I'm in Love with ChatGPT

[Deleted]

I’m working on a sci-fi story, and last week while on holiday I finally organized the chapters and wrote a brief summary for each one.

After my recent success with ChatGPT in uncovering the term “fractoluminescence,” I was eager to test its abilities again. This time, I fed it my chapter summaries and asked for suggestions on character arcs and purpose.

While many of its ideas overlapped with my own, ChatGPT still offered some fresh insights and useful perspectives. And here’s the exciting part: just like with the fractoluminescence exchange, I now find myself in a position to ask deeper questions, explore new angles, and gather more observations.

It feels almost like having a trusted colleague, mentor, or teacher to bounce ideas off. That’s a huge shift—because let’s be honest, asking someone to read your draft purely to give feedback rarely works well. Unless they’re genuinely interested (or being paid), it’s hard for either side to get much out of it.

Now, though, I’ve got something close to a personal editor—an AI sounding board that’s always ready to help refine my story and offer the best advice it can.

Things are definitely looking up on the writing front!

 

[jdraw]

Tony,
The idea of an AI personal mentor, advisor was discussed at length in the Jordan B Peterson interviewing Brian Roemmele video you linked to here. Quite an interesting subject well worth watching.

 

[The_Doc_Man]

The biggest issue I found when exploring the several AI art sites I found was that if I DID use them to illustrate my fantasy fiction novels, they could not be copyrighted.

 

[jdraw]

Steal or re-use or 99.9999% reasonable facsimile?====> share

 

[Lightwave]

Uncle G - I'm using it in a similar context... I see it both as a good editor/ great mentor and like the best cheapest intern ever.

I'm finding it very useful for just getting the right TSQL. I still prefer to use it to polish my stuff rather than just trusting everything that it comes out with. But so far its definitely increasing my consistency. I'm mainly using Bing as I get it for free. I am considering paying for ChatGPT.

Regards writing I do wonder whether anyone will ever have the time to read all the stuff that can be produced now. If everyone is producing top quality stuff there might not be enough people to read it all. I am subscribed to someone called Ethan Mollick who I found on twitter he is using AI (all forms) as an assistant to learn from and co-write articles. Some of his articles are so long and deep though as a result that I rarely get time to read through them..

Ethan Mollick | Substack

I am a professor at the Wharton School of the University of Pennsylvania. I study entrepreneurship & innovation and AI. I am trying to understand what our new AI-haunted era means for work and education.

substack.com

I think as a mentor and tutor it is great as well - it very much reminds me of the personal illustrated primer in Diamond age by Neal Stephenson - the main benefit is to the person who collaboartes with the primer - I loved that book.

 

[Isaac]

Careful that headline is just what the killer says before she draws a knife! Beautiful, though!

 

[Cotswold]

The biggest issue I found when exploring the several AI art sites I found was that if I DID use them to illustrate my fantasy fiction novels, they could not be copyrighted.

That being the case. If someone generated a program in Access, C#, or any other language. Would it not follow that the software created using AI could not be copyrighted?

 

[jdraw]

Tony,

Just saw this on youtube- I watched about 20 minutes. An interesting perspective.
I did watch the Cory Schafer video ---very good!

 

[jdraw]

Perhaps AI really could be alien intelligence. Once it influences/determines cultural norms...

Somewhere between Jordan P and Harari is the likely position.

 

[jdraw]

My comment wasn't so much geared to programming or any specific domain. It is more of where AI and its usage will be positioned in the next few years.
In one sense it is the ability to have a private mentor, and all of your lifelong memory and learning with 100% recall tempered by your point of view. In another sense it is inexorable, repeated miscommunication (purposely falsified info) of any and every potential subject( advertising, finance, economics, politics, religion, health....).

 

[jdraw]

Tony,
Have you been watching videos by "eye on AI"?

 

[Isaac]

That being the case. If someone generated a program in Access, C#, or any other language. Would it not follow that the software created using AI could not be copyrighted?

Or worse yet, would OpenAI have a claim?

 

[Isaac]

My comment wasn't so much geared to programming or any specific domain. It is more of where AI and its usage will be positioned in the next few years.
In one sense it is the ability to have a private mentor, and all of your lifelong memory and learning with 100% recall tempered by your point of view. In another sense it is inexorable, repeated miscommunication (purposely falsified info) of any and every potential subject( advertising, finance, economics, politics, religion, health....).

That's a great point. I feel that AI goes back to that saying our mother's probably taught us: "Garbage in, garbage out".
(or, my brother used to tell me as a child, 'If you ask a stupid question, you get a stupid answer', but Mom made him stop saying that!)

AI depends on available input. The process of gathering input itself may, theoretically, be able to be fine-tuned to near-perfection.
It's the available input I worry about.

Even take coding for instance. For years I used to see content on Mr. Excel (and that type of website, which I consider crap-sites), that used Active, Active Sheet, Range.Activate, stuff like that - which I consider to be garbage code. And yet, because it was on a Site with a veneer of credibility, it spread like wildfire. Before a few years had passed, the internet was full of pages using the same type of code. Microsoft documentation used it to a small extent. So AI would size up that situation and believe it to be perfectly acceptable, perhaps, depending on how the question was asked and whether there was a way to 'limit' or 'steer' its available inputs.

And the limiting, and the steering of inputs.....therein lies the rub. It will be used for good, it will be used for evil, it will be used indifferently without regard.

AI will likely become like most other tools: Used somewhat for good, and very heavily for Power and Money. The consumer's knowledge of when and where AI is screwing them over will come much slower than the pain.

 

[Isaac]

Think about how Google, et al control information. Then consider that the primary source of information starts with Google. So, not only do the authors of the code have a Socialist leaning, all conservative views are hidden as deeply as possible. It's not like Google deletes the stuff you are not allowed to see, it is that Google buries it so far down the list that you never actually find it.

It's not totally the fault of the search engines although they tag all sites so that they know their political leaning (I'm sure this forum is seriously affecting this forum's social ranking negatively), it is also the use of confirmation bias. The more you click on stories or "like" anything, the more you are fed "similar" stories or stories with a similar bent. Click on one negative Trump headline and you'll never see a positive one again. Because of the coronation, I'm being inundated with British Royals stories because I clicked on a few to see photos and videos of the event.

I can't even have a "news" feed because I can't find one that will give me actual news of the world. I can get political "news" or Royals "news" but there is no such thing as balanced news like you used to get on TV at 6 and 11 or in the NYT or Hartford Current or WSJ. The WSJ was my last news paper but i can't even read it any more. Every article is now opinion. I don't care what the writer's Opinion is. I want the NEWS. Who, what, where, when, sometimes why but that usually comes much later. I don't want to read a "news" article filled with purple words and innuendo leaving me no doubt as to how the writer "feels" about the subject.

Just yesterday I was marvelling at the difference in coverage of that story about a large mob of teenagers who beat up some marines.
So fox news covered it pretty accurately IMO, it seems that yes there was a beginning altercation, and a beginning asking of the teens to leave by the marines, and then there was a one-sided beating with the marines curled on the ground and 40 teens stomping them. Pretty cut and dried..

Then I go to nbc news just for sh*ts and giggles. The way they cover it is TOTALLY DIFFERENT - I mean, Totally. They basically say nothing more than, there was a Fight between some marines and some teenagers. (as if it were even - and definitely hoping you believe that these marines were some kind of toxically masculine macho bravados who were fighting a teenager). Really unbelievable.

But the proof is in the pudding. I notice the Sheriff is not confused about what really happened like MSM is. Although the marines involved requested nothing, including medical attention, law enforcement vowed to make sure they figured out who to arrest for the stomping, and already have.

But really interesting how deceptive and false the main media reporting is. You can see it everywhere, in almost every article. It's not usually blatantly false information, it's just partial pictures meant to support a certain side. Partial pictures painted for you - over and over and over.

The idea of out of control, undisciplined teenagers running wild through cities is something conservatives usually point out. Liberals for example won't mention that Adam Toledo was running from a crime and had a gun in his hand and was starting to turn around WITH the gun when he was shot - they'll just mention that a 12 year old was shot by a police officer and hope you fill in the rest.

It amazes me how dumb and gullible most of the population is to go along with the deceptive, half-painted pictures. Then again, I wasn't far off from that myself until I started examining and comparing news stories.
And from the impression AWF interactions have given me (which hopefully aren't broadly true), it seems that Europeans and people from other countries generally open wide and swallow that partially-painted MSM picture, thus concluding that Republicans are some kind of crazies. Really it is the opposite, it is crazy to continually describe only half of something and then consider yourself discerning.

 

[Isaac]

that's interesting. Yeah - msm has tried really hard to avoid coverage of rampant crime, and tried really hard to paint military people as bad, so when a bunch of rampant deliquents beat up some military people, they had to go easy on that one!

as for chatGPT, I just got approved for Google labs. Really amazing - but INTEGRATED into their products - literally 10 yr ahead of chat GPT in my opinion. As I said, it will be he who holds the keys to INTEGRATION between customer data/products + AI who will win. I was on the Labs waitlist forever it felt like. Now they have AI integrated into search results, sheets.

 

[Jon]

It occur to me load the document into chat GPT and ask chat GPT to ferret out any buried information! I have no idea if chat GPT can do this

You can do a prompt like: Summarise in bullet points the key ideas from the following...

 

[Jon]

@Uncle Gizmo, your post summarised. I changed the word "document" as it didn't like it.

Edit: Just noticed that the summary seems longer than the original!

 

[jdraw]

Tony,
good stuff- as usual. I see Bing is using ChatGPT and prompting me to try it out; ask me anything.Quite impressive.
A question regarding post #40. Did you use ChatGPT to assist you with the verbiage?
UPDATE:
Using Bing, I took the text from post #40 and asked Bing to give me a 10 line summary.
Result:
Sure, I can help you with that. Here is a possible 10 line summary of the text:

The text is a personal account of how the author used ChatGPT, an artificial intelligence tool, to assist them in various tasks such as programming, word finding, and story writing. The author describes how they initially dismissed ChatGPT as a trend, but later became impressed by its capabilities and benefits. The author shares some examples of how ChatGPT helped them solve problems, learn new things, and enhance their productivity and enjoyment. The author concludes by recommending ChatGPT as a valuable companion for anyone who wants to accelerate their learning curve in any discipline.

It then offered to give me a title for the summary.
Bing provided these options:

Sure, I can generate a title for the text. Here are some possible titles:

• How ChatGPT Changed My Perspective on Artificial Intelligence
• ChatGPT: A Powerful and Versatile Tool for Learning and Creativity
• My Experience with ChatGPT: From Skepticism to Appreciation
• The Benefits of Using ChatGPT for Various Tasks and Projects
• ChatGPT: An Artificial Intelligence Friend that Helps You Grow

 

[Isaac]

I credit ChatGPT for bringing AI in general to the public's attention.

I am starting to see signs that it might be just a bit like:

ChatGPT is to AI what iPhone was to Smart Phones at first, or what AOL was to the Internet at first.

At first, chatGPT seems like "everything". A slowly-increasing % of consumers realize that chatGPT is just one brand's implementation and marketing of AI, and they branch out. Some (like many iPhone users) kind of stay stuck in their conceptual bubble that iPhone is a thing in and of itself and there is nothing else like it - those consumers give rise to the easiest and biggest profits for the company. (These are the ones who say, I got an iPhone so I can facetime my daughter - Rather than, I got a smartphone so I can video chat [any one of 75 apps] with my daughter). Subtle but enormously important difference for the company and brand.

I was very excited to get approved for Google Labs the other day but I think I know why, Google is trying to make sure it keeps up to where it is perceived to ought to keep up.

I now have integrated AI in my Searches. For example, I can still use Google Search, but I get AI-interactive portions of the screen that seamlessly integrate with my question and continue conversating - much like ChatGPT, only integrated into Google products.

Google has a ways to go, but I think if they keep this trend up they will stay on top.

They are doing the same thing with Code. As you perform your well-formed Search question, it will provide--and continue to refine--code snippets that match the ongoing conversation. Again, like chatGPT, except imagine chatGPT knows everything about the context of who you are, what you do, and what you googled 5 minutes ago.

 

[jdraw]

Yes, I did Jack. I usually just blurt out what I want to say, with little regard to structure, paragraphs, sentences and the like.

However, there is hope! I see the way chat GPT writes and I like how it looks. I aspire to writing like that myself! Not too much though, I think the personality needs to remain. I didn't put any personality back into what chat GPT changed.

in my original dialogue I said something like instead of walking up the flat section of the learning curve, I am now running up it. Chat GPT took that piece out and I should have put it back in.

On Twitter someone explained how a paragraph should be constructed. I copied the instructions and placed them in Chat GPT. I asked chat GPT if this was a correct statement and it said yes. I asked it to give me an example paragraph based on a science fiction story.

Chat GPT wrote an excellent piece on how a spaceship was heading through an asteroid field to find a Portal belonging to an alien transport system.

It was a good piece of writing but also it lacked personality.

Thinking about it. It's probably because it's too perfect. It's plastic. It's not genuine. It's not real in some way.

I might have found a limitation of chat GPT with its writing. It's too good, if you like!

I thought so. I've read many of your comments/posts over time and I did notice/perceived a change in structure. Sort of like "a personal mentor/teacher looking over your work offering some constructive criticism".

Keep up your interest and keep the forum informed.

As a side note: A while back I tried to get some chatgpt assistance with some vbe code. I learned a little of "AI Hallucinations". Chatgpt repeatedly gave me objects that don't exist. I have since (via youtube) found that this can be an issue with the language models and training --some times the results include fabrications or "alternate facts" issued as if they were absolute truths.