I'm in Love with ChatGPT

I asked ChatGPT what is wrong with this code:

Code:
Sub ok()
Dim var As Integer
For var = 1 To 1000000000
    MsgBox "ok"
Next var
End Sub

It brought up 3 "problems", no fewer than TWO of which were flat-out WRONG (claimed that msgboxes will accumulate on the screen, completely missed the concept of a modal popup, secondly it claimed it will cause an infinite loop which is obviously not true), and NONE of which caught the glaring issue that's definitely wrong.

This seems very basic, if I were to give someone a VBA test who claimed intermediate familiarity I would expect them to catch this within seconds.

Doesn't bode well for getting code advice from ChatGPT in my opinion, I was just wanting to test it.
I would have a hard time trusting it after this.

Like I said - there has to be a way to objectively test it, rather than just feeding it questions that you yourself don't know the answer to and therefore, can't evaluate the response quality.
 
Doesn't bode well for getting code advice from ChatGPT

I found ChatGPT pretty good with Python. It was also pretty good in helping me build a Chrome extension. However I had to put quite a bit of effort in to counteracting its wokishness and its.... What...? I was thinking "arrogance" but that's not quite the right word. Chatty comes over as arrogant in that it appears to know what it's talking about, but once you start believing that, you are in trouble!

My advice, you MUST use ChatGPT to help you in a programming language you thoroughly understand before you start using it in a language you don't understand.

Why? I've started one of my "forever shelved" Flutter projects... It is an excruciatingly painful process learning Flutter with chapGPT. But only because I thought it knew what it was talking about! It does not know. I'm currently trying to get it to modularize the code by extracting a widget (a class) into a separate file and then replace the section of code with a reference to that file. Painful! It should be simple, but it's not ...

I've fallen back to watching YouTube videos to get a better idea of what to do, but there are a few and far between good ones!

I saw this comment posted on Reddit:-

Beware: LLMs generate plausible text. Not most likely text, or accurate text. Just believable text denoted with confidence. They are great as a creative assistant, but frequently bad as an authoritative source.

I thought that was a brilliant and spot on observation...
 
Last edited:
Yes, amongst the rest of her other plagiarised writings

Did someone use a chatGPT or some other large language model to expose her plagiarism?
 
Would you prefer to talk to your doctor or a large language model like chat GPT?
We are heading towards those nifty scanners that Bones used to use in Star Trek. Smartphones can already get attachments that do medical stuff.
 
There's a new plug-in for chat GPT called "Consensus":-

Consensus

Your AI Research Assistant. Search 200M academic papers from Consensus, get science-based answers, and draft content with accurate citations. (Might be handy for Harvard University faculty members! )

I tested it yesterday by asking about the puffer fish that builds a big circle in the sand to attract its mate. It provided me one new snippet of information which I hadn't heard before. They believe the design is somehow related to collecting fine Sands in the middle which are best for the development of the eggs.

I've just been listening to Lex Frideman speaking to a Glasgow University professor Lee Cronin, who has an interesting new Theory called "Assembly Theory" I am just about to ask the Chat GPT consensus plugin if it has any information on this...
 
Last edited:
Smartphones can already get attachments that do medical stuff.

There is a 79$ attachment for Android phones where the App connects to the device and takes your EKG, records it, and can send it to a doctor. It doesn't do the "advanced" EKG that looks at multiple leads, but it can detect common situations like bradycardia, tachycardia, atrial fibrillation, etc.
 
New add-on for chat GPT called "Creative Writing Coach"
I have been writing a sci-fi story now for years. However I've run up against "writers block". The writers block itself has probably been there for a couple of years at least! I didn't admit that to myself, I just didn't write anything. I would do all sorts of procrastination relating to the story, checking spelling and writing ancillary chapters which didn't have a bearing on the main story. However I noticed a new add-on for chat GPT called "Creative Writing Coach". Now chat GPT itself is pretty good at helping with your writing but this creative writing coach takes it to another level, it's brilliant! It's helping me build character story arcs and the whole story structure. It's helping me come to terms with the block which is a discussion between Malcolm who has an illegitimate son Rylan and this is their first meeting and I just could not get my head around the dialogue.

This is what the chat GPT "Creative Writing Coach" produced as a suggestion for the dialogue:-

Dialogue Between Malcolm and Rylan

Setting: Aboard the "Splendor of the Stars", Malcolm's luxurious yet enigmatic office. The atmosphere is tense yet charged with unspoken emotions.

Malcolm: [With a measured tone] Rylan, I appreciate your coming. I understand this must be... unconventional for you.

Rylan: [Skeptically] Unconventional is one way to put it. You pull me out of a base jump, and now here? What's this about?

Malcolm: [Pausing, choosing his words carefully] It's about your mother, Liz. And about something bigger, something... beyond us.

Rylan: [His interest piqued, but guarded] My mother? What about her? She's been gone for years.

Malcolm: [Gently] Yes, she has. But what if I told you that we might have a way to find out what really happened to her?

Rylan: [Suspicious] Why would you want to help me? What's in it for you?

Malcolm: [With a hint of emotion] It's not just about what I want. It's about... correcting a wrong. Your mother's disappearance, it's linked to the space elevator. To time anomalies we didn't foresee.

Rylan: [Incredulous] Time anomalies? Are you serious?

Malcolm: [Solemnly] Very. And I believe you might have inherited certain... abilities from your mother. Abilities that could help us.

Rylan: [Frowning] Abilities? What are you talking about?

Malcolm: [Hesitant] It's difficult to explain. But your mother had a unique genetic trait. One that we think is linked to these anomalies. You might have it too.

Rylan: [Reflecting, then with a sudden realization] Wait, how do you know so much about my mother? Who are you to her?

Malcolm: [Taking a deep breath, deciding against full disclosure] I... knew her. Quite well. She was... important to me.

Rylan: [Trying to piece things together] Knew her? What does that even mean?

Malcolm: [Avoiding the question] Right now, Rylan, what's important is that we find out what happened to her. I can't do it without you.

Rylan: [Suspicious but curious] And what if I say no?

Malcolm: [Sincerely] Then you walk away. But consider this, Rylan: this could be your chance to find out about your mother. About who she really was. About who you really are.

End of dialogue

I particularly like the bracketed emotional context.

This is what chat GPT "Creative Writing Coach" said about it:-
The bracketed emotional context is provided primarily as a guide to help you visualize and understand the underlying emotions and intentions of the characters in each line of dialogue. It's a tool for you as the writer to grasp the nuances of the scene more deeply.

Then chat GPT gave me an example of how to turn the above into writing:-
You might rewrite a portion of the dialogue for a narrative format:

Malcolm's office aboard the "Splendor of the Stars" was a blend of luxury and mystery, much like the man himself. He watched Rylan enter, his expression measured, betraying nothing of the tumultuous emotions beneath.

"Rylan, I appreciate you coming," Malcolm began, his voice even. "I understand this must be... unconventional for you."

"Unconventional is one way to put it," Rylan replied, skepticism lacing his words. His eyes, sharp and calculating, scanned the room as if searching for hidden truths. "You pull me out of a base jump, and now here? What's this about?"

"It's about your mother, Liz," Malcolm said gently, yet each word seemed to weigh heavily on him. "And about something bigger, something... beyond us."

Rylan's stance shifted, interest flickering in his eyes, quickly masked by a guarded expression. "My mother? What about her? She's been gone for years."

"Yes, she has. But what if I told you that we might have a way to find out what really happened to her?" Malcolm's voice held a hint of something unspoken, a depth of feeling he seemed to keep in check.

Rylan's suspicion was palpable. "Why would you want to help me? What's in it for you?"

Malcolm's gaze held Rylan's, a flicker of emotion passing through his eyes. "It's not just about what I want. It's about... correcting a wrong. Your mother's disappearance, it's linked to the space elevator. To time anomalies we didn't foresee."

"Time anomalies? Are you serious?" Rylan's tone was incredulous, yet a part of him appeared intrigued by the absurdity of it all.

Of course it remains quite likely that the "Creative Writing Coach" provides another form of procrastination!
 
Last edited:
Have you tried the new version of Baird?

 
Have you tried the new version of Baird?

I haven't! Me and Bard didn't get on that well.

But! >>>>>>

Reading this article:-


Which suggests using a combination of Bard and Chat GPT allows you to draw on both of the AI's expertise and come up with some excellent help with your story writing.

I can vouch for the capability of Chat GPT however I acknowledge that it doesn't have depth. I have yet to try bard, I will see if it gives me the required depth.

Interestingly the author of the afore mentioned blog post suggests you cite the ai's contributions...

I usually mention stuff that I've used AI to generate, (I have forgotten once or twice), and really, except for the necessity of the reader to know who is writing, is it necessary to mention the input from AI large language models? I don't see why. Accept for the benefit of the reader there's no benefit to the AI itself... .

I haven't used one yet but I understand there are websites which can help you produce a particular voice, like write for you in an Irish accent or a cockney accent. I've never seen any one cite and article, or even an online or book source for such information. Really, it would look a bit silly if you said oh "that last dialogue was got from the meridian encyclopedia of cockney accents!" It would be like saying I used this this or that thesaurus to look up this alternative word!

Don't know really! Needs hammering out to come to a genuinely thought out conclusion.
 
I mention it becasue a) its a new version, and b) I think its multi-modal, c) more up to date?
 
I think its multi-modal, c) more up to date?
Bard is substantially improved by all accounts...
From reading that article I'm definitely going to try a combination of chat GPT and Bard. One tip I picked up is to have ChatGPT do the dialogue from a different person's perspective. Instead of having the person you want to speak speak, get chatGPT to do the dialogue of the other person.
 
I believe Google's Gemini will be the best from them because it uses some techniques that DeepMind created in AlphaGo. I've lost track if it is already out or not.
 
This article discusses how to use chat GPT to respond to emails with emails!


So this means chat GPT is going to generate more emails, who wants that?

And the emails it generates are properly going to be intercepted by another chart GPT and responded to.

A total waste of time and space to be had by all I reckon!
 
I should mention that I took one look at the article and came to the conclusion it was a verbose peace of writing probably generated by chat GPT with very little substance.

So I asked chat GPT to summarise it for me, then I asked GPT if the article mentioned any user cases. No, was the answer, just some vague idea of how it could be used. in other words I don't think the writer has really considered and examined the process themselves.

They just had an idea of what could be done and thought I'll write an article about that!
 
Khan Academy has added a new plug-in for chat GPT to help you learn physics maths and things. I've just asked it a question and it came back with a helpful answer!

Tutor Me
By khanacademy.org
Your personal AI tutor by Khan Academy! I'm Khanmigo Lite - here to help you with math, science, and humanities questions. I won’t do your work for you, but I will help you learn how to solve them on your own. Can you tell me the problem or exercise you’d like to solve?
 

Users who are viewing this thread

Back
Top Bottom