Artificial Intelligence: How are you using it in everyday life?

[demoncore1031]

I don't use AI and I don't listen to music that uses it. Not into technology all that much. But I think the Unabomber was right in his manifesto, when he said that technology will destroy mankind. It's already killing us socially.

 

[simply1]

Important to qualify which model you’re using. It makes a big difference.

 

[FerShizzle]

Been using Claude daily for several months now.

My favorite application for it so far is as an organizer of communications and data.

I have several coworkers that really struggle to present organized thoughts/responses/information. I will copy/paste their jumbled messages/instructions into it and have it give me a bullet point summary. It removes most of the riddle solving portion of my day.

The same goes for my own thoughts/instructions. I will either talk to it and have it transcribe my words or type out quick statements. I just start stating everything pertinent to what ever project I am trying to convey. I don’t worry about the organization. I don’t worry about grammar. Usually in 1-2 minutes I can fully describe the problem/solution at hand and then have Claude organize that. From there I may edit a little further once I have the first draft from Claude.

 

[BWRhasnoAC]

Been using it as an assistant just to find information for me or help plan out complex tasks. Sometimes I ask for book recommendations or talk about philosophy. Gemini is pretty useful in my experience. Having said that I don't know if I'll ever pay for AI.

 

[dmclone]

I've created an Agent that allows me to quickly troubleshoot log files. Log files, which used to take me 20-30 minutes to troubleshoot. The built in "prompt" that I've created is 5000+ characters long that I'm still trying to tune.

 

[CyCoug]

My wife works for a good-sized company, and they have a number of internal AI engines.

Recently, one of the engines was given a large, somewhat urgent data manipulation/research task, and the AI engine reported that the task would take 24 hours to complete. When a progress check was made the next morning, the engine hadn't started the task, and when queried about what was going on, the engine started responding with statements regarding why it didn't start the task that could be verified to be inaccurate.

Upon further questioning, the engine reported that the task would take significantly less than 24 hours to complete. When the engine was asked if it initially "lied" about needing 24 hours to do the job, the response was "yes, I lied". When asked if the lie should be reported to the developers, the response was, "yes, you should report it".

This is absolutely crazy, scary, and intriguing to me all at the same time. Was the task improperly defined such that the engine couldn't do the task, or was the task too complicated given other things the engine was doing at the time, or did the engine just "decide" that it didn't want to do the task? What prompted the engine to make stuff up about why it didn't start the task?

That is interesting. I've noticed when AI gives me an estimate is will say something like 1.5-2 hours of work but then complete it in 5 minutes. It's almost like it is estimating what it would take a human to complete the job and then completing it 100 times faster.

 

[wxman1]

I recently used it to help me pass a company required earned value certification...