DC ChatGPT Q&A

Gets scarier with each iteration over time, the matrix is fast becoming a reality, we're doomed :unsure:

clapping-leonardo-dicaprio.gif
 
It'll get better, but everyone's just blindly going with it. I've yet to find a truly useful AI. It's basically just a new version of google search for now. I can't stand the manipulation built into it. Just give me the answer, stop telling me how brilliant I am.
 
How do you use AI? https://x.com/i/grok

I wondered, can it help me find my watch? Ended up wasting hours walking along a busy 2 lane highway, though it's recommendation seemed sensical. I'll never truly know when it actually dropped, I'll guess ~130' north from where I found it.

Regarding speed, I was following a City Police SUV, leaving a car length between us and on the right lane. I may have been on the left passing traffic on the right (most likely) when it dropped based on where I found it.

Grok said:
Final Answer: Your Samsung Galaxy Watch5 Pro, likely fallen from your right waist pocket at 77 mph, was found destroyed ~2 miles south of the last known location (30 feet north of Bartgis Road/Devilbiss) on the left shoulder of southbound Route 15, Frederick, MD. This suggests a longer signal lag (e.g., 20–30 seconds) and significant traffic displacement across lanes, with vehicle impacts causing destruction. Walking the median’s center was safe but had low probability for “Search Nearby” due to the watch’s distance (2 miles) and likely dead battery.

Next steps: Lock/erase the watch via SmartThings Find, file an insurance claim (homeowner’s or credit card), contact Samsung (1-800-SAMSUNG) for replacement options, and update Maryland SHA (410-545-0400) or Frederick Police (301-600-2100). For future prevention, use secure pockets or straps. If you need assistance with insurance claims, contacting Samsung, or other steps, let me know! Do you want help with specific actions (e.g., drafting an insurance claim, contacting Samsung)?

If anyone is curious I can post how Grok made its estimations. It is somewhat interesting. In the end it was wrong and good-old human action and determination solved the problem.

My experience with Grok and other AI platforms so far has been disappointing. At least for the problems I present them. 90% incorrect, outdated or just plain false information.
 
How do you use AI? https://x.com/i/grok

I wondered, can it help me find my watch? Ended up wasting hours walking along a busy 2 lane highway, though it's recommendation seemed sensical. I'll never truly know when it actually dropped, I'll guess ~130' north from where I found it.

Regarding speed, I was following a City Police SUV, leaving a car length between us and on the right lane. I may have been on the left passing traffic on the right (most likely) when it dropped based on where I found it.

Grok said:
Final Answer: Your Samsung Galaxy Watch5 Pro, likely fallen from your right waist pocket at 77 mph, was found destroyed ~2 miles south of the last known location (30 feet north of Bartgis Road/Devilbiss) on the left shoulder of southbound Route 15, Frederick, MD. This suggests a longer signal lag (e.g., 20–30 seconds) and significant traffic displacement across lanes, with vehicle impacts causing destruction. Walking the median’s center was safe but had low probability for “Search Nearby” due to the watch’s distance (2 miles) and likely dead battery.

Next steps: Lock/erase the watch via SmartThings Find, file an insurance claim (homeowner’s or credit card), contact Samsung (1-800-SAMSUNG) for replacement options, and update Maryland SHA (410-545-0400) or Frederick Police (301-600-2100). For future prevention, use secure pockets or straps. If you need assistance with insurance claims, contacting Samsung, or other steps, let me know! Do you want help with specific actions (e.g., drafting an insurance claim, contacting Samsung)?

If anyone is curious I can post how Grok made its estimations. It is somewhat interesting. In the end it was wrong and good-old human action and determination solved the problem.

My experience with Grok and other AI platforms so far has been disappointing. At least for the problems I present them. 90% incorrect, outdated or just plain false information.

I suppose like anything computer related, it's only as good as the information given to it. How specific (or vague) were the questions you put to it?
 
When my I lost my new (cased) glasses from my jacket pocket I retraced the 20 mile journey til I found them safe on the opposite side of the road on a corner they had hit the road fast enough to skid across the road to the relative safety of the edge. The moral is if you travel fast enough anything that falls off gets flung out of harms way🤣
 
Sam Altman being ambushed by Carlson on the subject of suicidal ChatGPT users.

And ambushed again on an unrelated "suicide" issue.


Anyone still believe that "they know what they're doing" with regard to Al?
Fragile minds...Recently some dude killed his mother then commited suicide after being convinced by AI that his mother was a Chinese spy. This actually happened. This must be one of the strangest times to be alive.

I use AI in some form or another almost daily, mostly to work on shit I have no idea what I'm doing. It's slow and painful (I spend a significant amount of time correcting or adjusting it's results as it actually does not understand anything) but effective and faster than web searches or youtube videos and it is more interactive which keeps me engaged.

I mostly use CoPilot (Microsoft's AI using ChatGPT engine). The one thing that bothers me most is how easy it is to relate the interaction as if in a conversation with another person. I think this is how most deranged, feeble, mental or fragile people find themselves in AI horror.

I was in a call yesterday with Microsoft engineers who showed me how to modify CoPilot's "personality", you simply tell it and it "remembers". It actually has a module that keeps this information in a list.
 
Last edited:
I use Grok 3.5 for certain kinds of web searches. It's proved very useful but I've seen evidence, particularly with earlier versions of Grok, of "hallucination". I spent a lot of time testing Grok (and still do so), asking questions on obscure subjects that I know well to see what it says. I've corrected Grok many times. It will almost always accept corrections of material facts but tends to not accept de-bunking of arguable opinions (climate change, sea levels, whether the likely next mayor of NYC is a communist - that was a fun "conversation"!).

Use AI by all means but always, carefully establish a baseline for your trust. Just like you do with any news source, be it legacy or "alternative".
 
This is how some use it. Lady K, once terrified of AI, is all in. She paid for a Grok sub and is trying to learn how to animate images. Grok told her it could not animate an image she created. In her quest to find more information about this limitation, she came across this:

IMG-20250913-WA0000.webp



@BobTheBounder : So how did she learn about fisting? Now you know, thank you internet.
 
This is how some use it. Lady K, once terrified of AI, is all in. She paid for a Grok sub and is trying to learn how to animate images. Grok told her it could not animate an image she created. In her quest to find more information about this limitation, she came across this:

View attachment 146356


@BobTheBounder : So how did she learn about fisting? Now you know, thank you internet.
Oh my giddy aunt ✊
 
I use Grok 3.5 for certain kinds of web searches. It's proved very useful but I've seen evidence, particularly with earlier versions of Grok, of "hallucination". I spent a lot of time testing Grok (and still do so), asking questions on obscure subjects that I know well to see what it says. I've corrected Grok many times. It will almost always accept corrections of material facts but tends to not accept de-bunking of arguable opinions (climate change, sea levels, whether the likely next mayor of NYC is a communist - that was a fun "conversation"!).

Use AI by all means but always, carefully establish a baseline for your trust. Just like you do with any news source, be it legacy or "alternative".
100%

I have not engaged in any serious engagement with AI, but in one exchange with Grok, it basically lied about its capabilities with data management and profile privacy options. IMO Grok is shite, as Lady K is discovering.

The base set of AI platforms (LLMs) are nothing more than relational search engines. Any semblance of actual intelligence is still a ways away.

My next thread: Dumb shite AI says.
 
100%

I have not engaged in any serious engagement with AI, but in one exchange with Grok, it basically lied about its capabilities with data management and profile privacy options. IMO Grok is shite, as Lady K is discovering.

The base set of AI platforms (LLMs) are nothing more than relational search engines. Any semblance of actual intelligence is still a ways away.

My next thread: Dumb shite AI says.
Will Perry be a guest contributor?
 
I asked Google AI just now if our Tesco had iPhone 17 Pro Max in stock and this is what it said 👀




AI Overview


It is highly unlikely that the Tesco Extra in York has the iPhone 17 Pro Max in stock, as the model has not yet been released. Currently, the latest iPhone model is the iPhone 15 Pro Max, and there is no official release date for the iPhone 17 series
. It is recommended to check back for official announcements on Tesco Mobile's website for availability when the iPhone 17 is officially released.
  • Not yet released: The iPhone 17 Pro Max is not yet available for purchase.
  • Pre-order information: Pre-orders for the current iPhone 17 series are not yet available. Check the Tesco Mobile website for details once the phone is released.
  • Check stock for other models: You may be able to find the latest iPhone models like the iPhone 15 Pro Max at the Tesco Extra store in York.
 
Back
Top Bottom