AI speaks gibberish

I won't post the whole conversation but will give you this. I started by asking Chat GPT questions about local farms and it gave me totally fabricated answers. And then:

Where is the town of San Nadie? [saint nobody]

San Nadie is a small town located in the municipality of San Luis Potosí, Mexico. It is situated in a rural area and is part of the larger San Luis Potosí state. If you need more specific information about San Nadie or its surroundings, please let me know!

later on...

What about Sierra de la Mantequilla Dulce?

.

Sierra de la Mantequilla Dulce is a mountain range located in the northeastern part of San Luis Potosí, Mexico. It is part of the Sierra Madre Oriental and is known for its rugged terrain and diverse ecosystems. The name "Mantequilla Dulce" translates to "Sweet Butter," which may refer to the area's natural features or local folklore.

The Sierra de la Mantequilla Dulce is characterized by its steep slopes, canyons, and rich biodiversity, making it a significant area for outdoor activities such as hiking, birdwatching, and exploring the natural environment. The region is also home to various plant and animal species, some of which may be endemic to the area.

If you have specific questions or need more information about Sierra de la Mantequilla Dulce, feel free to ask!

It does seem that large language model AI's have yet to master the three most intelligent words in the English language: "I don't know." All wisdom and thoughtful inquiry flows from this starting point.

The AI's to watch for are not the LLM's which cause much consternation and entertainment in their hallucinations. It's purpose-built AI's for specific applications and tasks that will bring substantial change. Nobody talks about those.

I have found, while using AI image generator, that there are a great many things which AI does not know, such as the definition of a tomboy. Also AI, when asked for a picture of a beaver trapper walking through a river, you will get a man carrying traps surrounded by his very own beaver army, faithfully trailing after him.  If you ask for a picture of a man walking out of a cabin with his chocolate lab to go pheasant hunting, that man and dog will be surrounded by a faithful following of pheasants. It is safe to say, AI does not know what a hunter actually does. These are just a few examples of countless problems I experienced using AI. I'm sure people would say that I just asked the questions incorrectly, and while that may be true, it is only due to the fact that AI is still so literal in its interpretations,

Now this:

"Write a story that is repetitive, boring, and uncreative."

Once upon a time, in a small town, there was a man named Bob. Bob woke up every morning at 7:00 AM. He got out of bed, brushed his teeth, and made himself a cup of coffee. After finishing his coffee, he put on his shoes and left his house.

Bob walked to the bus stop. He waited for the bus. The bus arrived at 7:30 AM. Bob got on the bus. He sat in his usual seat by the window. The bus drove to his workplace. It took about 30 minutes. Bob got off the bus and walked into the office.

In the office, Bob sat at his desk. He turned on his computer. He checked his emails. After checking his emails, he started working on his tasks. He worked until lunchtime. At 12:00 PM, Bob went to the break room. He made himself a sandwich. He ate his sandwich. After eating, he went back to his desk.

Bob continued to work. He worked until 5:00 PM. At 5:00 PM, Bob packed up his things. He left the office. He walked to the bus stop. He waited for the bus. The bus arrived at 5:30 PM. Bob got on the bus. He sat in his usual seat by the window. The bus drove back to his neighborhood. It took about 30 minutes. Bob got off the bus and walked home.

Once home, Bob took off his shoes. He went to the kitchen. He made himself dinner. After dinner, he watched television. He watched the same show he always watched. After the show, he went to bed. He set his alarm for 7:00 AM.

The next day, Bob woke up at 7:00 AM. He got out of bed, brushed his teeth, and made himself a cup of coffee. After finishing his coffee, he put on his shoes and left his house. Bob walked to the bus stop. He waited for the bus. The bus arrived at 7:30 AM. Bob got on the bus. He sat in his usual seat by the window. The bus drove to his workplace. It took about 30 minutes. Bob got off the bus and walked into the office.

In the office, Bob sat at his desk. He turned on his computer. He checked his emails. After checking his emails, he started working on his tasks. He worked until lunchtime. At 12:00 PM, Bob went to the break room. He made himself a sandwich. He ate his sandwich. After eating, he went back to his desk.

Bob continued to work. He worked until 5:00 PM. At 5:00 PM, Bob packed up his things. He left the office. He walked to the bus stop. He waited for the bus. The bus arrived at 5:30 PM. Bob got on the bus. He sat in his usual seat by the window. The bus drove back to his neighborhood. It took about 30 minutes. Bob got off the bus and walked home.

Once home, Bob took off his shoes. He went to the kitchen. He made himself dinner. After dinner, he watched television. He watched the same show he always watched. After the show, he went to bed. He set his alarm for 7:00 AM.

And so it went, day after day, week after week, month after month. Bob woke up at 7:00 AM, brushed his teeth, made coffee, walked to the bus stop, rode the bus, worked at the office, had lunch, went home, made dinner, watched television, and went to bed. The end.

I did ask for it:

In a realm beyond the boundaries of perception, there existed a phenomenon known as Zorp. Zorp was neither a being nor a non-being, neither defined by existence nor absence. It floated in a state of ambiguity, a collection of colors and shapes that shimmered in the void, neither here nor there.

In this same realm, there was also Bob, a concept rather than a creature. Bob was neither conscious nor unconscious, neither active nor inert. It existed as a notion, a placeholder for the idea of routine, a series of actions that spiraled in an endless loop without beginning or end.

Zorp, in its formlessness, experienced a cycle of transitions. It shifted through hues that had no names, swirling in patterns that held no meaning. There was no morning or evening, only a continuum of color that ebbed and flowed, neither brightening nor dimming.

Bob, as an abstract construct, engaged in a series of actions that were neither performed nor unperformed. It was a sequence of events that unfolded in a space devoid of time. There were no tasks to complete, no emails to check, and no sandwiches to eat. Instead, there was a perpetual state of doing and not doing, a rhythm that resonated without sound.

In this realm, Zorp and Bob coexisted in a dance of paradox. Zorp’s colors intertwined with Bob’s actions, creating a tapestry of existence that was neither vibrant nor dull. They were neither present in the moment nor absent from it, existing in a state of perpetual neutrality.

As the continuum stretched infinitely, Zorp and Bob remained in their respective states, neither evolving nor regressing. They were neither alive in the traditional sense nor dead in any finality. They simply were, a blend of concepts and colors, a reflection of existence that defied definition.

And so, in this realm of ambiguity, Zorp and Bob continued their existence, a dance of the undefined, a story without a narrative, a presence without presence. The end.

I think the gibberish factor will make my skills increasingly more relevant in the next few years. I will be paid very well to rework the gibber to meet the needs of the company and the key audience, research and fact check the content, and rewrite it into something that humans can use.

Go ahead, AI. Make my day!

I have been happy with what Google AI gives me.

I am sorry that you are not happy with your form of search for AI.

Maybe google might work like it does for me.

Would love to see the pictures of the beaver army.  Imagine if we added a prompt for gathering firewood.  We might be on to a good idea there.

Maieshe Ljin wrote:"Write a story that is repetitive, boring, and uncreative."

I just read the story.

I think it did an excellent job! That's about the most repetitive, boring and uncreative story I ever read. And yet it has undertones of  "You live like this, don't you? Maybe you should change, but you're not going to until you've switched your mind on and learned to think for yourself and realise how repetitive, boring and uncreative you have allowed your life to become"

Quite a masterpiece in its way...

The danger is that you don't know what you don't know. I think I would always fact check anything important before trusting a result. I do use it for auto translations, grammar checks, and amusingly AI text checking!
I generally prefer to ask a real person - hello permies! Or find a book.

To me it wrote a beautiful poem about Tai-chi, though I guided it towards it:

Whispers of Renewal

In gentle flow, the body sways,

A dance of peace in morning's rays.

With every breath, the spirit sings,

In harmony, the heart takes wing.

Soft whispers of the ancient art,

Where mind and body play their part.

With each slow step, the world unwinds,

A tapestry of calm it finds.

Flexibility in every move,

Strength emerges, as we improve.

The weight of stress begins to fade,

In tranquil moments, fears are laid.

Cognitive clarity, a brightening light,

In mindful stillness, thoughts take flight.

Connections bloom in shared embrace,

Together we find our sacred space.

As time unfolds, the cells renew,

With every practice, life feels true.

In unity, we rise and stand,

Rejuvenated, hand in hand.

So let us dance through life’s sweet song,

In Tai Chi’s grace, we all belong.

With open hearts and spirits free,

We weave our path to harmony.

Douglas Alpenstock wrote:I think the gibberish factor will make my skills increasingly more relevant in the next few years. I will be paid very well to rework the gibber to meet the needs of the company and the key audience, research and fact check the content, and rewrite it into something that humans can use.

This is pretty much what my field (translation) has evolved to right now. So far companies are so excited about "savings from AI" that they are very invested in using it, only a few are starting to come back to me saying they need a human to verify and smooth over, often in cases where there are legal implications. I'm doing a lot of editing for scientific publication, because while AI may talk a good game it often skips the parts it doesn't understand, which does not get you published in Nature.

Nancy Reading wrote:The danger is that you don't know what you don't know. I think I would always fact check anything important before trusting a result. I do use it for auto translations, grammar checks, and amusingly AI text checking!
I generally prefer to ask a real person - hello permies! Or find a book.

We can get gobs and gobs of these computer-generated images and texts now--and all this has really made me appreciative of humans. Humans can make mistakes transmitting information, sure--but humans give me more than merely information when I speak with them. They give me their unique perspectives and their experiences, their sense of humor, their unique appreciation for the topic under discussion, many of their related ideas, and their personal recommendations for me as me.  

The personal is utterly lost with computers. The personal is a thing we may now begin to value most highly--as I have begun to.