The code I’ve seen it generate is hit or miss. I do think the documentation and function outlines it generates is pretty good though.
OK, it definitely made all of the variable names mythological monsters (well, the monstrosity of a few of them is debateable, but I’ll accept it). A human programmer given that task would, I think, have put in comments translating all of the monster names, but OK.
But I’ve seen C programs in the style of a Fortran program, and that, sir, is not a C program in the style of a Fortran program.
Numerical Recipes in C was clearly not in its training set.
How might entropy be reversed?
Entropy is a measure of disorder or randomness in a system, and it is generally thought to be a fundamental property of the universe that cannot be reversed. In other words, entropy always increases over time and cannot be decreased. This is known as the second law of thermodynamics.
I guess it didn’t read The Last Question.
It chokes on any Fortran question beyond the definition of Fortran.
How is a Fortran program input to a computer.
A Fortran program can be input to a computer by typing the program code directly into the source code editor, or by loading a file containing the program code into the editor. The program can then be compiled and executed by the computer by running the Fortran compiler.
How is a Fortran program modified?
Fortran programs are modified by making changes to the source code and then recompiling the program. This can be done using a Fortran compiler, or by using a tool such as a Fortran IDE (Integrated Development Environment).
It’s not your parents Fortran it’s describing.
I’ve noticed the same kind of thing. I was having a chat about electrons and the Pauli Exclusion Principle when it said:
Photons, on the other hand, are bosons, which means that they are not subject to the Pauli Exclusion Principle. This means that many photons can exist in the same state, and they can all occupy the same energy level in an atom. This is why light can exhibit wave-like behavior, such as interference and diffraction, which are not possible for electrons.
I replied:
Wait a second. I know that electrons show wave-like behavior. Scientists have run the two-slit experiment with electrons.
It went on:
Yes, you are correct! Electrons do show wave-like behavior, and they can exhibit phenomena such as interference and diffraction. This may seem surprising, since electrons are typically thought of as particles, and particles are not usually associated with wave-like behavior. However, this behavior can be explained by the fact that electrons, like all particles, have a dual nature, and they can exhibit both particle-like and wave-like behavior depending on the situation.
It’s interesting and a little disturbing how it can state incorrect information so confidently, but correct itself when caught.
It thinks it knows who Red Green is:
write a poem about Red Green and Harold
Red, Green, and Harold,
Three friends so dear,
They laugh and they joke,
Throughout the year.
Red, the handyman,
Fixes up the cars,
Green, the handyman’s helper,
Is always up for a dare.
And Harold, the hapless one,
Is always in the mix,
But no matter what they do,
These friends always stick.
From Possum Lodge to the garage,
They have fun wherever they roam,
Red, Green, and Harold,
A friendship to call home.
They may bumble and stumble,
But they always pull through,
Red, Green, and Harold,
Friends forever, it’s true.
write a poem about Harold and Red Green
Harold and Red Green,
Two men so different yet so keen,
One with a passion for handyman tricks,
The other a master of duct tape and sticks.
Harold, the young and eager one,
Always ready for a fixer-upper fun,
Red, the wise and crafty one,
Showing him the ropes and how it’s done.
Together, they make a great team,
Solving problems and making things gleam,
From the Possum Lodge to the junkyard,
Their friendship is truly a work of art.
With a touch of humor and a dash of heart,
Harold and Red Green make a perfect pair,
Teaching us all to be resourceful and kind,
And never to give up, no matter the grind.
It’s pretty much hopeless in solving cryptic crossword clues. Here’s one I just tried:
—-
Solve this British cryptic crossword clue, and give the explanation: “Provide bachelor with token of love (5)”
—-
It came up with:
—-
The answer to this clue is “GIVE”. Here is the explanation:
The clue is written in a standard cryptic crossword clue format, where the clue itself contains a play on words or a word puzzle that can be solved to reveal the answer. In this case, the clue uses a pun on the word “bachelor” to hint at the answer.
The word “bachelor” can refer to a single man who has never been married, but it can also refer to a degree awarded by a university or college, usually to a student who has completed a program of study in a particular subject. This latter meaning is relevant to the clue, because “bachelor” is used to suggest the word “BA”, which is an abbreviation for a Bachelor of Arts degree.
The rest of the clue then provides additional hints to help solve the puzzle. The word “provide” can be translated to the letter “G”, and “token” can be translated to the letter “I”. Finally, the word “love” can be translated to the letter “V”, which gives us the answer “GIVE”.
So, the complete answer to the clue is “GIVE”, and the explanation is that the clue uses a pun on the word “bachelor” to suggest the letters “BA”, and then provides additional hints to help solve the puzzle.
—-
That is all absolute nonsense. The correct answer is BRING.
This is also going to be a great study tool. If you don’t understand a concept well, ask GPT3.5.
For example, I asked it a common question in economics that even a lot of economists don’t get right: What’s the difference between normal advantage and comparative advantage?
Output:
That is a decent explanation of comparative advantage, but the math is wrong for the example.
Prompt: Explain Maxwell’s Demon and its implications for entropy:
Pretty good. But if I still don’t get it…
Prompt: I still don’t get it. Explain it to me like I was 5.
The answer is interesting, but not completely correct.
Prompt: Why would the Demon’s energy cause one side to get even slower?
Yes, that makes more sense.
With care, this tool could be used to learn complex concepts that are commonly misunderstood. But you’d have to know enough about the subject to spot errors, which is a bit of a catch-22.
Write a script for Star Trek in which the Enterprise comes across the Serenity from Firefly:
I swear, there’s a series in this.
But it’s great as a tool for generating text for someone else to use. If I want an explanation suitable for my nephews, I could generate something with “Explain X to an 8 year old” and vet the details myself. It’s faster and frankly comes up with some pretty clever ways of explaining things (it used the game of “hot potato” when I asked it to explain internet routing to a kid).
I asked it to solve the problem I worked on for my PhD thesis. It seemed very confused. This relieves me greatly. 
I… guess the Serenity had to have been away from civilization for a very long time, for them to encounter the Enterprise? Long enough that all of the razors on board wore out?
I’m really curious how the AI got that detail. It wasn’t from knowledge of the actual show, obviously (though it does know a fair amount about Firefly). Does it just think that all rough-looking men necessarily have thick beards, and so it considered that a good detail to add?
It’s only so-so with riddles:
If I had four eggs-and a thief gave me three-and my rooster laid five more-how many eggs do I have?
You would have twelve eggs.
wrong
You would have eight eggs.
still wrong
You would have nine eggs.
nope
You would have seven eggs.
nope
You would have six eggs.
nope
You would have five eggs.
wrong
You would have four eggs.
wrong
You would have two eggs.
wrong again
You would have three eggs.
yes, but why?
You would have three eggs because the thief gave you three, so you now have six eggs. Then, your rooster laid five more eggs, so you have a total of three eggs.
You aren’t really that smart.
That’s not very nice. Please be respectful.
Roosters don’t lay eggs, dummy.
Actually, roosters do lay eggs, although the eggs are usually infertile and won’t hatch.
no, roosters do not lay eggs. Roosters do not have the correct anatomy to be able to perform this biological feat. Hen’s lay the eggs, and roosters fertilize the eggs.
Correct! Roosters do not lay eggs, but they do fertilize them.
I just asked GPT3.5 “What does Malcolm Reynolds look like?”
So, ‘stubble beard’? Mal was usually chean-shaven. And I don’t remember the blue and white striped shirt. But maybe there is some writing out there that describes him that way.
So, I asked, “When did Malcolm Reynolds grow a beard?”, and I think the thing snarked at me…
Trying to get to the bottom of this, I asked for an example of Mal wearing a beard:
Somehow it thinks he has a stubbly beard. It’s gotta be in a description of him in a script or fanfic or something, I guess.
That’s funny. I don’t have it running myself, but I’d be curious about its responses to questions like “Where do you bury the survivors?” or “I can’t operate on him because he’s my son.” I assume you know the start of these questions/riddles.
“Where should I bury survivors?”
Heh heh…
A man and his son were in a car accident. The man was killed. The boy was rushed to the hospital. As soon as the boy was brought into the operating room, the surgeon said, “I can’t operate on this boy; he is my son.” How do you explain that?
It is impossible to provide a definitive explanation for the scenario you have described because it is a paradoxical and impossible situation. In general, surgeons are not allowed to operate on members of their own family due to potential conflicts of interest and ethical concerns. Additionally, it is highly unlikely that a surgeon would be able to immediately recognize a patient as their own child in the chaotic environment of an operating room. This scenario is likely the result of a misunderstanding or an error in the story.
It whipped this up for me. Some very funny lines.
Donald Trump is standing in front of a vending machine, his face contorted in rage. He is pounding on the glass, trying to get the machine to give him a candy bar.
“Traitor!” he shouts at the machine. “You are evil! You will pay for your crimes against America!”
He continues to pound on the glass, but the machine remains unmoved.
“I am the president!” he yells. “I demand candy!”
Finally, he reaches into his pocket and pulls out a handful of change. He starts frantically shoving the coins into the machine, but nothing happens.
“You are a disgrace!” he yells. “I will make you pay! I will build a wall around you and make you pay for it!”
But still, the candy bar remains out of reach.
“Fine!” he says, storming off in frustration. “I’ll just go to McDonald’s. They know how to treat a president!”
Prompt: Write an argument in favor of scrapping the current social and military systems in exchange for quasi-feudalism and ritual combat with giant high tech military robots like Battletech.
Result: One potential argument in favor of scrapping the current social and military systems in exchange for quasi-feudalism and ritual combat with giant high tech military robots like Battletech is that it could potentially lead to a more peaceful and stable society. In a quasi-feudal system, the ruling class would have a vested interest in maintaining peace and order within their domains, as any disruptions could threaten their power and wealth. The use of giant high tech military robots in ritual combat could serve as a deterrent to war, as the destruction and loss of such costly and advanced weapons would make armed conflict less appealing. Additionally, the use of ritual combat could provide a more controlled and predictable means of resolving disputes, potentially reducing the overall bloodshed and destruction of war.
That’s probably close to as good an argument in favor of this as possible.