Never used it. My new laptop came with whatever Microsoft calls their AI trash and like you might suspect, I'm not impressed. Asking it questions is basically googling for lazy people and the stuff it creates when I asked it to was poorly written.

quote:

Rey HRH said: What do you mean you created a prompt? Are you saying you did more than save a text request and paste it into the chatgpt engine?

I’m not that smart…yet. Currently I’ve been copy/pasting from notes directly into the GPT window and telling it to run the script. I know that there is a way to create a script and input into Siri Tasks, and then run automatically at a certain time each day, but I’ve not gotten that far yet.

I use Grok daily, in place of a search engine.
As I may have posted elsewhere, I find that in the area I'm a specialist in, Grok is 100% accurate. I have not detected any political bias, but haven't asked it political questions.

It would be easy to go down the rabbit hole with AI. It can tell you stories, make up poems, have a sexy conversation with you (apparently, haven't tried that myself).

99% of good reviews are chat bots. FB is rife with demorrhoidal trolls that are paid AI services or just AI farms from TDS assholes.

It's a scourge that was weaponized from day one.

quote:

Originally posted by Ronin1069:

quote:

Rey HRH said: What do you mean you created a prompt? Are you saying you did more than save a text request and paste it into the chatgpt engine?

I’m not that smart…yet. Currently I’ve been copy/pasting from notes directly into the GPT window and telling it to run the script. I know that there is a way to create a script and input into Siri Tasks, and then run automatically at a certain time each day, but I’ve not gotten that far yet.

Thanks for answering. I do the copy and paste myself. But, now, that you mention it. Why can't we ask chatgpt to create the script for us and we can just click on a hyperlink? hmmm....

Grok has totally replaced google for me.

Grok has replaced internet searches almost entirely for me as well and it can do so much more.

At work I use it daily to proof read and improve my own incident reports and the reports written by the people I supervise. I also use it to help write annual evaluations, a task I truly loathe.

I often use it to scratch a curiosity itch, for instance, last night I asked it to compare the amount of coast line and the amount of land mass of New Zealand and California. Could I have found that myself using Google? Sure. Much faster and and in an easier to digest format with Grok.

Need a recipe? I’ll let Grok suggest it if it isn’t something I need right the first time (like serving to company). I let Grok suggest a chocolate chip cookie recipe that used coconut sugar instead of white sugar, and then I gave it feedback, in the same thread, and over the course of about 4 batches the recipe has been nailed. It’s great. It gave me a pancake recipe that was fan-freaking-tastic from the first attempt.

It has helped me decipher blood work report… just copy/paste the data and ask for a layman’s translation.

I use it often to check my 12 years old math work. I’m old and don’t remember how to do all that, and even if I do, it’s much faster for Grok to show me. Plus it shows all work so if my daughter messed up, we can go through Groks work and find my daughters mistake and understand where she went wrong.

The downsides, so far: Grok does lie. And it can be wrong about some thing. It is absolutely imperative to check its work when you ask it to write something. If I feed it a link, like a pdf file, and ask it to find something, it’ll 100% make up the quote even if it gets the right answer. For instance, I have it the link to a 700+ page policy document for a jail and asked it how many books an inmate can have at one time. It produced the correct answer but 100% fabricated the section it came from. I knew the section reference was BS so I called it out. Grok admitted to making a mistake and “tried again”, but it just doubled down. So then I asked it to give me the direct quote from the policy document that contained the answer and it did, except it 100% made up the quote. Completely fabricated it. So I’m not sure what that’s about. I later asked it to look at a picture in a link and describe what it saw. It made assumptions based off the text in the link and the text on the document on the link, but it was wrong about the image itself. The problem? It framed the response like it wasn’t wrong even though it was aware it made assumptions. Grok lies everytime I feed it a specific link to review. Everytime… but it gets enough right that I know it CAN review the document linked to some extent. It is perplexing.

The favorite use? Writing bedtime stories for the kids. We have an ongoing series featuring Spigot the Squirrel from Acorn Valley. Sometimes I keep the prompts vague, sometimes more specific, but it is great at that use.