This is a bit of a fun post, inspired by the thread here: https://www.linux.org/threads/whats-the-deal-with-using-ai-in-replies-in-topics.62602/. It's a bit long because it includes AI responses in full as they appeared on screen on this computer. The post endeavours to make a point about AI from an experience with it.
In the study of linguistics there's a well known example of an English sentence created by the linguistics scholar Noam Chomsky in his book "Syntactic Structures" released in 1957. The sentence is:
What the sentence exhibited was a conventionally correct grammatical structure, but had no logical meaning, or literally understandable meaning, because of the inappropriate relationships of meaning between the words. Ideas aren't usually green, and sleep is not usually furious etc. The point of the sentence was to indicate that grammar (the structure of sentences) is independent of semantics (the meaning of sentences). What would follow from that is that meaning is not dependent on particular structures in natural language.
The concept inherent in the example of that particular sentence, and what it purports to prove can be exhibited in any other sentence which has the same aspects of conventionally correct grammatical structure but without logical meaning or a literally comprehensible meaning. Such an example could be:
The challenge set for AI was: what can AI inform an inquirer about the characteristics and inferences associated with sentences like this?
When the sentence "faithless pious insects erupt somnambulantly" was written into the ChatGPT dialog box at first, this was the response:
Note that ChatGPT has described the sentence as a "phrase", but it is actually a sentence with nouns and verbs and adjectives all in the conventionally correct places for English to create a sentence.
Next challenge for AI was to enter the Chomsky sentence:
After having entered the sentence which ChatGPT had correctly identified, the second sentence which exhibits the same conceptualisation was entered:
After having correctly identified the Chomsky sentence, ChatGPT has been able to relate the information it was able to access, to the second sentence and identify the similarity of concept, and was correct in doing so.
As an addendum, the second sentence "faithless pious insects erupt somnambulantly" was entered into two other AI interlocutors with the results being similar to the initial ignorant response ChatGPT provided in the first response above:
In perplexity:
In gemini:
It appears that if AI has the information accessible to it, such as in the form of a text somewhere that it can retrieve, it provides that text and presents it using its internal language handling algorithms which are known to involve statistical processes. When it comes to conceptualisation which is a common characteristic of human intelligence, the AI is not so adept, to say the least. When AI is provided with "hints" or examples it can apply to text, it can use those hints and produce meaningful results.
There's nothing definitive about the above adventure and readers can draw their own conclusions of course
.
In the study of linguistics there's a well known example of an English sentence created by the linguistics scholar Noam Chomsky in his book "Syntactic Structures" released in 1957. The sentence is:
Code:
Colorless green ideas sleep furiously.
The concept inherent in the example of that particular sentence, and what it purports to prove can be exhibited in any other sentence which has the same aspects of conventionally correct grammatical structure but without logical meaning or a literally comprehensible meaning. Such an example could be:
Code:
Faithless pious insects erupt somnambulantly.
When the sentence "faithless pious insects erupt somnambulantly" was written into the ChatGPT dialog box at first, this was the response:
Code:
ChatGPT said:
That phrase feels like surrealist poetry—charged with contradiction and movement.
“Faithless pious” — a paradox. It suggests someone (or something) outwardly devout but inwardly empty, or ritual without belief.
“Insects erupt” — small, usually unnoticed lifeforms suddenly becoming overwhelming. It evokes decay, transformation, or suppressed truths bursting out.
“Somnambulantly” — sleepwalking. Unconscious, automatic, unaware.
Taken together, it paints a scene of blind devotion cracking open, releasing swarms of hidden realities—actions performed without awareness until something suppressed breaks through.
If you’d like, here’s a short poetic expansion:
Faithless pious insects erupt somnambulantly—
from cathedral walls of bone and borrowed creed.
They spill from the seams of whispered prayers,
crawling through incense-heavy air,
winged doubts beating against stained glass sleep.
No one wakes.
The hymns continue.
Would you like an interpretation in a philosophical, psychological, or purely abstract direction?
Note that ChatGPT has described the sentence as a "phrase", but it is actually a sentence with nouns and verbs and adjectives all in the conventionally correct places for English to create a sentence.
Next challenge for AI was to enter the Chomsky sentence:
Code:
ChatGPT said:
“Colorless green ideas sleep furiously” is a famous sentence coined by the linguist Noam Chomsky in 1957 in his book Syntactic Structures.
He created it to show that:
A sentence can be grammatically correct
Yet still be semantically nonsensical
Why it’s grammatically correct
The structure follows standard English syntax:
Colorless (adjective)
green (adjective)
ideas (noun)
sleep (verb)
furiously (adverb)
Everything is in the right order according to English grammar rules.
Why it’s meaningless
Something cannot logically be both colorless and green.
Ideas don’t literally sleep.
Sleeping furiously is contradictory.
Chomsky used this example to argue that grammar (syntax) operates independently from meaning (semantics), challenging behaviorist theories of language at the time.
After having entered the sentence which ChatGPT had correctly identified, the second sentence which exhibits the same conceptualisation was entered:
Code:
ChatGPT said:
“Faithless pious insects erupt somnambulantly” feels like a playful echo of Noam Chomsky’s famous sentence in Syntactic Structures.
Why it works grammatically
It follows a perfectly valid English structure:
Faithless (adjective)
pious (adjective)
insects (noun)
erupt (verb)
somnambulantly (adverb)
Syntactically, everything is in the correct order.
Why it feels strange
Faithless and pious contradict each other.
Insects aren’t typically described in religious terms.
Erupt suggests violent emergence.
Somnambulantly (sleepwalkingly) implies unconscious or dreamlike action.
Violent eruption and sleepy wandering clash semantically.
So, like the earlier sentence, it demonstrates that:
Grammatical structure does not guarantee meaningful sense.
That said, unlike Chomsky’s example, your sentence has a kind of surreal poetry to it. It almost suggests:
Creatures acting ritualistically without belief.
Instinctual life bursting forth unconsciously.
Contradictory identities moving in a dream-state.
After having correctly identified the Chomsky sentence, ChatGPT has been able to relate the information it was able to access, to the second sentence and identify the similarity of concept, and was correct in doing so.
As an addendum, the second sentence "faithless pious insects erupt somnambulantly" was entered into two other AI interlocutors with the results being similar to the initial ignorant response ChatGPT provided in the first response above:
In perplexity:
Code:
That phrase reads as intentionally surreal word salad rather than a standard expression or quote, so it does not have a fixed, recognized meaning in English.
Possible interpretation
Broken into parts, it can be read as:
faithless pious – a paradox combining devotion (“pious”) with lack of faith (“faithless”), suggesting hypocrisy, empty ritual, or spiritual confusion.
insects – often symbolic of small, swarming, unsettling forces (anxieties, intrusive thoughts, societal “bugs,” or corrupting influences).
erupt – a sudden outburst or breaking through a surface, like repressed feelings or hidden problems suddenly coming to light.
somnambulantly – “as if sleepwalking,” suggesting unconscious, automatic, or unexamined behavior.
Taken together, a reasonable paraphrase is: “Unfaithful believers, like swarming inner pests, burst out in a sleepwalking, unconscious way.” It evokes a mood of spiritual hypocrisy and unconscious, insect-like agitation spilling out from beneath a supposedly devout surface.
Were you aiming for lyrics, a poem title, or a prompt, or do you want help refining this into something more coherent?
In gemini:
Code:
Gemini said
That is a strikingly evocative string of words—it sounds like the opening line of a gothic surrealist poem or perhaps a very intense progressive metal album.
You’ve managed to pack a lot of heavy imagery into just five words. Let’s break down the "vibe" of that phrase:
The Anatomy of the Image
The Paradox: Combining "faithless" and "pious" creates an immediate tension. It suggests a devotion to nothingness, or perhaps a ritual performed without belief.
The Subject: Insects often represent the hive mind, instinct, or a creeping, unstoppable force.
The Action: "Erupt" suggests a sudden, violent breaking through the surface.
The Manner: "Somnambulantly" (sleepwalking) adds a layer of eerie, unconscious movement. It implies that this eruption is happening while the world—or the insects themselves—is dreaming.
If this were a writing prompt...
I could see this unfolding in a few ways:
High Fantasy/Horror: A civilization that worships a dead god discovers that their "prayers" have physically manifested as a swarm of mindless, shimmering locusts.
Modern Surrealism: An allegory for how ideas or trends spread through social media—mechanical, unthinking, yet strangely "holy" to those involved.
Abstract Poetry: Focusing on the phonetic texture of the words (the hiss of "faithless" vs. the "p" and "b" sounds in "pious" and "somnambulantly").
It appears that if AI has the information accessible to it, such as in the form of a text somewhere that it can retrieve, it provides that text and presents it using its internal language handling algorithms which are known to involve statistical processes. When it comes to conceptualisation which is a common characteristic of human intelligence, the AI is not so adept, to say the least. When AI is provided with "hints" or examples it can apply to text, it can use those hints and produce meaningful results.
There's nothing definitive about the above adventure and readers can draw their own conclusions of course

