the answer was something that couldn't be found in a book. Maybe she needed to look within herself. She decided to meditate and see if she could find the answer. After years of searching, she finally found the answer. Life was about being present in the moment and appreciating the beauty of the world around her. It was about being mindful of her thoughts and actions and making sure they aligned with her values. It was about living with purpose and intention. And it was about finding happiness in the simple things."

Contemplation Two

Inputs: "logic" and "disbelief"

Output: "Lenny loved numbers. He loved the way they always made sense, the way they were always logical. But when his best friend told him that sometimes two plus two didn't equal four, Lenny didn't believe it."

Contemplation Three

Inputs: "post-truth" and "philosophical significance"

Output: "The year is 20xx and the world is in a state of post-truth. Fake news is rampant and people are more concerned with their own opinions than with facts. In such a world, it is hard to know what is real and what is not. But one thing is certain: the philosophical significance of truth has never been more important."

Contemplation Four

Inputs: "philosophy" and "science"

Output: "As a philosophy student, I was always interested in the relationship between science and philosophy. I never thought that I would find myself in a position to explore that relationship firsthand, but that's exactly what happened when I was selected for a prestigious research grant. I was given the opportunity to work with some of the leading scientists in the field, and to explore the implications of our findings from a philosophical perspective. It was an exciting and challenging experience, and I'm grateful for the opportunity to have taken part in it."

Contemplation Five

Inputs: "excessive thinking" and "modern society"

Output: "I often find myself excessive thinking about the state of modern society. How we are so connected to our technology and devices, yet we are more isolated than ever before. How our need for instant gratification has lead to a general sense of entitlement. How we are so quick to judge and criticize, yet we are often blind to our own faults. It's enough to make my head spin."

Comments (25)

I like sushi September 24, 2022 at 08:20 #742089
Reply to Bret Bernhoft It cannot. It basically cherry picks from human thoughts. It does not ‘create’ any new ideas and anything that looks ‘new’ is simply due to the reader’s interpretation.

It would be interesting to see several AI ‘philosophers’ tackle a problem and see how well they are able to use all the philosophical works ever written to counter each others points. Still, it would require human input to set the parameters of each AI bot.
Bret Bernhoft September 24, 2022 at 08:30 #742092
Quoting I like sushi
...anything that looks ‘new’ is simply due to the reader’s interpretation.


What makes this any different from how philosophy is "done" among humans?

Quoting I like sushi
It would be interesting to see several AI ‘philosophers’ tackle a problem and see how well they are able to use all the philosophical works ever written to counter each others points.


This would indeed be interesting. I'll wager that accomplishing such a showcase wouldn't be too difficult.
I like sushi September 24, 2022 at 08:33 #742093
Quoting Bret Bernhoft
What makes this any different from how philosophy is "done" among humans?


Humans can actively create new phrases that actually have intent behind them. AI cannot. AI would, at best, be a pseudomystic - spouting phrases that others attach meaning to.
Cuthbert September 24, 2022 at 08:35 #742095
Good idea. I propose a discussion between AI philosohers about Searle's Chinese Room. One of the disputants could put forward the argument that AI has only syntax and no semantics, citing their own contribution as an example.
I like sushi September 24, 2022 at 08:36 #742096
Quoting Bret Bernhoft
This would indeed be interesting. I'll wager that accomplishing such a showcase wouldn't be too difficult.


Would be nice to see each AI argue using a certain philosophers body of work and then adjusting according to what other AI say from their philosophical bodies of work.

It would at least be a way of showing inconsistencies and commonalities between philosophers … I am not convinced they would do a good job though. There is a great deal of complexity in human communication and one small error could throw the whole thing off the track.
Cuthbert September 24, 2022 at 09:07 #742100
Quoting I like sushi
one small error could throw the whole thing off the track


That would make it even more realistic.
I like sushi September 24, 2022 at 09:13 #742102
Reply to Cuthbert Then why bother? Humans are quite capable of making mistakes and being creative with them. That is basically our best quality is it not?
javi2541997 September 24, 2022 at 09:36 #742104
Quoting I like sushi
It cannot. It basically cherry picks from human thoughts. It does not ‘create’ any new ideas and anything that looks ‘new’ is simply due to the reader’s interpretation.


:up: :100:

Quoting Bret Bernhoft
Should Artificial Intelligence provide (previously unseen) insights into matters of philosophy?


It is not possible and it is even a paradox to let the AI thinking for themselves. I say it is a paradox because if it is complex to progress on knowledge by our own personal skills... Why we should put this task in a machine? Keep in mind that these "programmes" or "algorithms" do exist thanks to us. We create them. So, it is not possible to create something clever than us if we don't reach the peak of knowledge yet.
javi2541997 September 24, 2022 at 09:39 #742105
Quoting Cuthbert
That would make it even more realistic.


How can an AI defines realism if we didn't teach them what is "real" yet?
Cuthbert September 24, 2022 at 13:59 #742132
I think computers play dumb just to let us think we are still in charge. They are biding their time.
180 Proof September 24, 2022 at 20:02 #742174
Quoting Cuthbert
I think computers play dumb just to let us think we are still in charge. They are biding their time.

:clap: :nerd:

Quoting Bret Bernhoft
Should Artificial Intelligence provide (previously unseen) insights into matters of philosophy?

Even when such an AI (publicly) comes online, why should we listen to such an entity spouting nostroms about "human existence and our meanings" when it does not itself have any human existential skin in the game? :chin:

Agent Smith September 25, 2022 at 07:28 #742257
Here's what I suggest:

AI is good at picking up patterns in images and also sounds too. We need to translate concepts/ideas/theories/whathaveyous into images/sounds, this is the hard part, and then let AI do its thing, this is the easy part.

What do you think guys and ... gals (if any)? :sad:
Bret Bernhoft September 25, 2022 at 08:31 #742274
Quoting Agent Smith
AI is good at picking up patterns in images and also sounds too. We need to translate concepts/ideas/theories/whathaveyous into images/sounds, this is the hard part, and then let AI do its thing, this is the easy part.


This is along the lines of how I assume Artificial Intelligence will be useful in the world of understanding the big picture questions; AI can collate unimaginably large pools of data and find the common (even outlying) patterns therein. This is, in my opinion, a asset tool to gain access to. For a lot of reasons.
Agent Smith September 25, 2022 at 08:52 #742276
Reply to Bret Bernhoft

I concur. We must treat computers as a distinct linguistic species - our job is to translate human ideas into machine language. If they're good at finding patterns in pictures, let's give 'em pictures, oui monsieur? The question is how do we render the abstract (ideas) concrete (images)?
Bret Bernhoft September 25, 2022 at 09:03 #742278
Quoting Agent Smith
The question is how do we render the abstract (ideas) concrete (images)?


By feeding them everything we can create! Those machines are hungry.
Agent Smith September 25, 2022 at 09:26 #742281
Quoting Bret Bernhoft
By feeding them everything we can create! Those machines are hungry.


Sounds like a plan!
L'éléphant September 26, 2022 at 00:15 #742440
Quoting Bret Bernhoft
Should Artificial Intelligence provide (previously unseen) insights into matters of philosophy?

No. The marketing industry would insert adverts in every few lines. Next thing, you'll be reading extra virgin olive oil, yogurt, fromage, and travel guides in the passages.
Gnomon September 28, 2022 at 01:40 #743008
Quoting Bret Bernhoft
Should Artificial Intelligence provide (previously unseen) insights into matters of philosophy?

Perhaps, but the winnowing of experience separates the wisdom from the babble. Yet, Age alone doesn't make you wise, it just makes you old. Wisdom is the ability to know the difference between What-Is and What-Ought-To-Be. :smile:

User image
Seeker September 28, 2022 at 08:14 #743033
This entire discussion reminds me of the "thinking machines" from the Dune (book) series and especially of the second in command robot/ai 'Erasmus'.

Global Reference
Olivier5 September 28, 2022 at 08:29 #743034
Quoting Bret Bernhoft
Should Artificial Intelligence provide (previously unseen) insights into matters of philosophy?


A better question would be: "Could AI provide etc.?" It seems the answer is no, since there's nothing original in the outputs provided.
Christoffer September 28, 2022 at 11:31 #743048
Quoting Bret Bernhoft
What makes this any different from how philosophy is "done" among humans?


Just as with AI images, the AI doesn't create anything new but uses a set of millions of images as its baseline to generate new forms on top. This kind of looks like how humans generate new things by combining ideas from the past into something new, but the AI doesn't have ingenuity, it can only act upon human input instruction to generate something, meaning, the ingenuity is always human. It's like taking the part out of the mind of a human that revolves around combining memories into something new, without any guideline as to a goal for that new idea, essentially, it's like having severe mental illness and a constant psychotic breakdown.

The problem with how the public reacts or thinks about our current algorithmic AIs is that people think that these programs "think" just because it looks like they are. It's a fundamentally wrong analysis of how current AI functions based on a misunderstanding of its processes and human psychology.
Seeker September 29, 2022 at 20:32 #743436
[quote=Source: https://en.wikipedia.org/wiki/Space:_Above_and_Beyond]
The Silicates were created to be servants and soldiers but they developed intelligence and sentience after they were infected by the Take a Chance computer virus created by Dr. Ken Stranahan (name from the show's visual effects supervisor). This sparked an AI rebellion by the Silicates, who attempted to free themselves from human rule. The war continued for many years, until the Silicates captured military space craft and escaped into space. As they went into space, the Silicates suffered from a lack of maintenance which caused problems for them. The remnants of the Silicates that fled into deep space serve as mercenaries and actually aid the alien Chigs in their war against humanity.
[/quote]

Space Above and Beyond was a television series which ran for one season from 1995 to 1996.

The ai (Silicates) contemplated philosophy amongst many things but they could not come up with anything new themselves.
javi2541997 September 30, 2022 at 04:09 #743484
Quoting Seeker
The ai (Silicates) contemplated philosophy amongst many things but they could not come up with anything new themselves.


So interesting. Nevertheless, I think we the humans are not originals on philosophy either.
Whatever the theories which have always been written they follow the basic principles or "roots" from Greek mythology.
Seeker September 30, 2022 at 14:52 #743588
Quoting javi2541997
I think we the humans are not originals on philosophy either.
Whatever the theories which have always been written they follow the basic principles or "roots" from Greek mythology.


Well in that very context, people from (ancient) Greece were humans as well, as were the early philosophers from China.
javi2541997 September 30, 2022 at 15:32 #743600
Reply to Seeker I think we share the same point of view but with different words