Chatgpt - Seriously good potential (or just some Internet fun)

I've been working on quite a long powershell script over the past few weeks in order to fully automate one of our systems. There was one nifty feature I wanted to add, not important, but it would have been the icing on the cake for me.

Anyway, I was struggling with it, and even after looking on various websites I just couldn't work out the code. I've never used chatgpt before, but thought I'd give it a go. I asked it the question, and in less than 10 seconds it wrote it out for me. The code was efficient, and it worked!!!

I feel like a bit of a cheat if I'm honest.
 
I've been working on quite a long powershell script over the past few weeks in order to fully automate one of our systems. There was one nifty feature I wanted to add, not important, but it would have been the icing on the cake for me.

Anyway, I was struggling with it, and even after looking on various websites I just couldn't work out the code. I've never used chatgpt before, but thought I'd give it a go. I asked it the question, and in less than 10 seconds it wrote it out for me. The code was efficient, and it worked!!!

I feel like a bit of a cheat if I'm honest.

Not really cheating. Cutting and pasting from stackoverflow was so common for professional developers that it was meme'd to hell. Now its cut and paste from <favourite AI tool>. There is still a silly myth that we're supposed to know it all, which is utter balls. Knowing how to find the answer and use it quickly is what differentiates the ability.

I started writing code pre-internet. If you couldn't figure it out yourself or find it in a textbook / magazine article (anyone still reading Proceedings of the ACM ???) you were basically stuffed.
 
Not really cheating. Cutting and pasting from stackoverflow was so common for professional developers that it was meme'd to hell. Now its cut and paste from <favourite AI tool>. There is still a silly myth that we're supposed to know it all, which is utter balls. Knowing how to find the answer and use it quickly is what differentiates the ability.

I started writing code pre-internet. If you couldn't figure it out yourself or find it in a textbook / magazine article (anyone still reading Proceedings of the ACM ???) you were basically stuffed.
Agreed, we live in the information age - nobody knows everything and nobody needs to know everything - you just need to know how to effectively find and identify what you need.
 
I know I'm late to the party here but i've been having some great fun coding an npc in unity that is powered by chatgpt 3.5 turbo. It soon forgets everything i tell it so I've been trying to code a memory for it and it now works (most of the time :p ) but i hit the token limit quickly. I've just been trying gpt 4 1106 preview in the playground, it has a much better memory than 3.5 turbo. Anyway, i've just signed up for the chatgpt plus waiting list for a subscription to it, hope i hear something soon.
 
Games up

  • NEWS
  • 06 November 2023

‘ChatGPT detector’ catches AI-generated papers with unprecedented accuracy​

Tool based on machine learning uses features of writing style to distinguish between human and AI authors.

A machine-learning tool can easily spot when chemistry papers are written using the chatbot ChatGPT, according to a study published on 6 November in Cell Reports Physical Science1. The specialized classifier, which outperformed two existing artificial intelligence (AI) detectors, could help academic publishers to identify papers created by AI text generators.
“Most of the field of text analysis wants a really general detector that will work on anything,” says co-author Heather Desaire, a chemist at the University of Kansas in Lawrence. But by making a tool that focuses on a particular type of paper, “we were really going after accuracy”.

When tested on introductions written by people and those generated by AI from the same journals, the tool identified ChatGPT-3.5-written sections based on titles with 100% accuracy. For the ChatGPT-generated introductions based on abstracts, the accuracy was slightly lower, at 98%. The tool worked just as well with text written by ChatGPT-4, the latest version of the chatbot. By contrast, the AI detector ZeroGPT identified AI-written introductions with an accuracy of only about 35–65%, depending on the version of ChatGPT used and whether the introduction had been generated from the title or the abstract of the paper. A text-classifier tool produced by OpenAI, the maker of ChatGPT, also performed poorly — it was able to spot AI-written introductions with an accuracy of around 10–55%.
 
I've been working on quite a long powershell script over the past few weeks in order to fully automate one of our systems. There was one nifty feature I wanted to add, not important, but it would have been the icing on the cake for me.

Anyway, I was struggling with it, and even after looking on various websites I just couldn't work out the code. I've never used chatgpt before, but thought I'd give it a go. I asked it the question, and in less than 10 seconds it wrote it out for me. The code was efficient, and it worked!!!

I feel like a bit of a cheat if I'm honest.

It's not going to be anything novel really, you're just getting it to spew out boilerplate code that you might have otherwise (at least partially) got from stack overflow or some cookbook textbook.

The vast majority of programming is just doing that sort of thing, it's not like everyone is reinventing things from first principles. Maybe you do come up with some creative way to solve a problem but even then the basic steps you're taking to do it probably still involve some standard routines etc.. for which you can just ask ChatGPT.

It's not really able to do much reasoning (thus it's not great at mathematics or previously unseen logic problems etc..) so what you're getting out of it at this point is mostly the result of compression of its training set.
 
I've been working on quite a long powershell script over the past few weeks in order to fully automate one of our systems. There was one nifty feature I wanted to add, not important, but it would have been the icing on the cake for me.

Anyway, I was struggling with it, and even after looking on various websites I just couldn't work out the code. I've never used chatgpt before, but thought I'd give it a go. I asked it the question, and in less than 10 seconds it wrote it out for me. The code was efficient, and it worked!!!

I feel like a bit of a cheat if I'm honest.

Not cheating, just rolling with the times, just think, all that time wasted over the past few weeks if you asked it earlier :)
 
This is quite funny, people have been complaining about a drop in quality and one hypothesis is that it gets a bit lazy during the holiday season... after all, it's tuned via RLHF. :D (There may also be some issues with the base model too if this really is a thing, perhaps this human trait was learned there too.)

 
Last edited:
ME

Can you finish this rhyming joke.

Little birdy on the wall
Have you got no sense at all
Can't you see the wall's been plastered
Now you're stuck you silly _______


CHATGPT

Little birdy on the wall
Have you got no sense at all
Can't you see the wall's been plastered
Now you're stuck, you silly feathered disaster
 
Forgive my juvenile humour.
I asked Chatgpt 'can you write a story entitled ' on the verge of a poo'

Chatgpt said:
Once upon a time in the quirky town of Blunderburg, there lived a peculiar character named Oliver O'Leary. Oliver was an ordinary man with an extraordinary ability—he could predict the future, but only when he was on the verge of a particularly urgent bodily function.

One sunny morning, as the townsfolk went about their daily routines, Oliver found himself in the middle of the bustling Blunderburg market square. He could feel a strange sensation in his stomach, a sign that something peculiar was about to happen. As he clutched his stomach, a crowd began to gather around him, intrigued by the odd expression on his face.

Word quickly spread that Oliver was "on the verge of a poo," and the townspeople, always up for a bit of excitement, gathered in anticipation. Children stopped playing, vendors put down their wares, and even the mayor abandoned his important paperwork to witness Oliver's unusual talent.

With a theatrical gasp, Oliver grabbed onto a lamppost, his eyes widening as if he had just glimpsed into another dimension. The crowd held its collective breath, eager to hear what Oliver would reveal about the future of Blunderburg.

"In the near future," Oliver began, his voice trembling with urgency, "a great calamity will befall our beloved town. But fear not, for it shall pass if we all work together!"

The townspeople exchanged puzzled glances, wondering what kind of calamity could be averted by their collective efforts. Just as they were about to question Oliver further, a loud rumble echoed through the square. It wasn't the predicted calamity but something more immediate and personal.

Realization struck Oliver, and he blushed, realizing the awkward situation. "Uh, excuse me," he mumbled, looking for the nearest restroom. The crowd erupted into laughter, realizing that Oliver's unique talent wasn't as profound as they had hoped.

Despite the initial disappointment, the people of Blunderburg couldn't help but chuckle at the absurdity of the situation. From that day forward, the town embraced Oliver's quirkiness, turning the phrase "on the verge of a poo" into a local idiom for unexpected surprises.

And so, in the quirky town of Blunderburg, life continued with a touch of humor and a newfound appreciation for the unpredictable twists that fate could bring, especially when one found themselves on the verge of a poo.

It wouldn't win prizes but it was an alright effort by the AI.
 
Back
Top Bottom