I dunno how it's use can be stopped. Or why it should. I received approval for yet another shop with ChatGPT narrative help. I use it to draft a section and then I go back and revise it with my actual shop findings. It's factual. It's helpful.

Isn't this just a natural progression from a dictionary? An APA Guide? An MLA Guide? A spell-check app?

While ChatGPT and other AI systems can be helpful, their text does not always read natural. As you said, you have to revise it. What if you miss something? What if there is a contradiction within the narrative that you missed? I used ChatGPT over the summer on a few shops. I wasn't impressed. I found myself having to edit and rewrite so many things that I think it took more time than had I just written the narrative myself without any help. I also have used it to list items online to sell, even before eBay implemented AI into their listing platform. Again, didn't like it.

There are programs that can detect if something is written by AI. I'm sure the MSC's will employ them, if they haven't already. Still, as you noted, it may not be able to be stopped. Then what? If you can't beat them, join them? Does the MSC decide to turn all shops into simply hard data, no personal touch? Will the MSCs decide to do narratives on the backend? It would be a great way to lower shopper pay, yet again. Most stores have cameras and probably microphones, somewhere. What happens when AI reaches that point that one can feed video and audio to it, then have it analyze everything an employee did during a real customer interaction. Would IC work cease to exists? I don't like anything about Hollywood, but they can see how AI will replace many of them if the brakes are not pumped on this.

A machine did not experience the shop. A machine did not interact with the employee. A machine can't bring any context to the data. You can get a cliff notes version out of ChatGPT, sure. You won't get the human touch. I, for one, think we've lost enough of that already.

There is a constant risk of too much subjectivity and too little humanity. Somewhere, there is a happy medium between too much human subjectivity which could skew dry data and the complete absence of human input which includes nuances that tech cannot supply. AI and ChatGPT have become too much of a good thing.

IC work will continue to exist because businesses want to know the customer's EXPERIENCE. You can't fully capture a customer's experience with just raw audio and video files. That is where a factually crafted narrative comes in.

If you probe ChatGPT with the correct input, you can get a suitable output that needs less massaging than a total rewrite of a narrative. It takes practice.

Edited 1 time(s). Last edit at 01/13/2024 01:01AM by maverick1.

I would not be doing SS if Chat did not exist. The grammar they require you have to damn near need to be a English major.

You have to tell it what you're actually writing. For example. Revise my REVIEW,....revise my COMPLAINT....revise my LETTER. Chat is a life saver and saves me soooooooo much time.

its getting stupider the more and more it gets flooded with information. ironic in how thats a reflection of society as a whole

I would not be doing this stuff if i couldn't use chat. It has also improved my grammar

I just said this! So true! Lol

@Amiking wrote:

I would not be doing this stuff if i couldn't use chat. It has also improved my grammar

Nonsense. When you use AI to write *your* reports for you, those aren't *your* words, so it's not *your* grammar that's seeing an improvement.

If you want to do yourself a favor, purchase a Business English textbook and begin reading.

Yes, to what @drdoggie00 said.

As a follow-up, Gemini is better than ChatGPT IMHO. Although I am still against using it to write reports for mystery shopping, I have found it helpful for other things.

Yes, to what @drdoggie00 said.

As a follow-up, Gemini is better than ChatGPT IMHO. Although I am still against using it to write reports for mystery shopping, I have found it helpful for other things.

Yes, to what @drdoggie00 said.

As a follow-up, Gemini is better than ChatGPT IMHO. Although I am still against using it to write reports for mystery shopping, I have found it helpful for other things.

EduWriting I regret to share that my negative experience with the site left me dissatisfied. With this bad experience, I cannot endorse this or any other "AI essay tool" that allegedly writes essays for students. They are located overseas and steal your content even if you don't use their service, but it's on a trial! My expectations were not met, and the tool proved to be unreliable. It is clear that whoever wants to use it for academic research and essay writing will go somewhere else (this is what I did). When it comes to relying on tools for critical tasks like writing essays, my advice is to exercise extreme caution and actively seek out alternative options (for example working with an experienced freelancer or research service).

Real human writers are more likely to fulfill your coursework requirements than an automated bot generating gibberish.

This website or service is not trustworthy as it has been found to be fake. Therefore, students are not recommended to use it. Additionally, there have been reports of data theft or unauthorized access. Therefore, students should avoid sharing any sensitive information with this service to prevent any potential harm to one's privacy and security. In general, AI writing tools are like soft automated plagiarism generators that are going to be detected by your school.

There is a reason why flesh and blood academic writers are still the way to go. AI is still too much in its writing infancy to be a truly reliable tool for anybody to use. Human beings, on the other hand, have a wealth of experience when it comes to writing academic papers. We started writing for ourselves first, then moved on to writing for other students. So we have the kind of experience that an AI can never have. The human experience factor will always beat an AI every single time. So use a human being for your writing needs instead. Unless, you actually want to write the paper yourself, then you can use an AI to get you started.
The opinions are that of the author's alone based on an individual capacity. Opinions are provided "as is" and are not error-free.

Some of my clients tried out AI and immediately realized that essays generated by AI are worthless, unless you happen to know that your professor doesn't even bother reading submitted projects.
Highly experienced, versatile, honest writer with a US Law degree (JD) located in NYC.

I think it's useful in long shops where you can just voice record the narrative. The shorter ones with just clicking ratings it's next to useless.

What other things do you use AI for?

@ServiceAward wrote:

Yes, to what @drdoggie00 said.

As a follow-up, Gemini is better than ChatGPT IMHO. Although I am still against using it to write reports for mystery shopping, I have found it helpful for other things.

My experiences with AI continue to be laughable.

Or sad, in the case of a student who used it twice for a research project and generated a pair of abject fails. Not a single relevant fact was presented in either 4+ page, double-spaced document. All fluff with appropriate names and dates plugged in after the bot scrubbed the intertubes.

Or annoying and costly, in the case of the project I'm working on for the non-profit I do. It involves plugging the output of a sound-generating device into an iPad so an app can show stuff on its screen relative to the sound coming in. My developer uses ChapGPT to answer questions about things he is coding. Asked about the nature of the connection we need to make and received an inaccurate response which led him down a paid rabbit hole he didn't need to visit.

Last time I tried it (emphasis on 'last'), Google's Gemini couldn't properly answer a question about gas station hours.

If it can't get simple stuff like that right, I'm certainly not going to trust with it anything important.

@CoolMusic wrote:

My experiences with AI continue to be laughable.

Or sad, in the case of a student who used it twice for a research project and generated a pair of abject fails. Not a single relevant fact was presented in either 4+ page, double-spaced document. All fluff with appropriate names and dates plugged in after the bot scrubbed the intertubes.

Or annoying and costly, in the case of the project I'm working on for the non-profit I do. It involves plugging the output of a sound-generating device into an iPad so an app can show stuff on its screen relative to the sound coming in. My developer uses ChapGPT to answer questions about things he is coding. Asked about the nature of the connection we need to make and received an inaccurate response which led him down a paid rabbit hole he didn't need to visit.

Last time I tried it (emphasis on 'last'), Google's Gemini couldn't properly answer a question about gas station hours.

If it can't get simple stuff like that right, I'm certainly not going to trust with it anything important.

It's not like you're going to have a choice, unless you remove yourself from tech. They is continuously lobbed in politics to become the default use and regulated out the competition. Companies are "saving" costs and shifting blame when things go wrong constantly. Bribery will be a thing regardless of region, so good look regulating that. More online companies are spinning up, taking data and disappearing when things go south.

So while you're not going to trust it with anything important, the important things that other people are going to be responsible for? They're probably going to offload their workload to those machine learning models. Think of all the data breaches that has been found this year. From downloading free programs that took saved cookies from hardware and uploaded once connected to the internet, to simply calling and faking as someone else to obtain credentials, the enshittification of life is clear as day.

Kaito, you said more than a mouthful here and what you wrote of - I am seeing it every day, in the craziest of ways and places. Enshittification has become ubiquitous.

<<<this comment was produced by the brain of a human>>>

Edited 3 time(s). Last edit at 11/17/2024 01:03AM by BarefootBliss.

I used AI to assist in writing a few narratives for an apartment shop. It was very useful. I used my words and incorporated theirs.

@ServiceAward wrote:

As a follow-up, Gemini is better than ChatGPT IMHO

Gemini was a total joke the first (and only) time I tried it. Asked it for the hours at a gas station on a route and it came back with a totally incorrect response.

While only one of my students has been stupid enough to try to do a research project using AI, the results were laughable.

Not a single fact about the subject in 4+ pages of double-spaced text. Just a bunch of computer-generated puffery. They received 1 out of 250 available points because the subject's name was at the top of the paper.

AI is aptly named. It's artificial, and not worth a flying crap in my book. I require real intelligence...



Edited 2 time(s). Last edit at 11/22/2024 02:52PM by CoolMusic.

For mystery shopping narratives, I write it from scratch. It's not as efficient. However in the past, I ran into some issues. It helps force me to write each report as a unique experience independent from each other.

For some of the MSCs, I'll get the same editor, where I'll get feedback that builds upon previous feedback. If you do a high number of reports, the same editor may be familiar with your reports over time. To me, I personally like this. The one thing I would try is Grammarly to polish up my reports, but I never got to it.

For work, plugins like GitHub Copilot are useful. Having started out in a different industry and switching between technical and non-technical roles, and identifying as a non-technical person, it can be helpful to sometimes get you started. You can approve/reject suggestions, and over time through your feedback, it'll return more relevant suggestions in context with where you are. But it's not as beneficial as having a peer review, where a more experienced person reviews the work, and can point out things like company standards, best practices, performance, and efficiency. For work, we also use an internal form of Gen AI, where I found it more useful to translate things like patterns and restrictions that are black and white, and formulate business documents.

An analogy is if English is not your first language, it can help formulate something you struggle with from scratch and serve as a crutch. But ideally, as you become fluent in something and have a good understanding of it, as others have mentioned, it becomes more about refining the output to your needs.

@CoolMusic wrote:

@ServiceAward wrote:

As a follow-up, Gemini is better than ChatGPT IMHO

Gemini was a total joke the first (and only) time I tried it. Asked it for the hours at a gas station on a route and it came back with a totally incorrect response.

While only one of my students has been stupid enough to try to do a research project using AI, the results were laughable.

Not a single fact about the subject in 4+ pages of double-spaced text. Just a bunch of computer-generated puffery. They received 1 out of 250 available points because the subject's name was at the top of the paper.

AI is aptly named. It's artificial, and not worth a flying crap in my book. I require real intelligence...

It is improving by leaps and bounds. The key is knowing what to input and how to word it in order to get the results you want. I use it to design graphics that would take me hours to do in Photoshop. OpenAI has improved since my previous post.

It is likely your student has no clue how to properly utilize it. From the sound of it, you didn't either.



Edited 1 time(s). Last edit at 12/03/2024 02:46AM by ServiceAward.

@ServiceAward wrote:

From the sound of it, you didn't either.

You just earned a hearty GFY.

In response to a Google prompt, instead of typing in the typical search thread in the the appropriate field as is my MO, I decided to play along and 'asked' Gemini, "What are the convenience store hours for the XXXXX gas station in XXXX, XX?"

It came back with incorrect information, presented in a flowery screed.

Please tell us how you, The Obvious God Of All Things Digital, would have used it to obtain the desired information.

SMH

@CoolMusic wrote:

@ServiceAward wrote:

From the sound of it, you didn't either.

You just earned a hearty GFY.

In response to a Google prompt, instead of typing in the typical search thread in the the appropriate field as is my MO, I decided to play along and 'asked' Gemini, "What are the convenience store hours for the XXXXX gas station in XXXX, XX?"

It came back with incorrect information, presented in a flowery screed.

Please tell us how you, The Obvious God Of All Things Digital, would have used it to obtain the desired information.

SMH

I've never asked it for store hours or a location, because I would Google that type of thing. If that is your idea of what Gemini and other AI system are for, then you do not understand the premise of AI. It isn't being designed to only handle the basic things we can simply Google. It is being designed to actually do work. That said, after reading your post, I did ask Gemini random questions about store hours for stores here in town, the address to random Walmarts across the country, and the nearest 7-11 to my location. Gemini gave accurate answers each time. If you got a wrong answer, simply hit the thumbs down button on it so the programmers know. It is not yet 100% accurate. It is called machine learning for a reason. It learns over time. The information it gives should always be checked - it even says to double-check. That said, this thread was started in January. Since then, I've seen it improve exponentially in the months since. In January, I could ask AI to write me a synopsis on the book "Charlotte's Web." Back then, ChatGPT has no idea what to do. Today, it can give back a report as detailed as you would like on any aspect of the book. And, it is generally quite accurate. It is definitely not garbage. While you may be able to detect an AI written paper now, a year from now it will be increasingly harder to do so. If I were a teacher, I would get on board, I would learn the ins and outs, rather than mock it, and figure out how to harness it to my advantage. If you don't, your students will be light-years ahead of you and they'll be knocking anything you throw at them out of the park because they will be utilizing this tech.

I've had to modify my opinion on this over the past year, as I've used it more and I've seen what it can do expand.

You came in here in effect putting down anybody who might use AI. Don't knock me because I hit you back.



Edited 2 time(s). Last edit at 12/03/2024 10:51PM by ServiceAward.