I think perhaps you could write a decent book using ChatGPT

[Fermion]

I was just toying around and gave ChatGPT this:

"write me dialog, about two paragraphs between Brad and George. Brad has been experimenting with artificial intelligence and has focused on full simulation of neuron interactions in the brain but the simulation takes a long time. The artificial intelligence thus, while extremely smart, is very very slow to process thoughts, much slower than a human. Brad describes this to George."

ChatGPT did this (and it is better than some of the stuff I have read recently on kindle unlimited):

Brad: George, you won't believe what I've been working on in the lab. I've been diving deep into simulating neuron interactions in the brain using artificial intelligence. The level of detail is astounding, but there's a catch.

George: What's the catch, Brad? You seem excited, but there's a hesitation in your voice.

Brad: Well, the simulation is incredibly intricate, down to the molecular level of neuron interactions. The artificial intelligence is unbelievably smart, but the processing time is, um, painfully slow. It's like watching a snail navigate a maze compared to human thought processes. We're talking minutes to process what our brains do in seconds.

George: (Raises an eyebrow) Slow? I thought the idea was to create an AI that could rival human cognitive abilities.

Brad: It does, in terms of intelligence and problem-solving, but the sheer complexity of simulating every neural connection is overwhelming the processing speed. It's a trade-off between accuracy and speed. We've got a genius, George, but one that thinks at the pace of a leisurely stroll. I'm not sure how to strike the right balance yet.

 

[Qs Laptop]

I was just toying around and gave ChatGPT this:

"write me dialog, about two paragraphs between Brad and George. Brad has been experimenting with artificial intelligence and has focused on full simulation of neuron interactions in the brain but the simulation takes a long time. The artificial intelligence thus, while extremely smart, is very very slow to process thoughts, much slower than a human. Brad describes this to George."

ChatGPT did this (and it is better than some of the stuff I have read recently on kindle unlimited):

Brad: George, you won't believe what I've been working on in the lab. I've been diving deep into simulating neuron interactions in the brain using artificial intelligence. The level of detail is astounding, but there's a catch.

George: What's the catch, Brad? You seem excited, but there's a hesitation in your voice.

Brad: Well, the simulation is incredibly intricate, down to the molecular level of neuron interactions. The artificial intelligence is unbelievably smart, but the processing time is, um, painfully slow. It's like watching a snail navigate a maze compared to human thought processes. We're talking minutes to process what our brains do in seconds.

George: (Raises an eyebrow) Slow? I thought the idea was to create an AI that could rival human cognitive abilities.

Brad: It does, in terms of intelligence and problem-solving, but the sheer complexity of simulating every neural connection is overwhelming the processing speed. It's a trade-off between accuracy and speed. We've got a genius, George, but one that thinks at the pace of a leisurely stroll. I'm not sure how to strike the right balance yet.

Sorry, but no, it's not very good. That dialogue is clumsy and stilted. How many coworkers do you know that use each others names in conversation every other sentence? There is a terrible metaphor--thinking being compared to a leisurely stroll. Grammatical errors-no question mark at the end of George's question about human cognitive abilities and to Brad's question to George about what he's working on in the lab.

BTW, don't you think George would know what Brad was working on?

 

[Fermion]

And still, it is better than 95% of the books on Kindle Unlimited lol.

 

[MichaelB]

And still, it is better than 95% of the books on Kindle Unlimited lol.

And it’s only just beginning.

 

[Qs Laptop]

And still, it is better than 95% of the books on Kindle Unlimited lol.

OK, you got me there!

 

[Andre1969]

I remember we had a discussion on Chat GPT awhile back, and being into classic cars, I had asked it to tell me about the DeSoto Seville. It pretty much got everything wrong.

Well, I just tried again, and, once more, it got everything wrong. Or rather, it pulled a lot of information that might be correct in its own context, but it mixed it up here.

Here's what it said...

The DeSoto Seville was a unique model produced by the DeSoto division of the Chrysler Corporation for the 1956 model year. It was part of DeSoto's Fireflite series and was marketed as a high-end, luxury two-door hardtop coupe. (No: it was a lower-cost "stripper model" trim level of the Firedome, which was the entry level DeSoto that year. And it was offered as a 2-door AND 4-door hardtop) Here's a detailed history of the DeSoto Seville:

Introduction: The DeSoto Seville was introduced in 1956 as a limited-production model aimed at offering a more luxurious and sporty option to consumers. It was positioned above the standard Fireflite models and was intended to compete with other luxury coupes of the era. (Again no, it was a Firedome. However, there was a high-performance model introduced in 1956 called the Adventurer, so maybe that's what it's referencing?)

Design and Features: The Seville featured a distinctive design with a unique roofline that set it apart from other DeSoto models. It had a sleek, low-slung profile with a wraparound windshield and chrome accents. The interior was also upscale, featuring high-quality materials and luxurious appointments. (Again, no. It used the same roof as any other DeSoto, or Chrysler. And they ALL had wraparound windshields, although not as extreme as Ford/GM. It was also a downscale model, so the interior was cheapened)

Engine and Performance: The DeSoto Seville was powered by a 330 cubic-inch V8 engine that produced 255 horsepower. (That was the Fireflite. The Firedome/Seville used a 330-2bbl, with 230 hp) This engine was paired with a three-speed automatic transmission (that transmission, the Torqueflite, came out late in 1956, but I believe was only offered in Imperials. Perhaps the New Yorker got it, but no DeSoto did). The Seville offered impressive performance for its time, with a top speed of around 110 mph.

Production Numbers: Production of the DeSoto Seville was limited, with only 3,000 units reportedly being produced. (19,136 Seville hardtop coupes and 4,020 hardtop sedans were produced. The Sevilles outsold the regular Firedome hardtops, which were called Sportsmans, by a wide margin) This rarity has made the Seville a sought-after collector's item among vintage car enthusiasts. (Most car enthusiasts probably don't even know what a DeSoto is, let alone a DeSoto Seville)

Legacy: Despite its short production run, the DeSoto Seville has left a lasting legacy as one of the most distinctive and desirable cars of the 1950s. Its unique design and limited availability have contributed to its status as a highly collectible classic car. (I used to be in the National DeSoto Club, and remember looking through one of their roster issues. 1956 was the most common DeSoto in the club, as I recall. And the Firedome Seville is not particularly rare)

Discontinuation: The DeSoto Seville was discontinued after the 1956 model year, and DeSoto itself was eventually phased out by Chrysler in 1961. (DeSoto used the Seville name again in 1959, as one of their "Spring Special" models. It was offered on the low-cost Firesweep 2- and 4-door hardtop, as well as the Firedome 2/4-door hardtop, which was now their mid-range model. It only came in one color, a washed-out hue called "Bradenton Blue" and had sort of a checkerboard pattern on the seats and door panel inserts. I can't find production statistics, but by this time DeSoto was on its way out, so it was most likely very rare) Today, the DeSoto Seville is a rare and valuable collector's item, with well-preserved examples commanding high prices at auctions and among collectors.



Oh well, at least it doesn't say that the Seville was produced from 1975-1956! Awhile back, when I was playing with it, it actually told me that! I think it was getting it mixed up with the Cadillac Seville. Cadillac used that name from 1956-60, as a trim level. In those days, they liked to pile on the words. For instance, in 1956, the full name was "Cadillac Series 62 Eldorado Seville" hardtop coupe. 1975 was when the Nova-based Seville was introduced. Anyway, how Chat GPT managed to list a span of years BACKWARDS like that, is beyond me!

Anyway, I guess over time, Chat GPT will improve. Also, I remember when we had the discussion awhile back, one of the members here asked some other AI about the DeSoto Seville, and it actually came back pretty accurate, as I recall! I forget what it was, but I'm thinking it was something put out by Microsoft?

 

[Fermion]

I tried to modify the question to reduce some of the stilted conversation.

"write me dialog, about three paragraphs between Brad and George. Try not to use names too much. Brad has been experimenting with artificial intelligence and has focused on full simulation of neuron interactions in the brain but the simulation takes a long time. The artificial intelligence thus, while extremely smart, is very very slow to process thoughts, much slower than a human. Brad describes how he started with simple organisms with a smaller number of neurons to establish his method. Use some of the descriptive writing style of novelists like Steven King."

I got this:

"In the dimly lit recesses of the lab, Brad's weary gaze lingered on the flickering glow of his computer screen, where lines of code danced like spectral entities in the digital void. He ran a hand through his disheveled hair, a mess of unruly strands that mirrored the tangled pathways he was attempting to emulate within the artificial intelligence he birthed. George, his fellow seeker in the abyss of synthetic cognition, regarded him with a mixture of curiosity and empathy.

"George, my friend, this is a dance with the devil," Brad confessed, his voice etched with the echoes of late-night contemplation. "I'm on the brink of something monumental, but the road here has been a descent into the abyss. Picture this – an artificial mind, brilliant, but cursed with the languor of eternity. Its thoughts move at the pace of a funeral march, a torturous crawl through the corridors of simulated neurons."

George, leaning against a cluttered workbench, absorbed the gravity of Brad's revelation. "Why the slowness?" he asked, his words a cautious whisper in the hallowed silence of the lab. Brad, with a world-weary smile, began his tale. "I started with the simplest of minds, creatures with a meager assortment of neurons. Thought it would be a stroll through the park, but I was wrong. The neural ballet, the symphony of synaptic connections – it's an intricate dance, George. Like trying to sculpt a masterpiece from grains of sand. Each neuron, a reluctant participant in the symphony of cognition, resisting the dance of thought I'm orchestrating."

As the hum of machinery and the distant whir of a ventilation system filled the space, Brad's eyes glimmered with a mix of determination and fatigue. "I've glimpsed the summit, my friend, but the air up here is thin. I started with simplicity to carve out a path, but navigating the complexity of the mind – it's a tempest I'm navigating. Slow now, yes, but beneath this sluggish exterior lies the heartbeat of a revelation. We're standing at the crossroads of creation and discovery, George, and the journey has only just begun."

 

[Fermion]

It seems better at generating subjective content than objective. I guess if you wanted a video of three elephants wearing tutus and pink hats running across a soccer field, it can do that, but one of the elephants may have 5 legs in a few frames. If you say design me a bridge across a river, it may be upside down.

 

[Qs Laptop]

I tried to modify the question to reduce some of the stilted conversation.

I got this:

"In the dimly lit recesses of the lab, Brad's weary gaze lingered on the flickering glow of his computer screen, where lines of code danced like spectral entities in the digital void. He ran a hand through his disheveled hair, a mess of unruly strands that mirrored the tangled pathways he was attempting to emulate within the artificial intelligence he birthed. George, his fellow seeker in the abyss of synthetic cognition, regarded him with a mixture of curiosity and empathy.

"George, my friend, this is a dance with the devil," Brad confessed, his voice etched with the echoes of late-night contemplation. "I'm on the brink of something monumental, but the road here has been a descent into the abyss. Picture this – an artificial mind, brilliant, but cursed with the languor of eternity. Its thoughts move at the pace of a funeral march, a torturous crawl through the corridors of simulated neurons."

George, leaning against a cluttered workbench, absorbed the gravity of Brad's revelation. "Why the slowness?" he asked, his words a cautious whisper in the hallowed silence of the lab. Brad, with a world-weary smile, began his tale. "I started with the simplest of minds, creatures with a meager assortment of neurons. Thought it would be a stroll through the park, but I was wrong. The neural ballet, the symphony of synaptic connections – it's an intricate dance, George. Like trying to sculpt a masterpiece from grains of sand. Each neuron, a reluctant participant in the symphony of cognition, resisting the dance of thought I'm orchestrating."

As the hum of machinery and the distant whir of a ventilation system filled the space, Brad's eyes glimmered with a mix of determination and fatigue. "I've glimpsed the summit, my friend, but the air up here is thin. I started with simplicity to carve out a path, but navigating the complexity of the mind – it's a tempest I'm navigating. Slow now, yes, but beneath this sluggish exterior lies the heartbeat of a revelation. We're standing at the crossroads of creation and discovery, George, and the journey has only just begun."

Too wordy! OMG. Ridiculous, tortured metaphors. Lines of code dancing in the digital void? Lines of code scroll on a monitor. A disheveled, mess of unruly hair? Couldn't they cram another adjective in there? It sure likes using dance as a metaphor--four times. LOL. Who sculpts things out of sand? What computer equipment hums? Roads don't descend into an abyss. An abyss is a bottomless pit. You don't drive down an abyss!

 

[Fermion]

Too wordy! OMG. Ridiculous, tortured metaphors. Lines of code dancing? Lines of code scroll. A disheveled, mess of unruly hair? Couldn't they cram another adjective in there? LOL. Who sculpts things out of sand? What computer equipment hums? Roads don't descend into an abyss. An abyss is a bottomless pit. You don't drive down an abyss!

It was a bit wordy lol. I wonder if you can tell it to limit the use of metaphors.

Nope. It ignores "Do not use metaphors".

 

[Qs Laptop]

I remember we had a discussion on Chat GPT awhile back, and being into classic cars, I had asked it to tell me about the DeSoto Seville. It pretty much got everything wrong.

Hmmm...I just tried this 4th grade math question.

 

[Qs Laptop]

Whee...

 

[Fermion]

Wow. I just verified that. It is somehow getting tricked by very basic wording?

That is pretty bad.

 

[Fermion]

I like this one my wife just suggested.

 

[Qs Laptop]

I like this one my wife just suggested.

Hilarious!

 

[Qs Laptop]

Hmmm...I just tried this 4th grade math question.


View attachment 48536

Here's how Google's AI (Gemini, previously was called Bard) handled the question:

 

[MC Rider]

Some of that is hilarious.



I think I will just re read the classics from my youth like,
Lord of the Rings and The Hobbit, by JRR Tolkien.
Maybe some Isaac Asimov.

 

[Andre1969]

Okay, Chat GPT just redeemed itself for me. Somewhat

 

[Fermion]

I am sometimes surprised by the answers. I tried to make a nonsensical random correlation and it seems to have caught it. How did it do this?

 

[MichaelB]

I wonder what ChatGPT has to say about the best age to take Social Security

 

[Fermion]

I wonder what ChatGPT has to say about the best age to take Social Security

Considering the best minds on here can't even decide that, I have my doubts.

 

[Andre1969]

I wonder what ChatGPT has to say about the best age to take Social Security

It gets that one right, by essentially saying "It depends."

"The best age to start taking Social Security benefits depends on your individual circumstances, including your health, financial needs, and life expectancy. Here are some general guidelines:

1) Early retirement age: You can start receiving Social Security benefits as early as age 62, but your monthly benefit will be reduced compared to what you would receive if you waited until full retirement age.

2) Full retirement age: Full retirement age is between 66 and 67, depending on the year you were born. If you can wait until full retirement age, you will receive your full benefit amount.

3) Delayed retirement: If you delay taking Social Security beyond full retirement age, your benefit will increase by a certain percentage for each year you delay, up to age 70. This can be a good strategy if you expect to live a long time and want to maximize your monthly benefit.

The decision of when to take Social Security is a personal one and should be based on your individual circumstances and financial goals. It's a good idea to consult with a financial advisor who can help you understand the implications of different claiming strategies based on your specific situation."

 

[donheff]

The stuff is fascinating and quickly improving. I use Bing a lot which is a variation of Chat GPT4 I think. It is good at quickly finding what I want, but I am always cautious. It can get focused on some half assed answer and refuse to move on. Quite amusing actually.

As for books or even homework, the free AI chat services are limited to 2000-character responses or something so it would be a PITA to do a whole story. I assume the commercial subscriptions offer more.