Exploring Diverse Perspectives: ChatGPT and Personal Prompts

[phillyexpressgoods]

In this lively eBay discussion, we delve into the fascinating world of ChatGPT and its interaction with personal prompts. Everyone brings their unique viewpoint to the table—some love it, while others remain skeptical. Do you craft your own prompts for ChatGPT? What intriguing questions do you pose to this AI companion? Share your experiences and insights! By the way, I’m using ChatGPT 4.0—curious minds want to know! 😊

And speaking of selling, what items do you have up for grabs? Let’s chat about that too! Plus, what quirky queries have you thrown at the robot? Join the conversation and let’s explore together! 🤖🛒

[itsjustasprain]

---

@chapeau-noir wrote:

Soon all Gen AI will have to feed on will be itself. Someone called that the Habsburg Effect, where successive inbreeding begins to cause more and worse deformity (here be monsters).

---

In Ebay terms I would say that I see it going the other way: If more sellers rely on AI to write descriptions then all AI will find for references in the future are its own descriptions from the past.

There is just no way you are going to get an original or interesting description for your item if AI was responsible for generating more and more of the past listings to which it will refer when generating the latest.

[chapeau-noir]

Is this lively yet?

[broto_64]

I'm not disagreeing with what you said, but in all of my life I have never ever been impressed with "smart" or any other kind of marketing. Every once in a blue moon an ad does come along that I fall head over heels for, but I am sure I can count the instances that actually happened on one hand.

AI to me, this "targeted" nonsense, even the best ads are still off-target. Like they're trying to sell me something I researched, not because I intend to buy it but because I needed to know something.

Better yet when bought something and these so-called "targeted" ads come at me days after I've already made my purchase.

That to me is just dumb, might as well spam me with random ads, same end result.

I too at one point was concerned about this "smart" advertising, but the only thing it did was go from one version of terrible to another and I expect nothing more to come from even the most advanced AI targeted super intelligent (but nevertheless ELIZA level of intellect) advertising.

Spend any time with any AI driven app and it's no more exciting than playing a game on a computer against computer-driven adversaries or partnering with NPC's. Yes they've come a long ways but ultimately it still boils down to the same program-based decision making processes. AI will always be limited by its database, CPU power, and human programmed interactions it is allowed to perform.

AI is NOT ever truly "self learning," is NOT "self aware" and as things stand never will be. Google does spread the FUD with its supposed "Quantum" computers but so far that has been dreams and ideas that I have not seen materialize, at all.

The scariest part of AI to me is its absolute ineptitude and the fact all the high echelon CEO's and the media flaunt it like it's the next greatest thing... Although I do wonder if CEO's really do spin it or whether it's just the media.  Doesn't matter, to me AI will always be predictable, hackable, and it will glitch and on occasion freeze and even crash completely requiring a reboot.

So, I am not impressed with AI.

It's good we have self-checkout lanes, it's good that cars have proximity sensors...

It's not always so good when the software makes decisions based on the input of sensors and such, but not because it's going to take over the world and cause World War Z, more like because it errs in its conclusions and then we spend more time correcting the error than if we had just done it ourselves all along.

@onlinecentral wrote:

What items do you have up for grabs? Used automotive parts.

What quirky queries have you thrown at the robot?
I include two sections within my descriptions entitled Removal Instructions and Installation Instructions.
I ran a couple of AI tests by asking how to remove or install a specific part.
The reply plagiarized my eBay description with missing and duplicated steps. 

---

Exactly, case in point.

[itsjustasprain]

---

@broto_64 wrote:

The scariest part of AI to me, is its absolute ineptitude and the fact all the high echelon CEO's and the media flaunt it like it's the next greatest thing...

---

I would put the ineptitude more in the hands of the user than the software. Here for example the OP states that he used ChatGPT 4.0 to generate the opening post here (which he presumably did do as his other posts or replies in the Community groups show the usual level of grammar and punctuation errors seen in human-written posts).

So the result was a technically correctly-written post with a couple of paragraphs that show little to no relevance to what the charter of this group is about. It talks about a lively Ebay discussion as if one had already occurred or as if it was launching the discussion that it would then continue. (The OP has not logged in since posting this last Friday morning.) Its questions about crafting prompts for ChatGPT do not seem to relate to how AI is being offered here: as a button to click on for obtaining an AI-written description. 

It seems to be more of an inadvertent demonstration of how to use AI badly: generate some text and then post it without actually reading it to see if it is a workable or even relevant solution for the purpose (or Community group) for which it was intended. The AI interface may do as it was asked. The user still has to know what he is asking for and how to evaluate his results.

[phillyexpressgoods]

My three e-commerce stores practically run themselves, aside from the initial product listings. It’s like having an automated system in place. I assume you’re more inclined to work with a large company like Pitney Bowes or a third-party provider that doesn’t involve coding. Additionally, the private shipping software I use seamlessly integrates with all carriers and automatically updates pricing. I’ve never had to directly contact any carrier for pricing information. Perhaps you don’t have the same perspective as we do.” 🚚💻📦

[phillyexpressgoods]

I wanted to know if anyone else uses chat and their opinions but I see it is what it is and sad.   The shipping software I use I never use pitney bowes.  Funny,  the developer worked for pitney bowes contractor.    Pitney bowes has history.   Sometimes its nice to think outside the box.  if it were the case that my sales were going through the roof,  I wouldn't be asking the question. 

[phillyexpressgoods]

Just to get an opinion.  I guess some people have nothing better to do.  

[toomuchstuffagain35]

Knew the "private shipping software" would come up eventually.

[broto_64]

---

@itsjustasprain wrote:

---

@broto_64 wrote:

The scariest part of AI to me, is its absolute ineptitude and the fact all the high echelon CEO's and the media flaunt it like it's the next greatest thing...

---

I would put the ineptitude more in the hands of the user than the software. Here for example the OP states that he used ChatGPT 4.0 to generate the opening post here (which he presumably did do as his other posts or replies in the Community groups show the usual level of grammar and punctuation errors seen in human-written posts).

So the result was a technically correctly-written post with a couple of paragraphs that show little to no relevance to what the charter of this group is about. It talks about a lively Ebay discussion as if one had already occurred or as if it was launching the discussion that it would then continue. (The OP has not logged in since posting this last Friday morning.) Its questions about crafting prompts for ChatGPT do not seem to relate to how AI is being offered here: as a button to click on for obtaining an AI-written description. 

It seems to be more of an inadvertent demonstration of how to use AI badly: generate some text and then post it without actually reading it to see if it is a workable or even relevant solution for the purpose (or Community group) for which it was intended. The AI interface may do as it was asked. The user still has to know what he is asking for and how to evaluate his results.

---

Well yes, but by the time I've done that I could've also done it without the extra "assistance."
I will admit, for the most part foreign language translation has gotten much better (but like you said it helps if you actually at least somewhat know the other language, and if you don't the results can be unexpected). And that is what I'm talking about, this randomness and that is what scares me.

It's in my wife's car, in the form of "adaptive" cruise control.

When it works like it should it's a dream, but when it doesn't it's dangerous.

And that is what scares me, AI's inability to distinguish real from irrelevant.

Because again by the time I "evaluate" the results it likely would've been better if I had just used regular cruise control, this particular piece of AI reacts in real time so there's a good surprise factor involved when our car's AI suddenly hits the brakes hard because it doesn't realize the car that's moving slower is in the exit lane and not in ours... What would happen if someone was tailgating, in most any traffic the majority of drivers follow too closely, actual speed makes no difference.

Sometimes, I sorely miss the 80's.

[chapeau-noir]

The 80s was a good decade - it's when I came of age.