Using AI

[bjl584]

Is anyone using AI for anything productive?
There are lots of fake videos popping up online, but is anyone actively using AI for work or meaningful research?

[james.hendrickson]

So, its been a ton better for summarizing information about specific topics - instead of reading say, 15 news stories, you can just do an AI query.

[kork13]

Yes! I'm ordinarily an AI hater .... but the military uses a version of Google's Gemini that is approved for certain types of controlled information. The last couple weeks I've been writing an advocacy paper for my boss's boss to go to bat with at our higher headquarters ... but I have a tendency to be quite wordy (as many here may have recognized over time). After I sat down with my boss & worked through some ideas for edits, he suggested I try using the AI thing.

I gave it a shot, uploaded my file, and asked it to condense it down to roughly 1 page. It took 2.5 pages of information down to just over 1 page, consolidated ideas that I had duplicated in a few places, and generally used much more efficient language to get generally the same ideas across. I still had to go through & edit it to get the message exactly right, which took it back to 1.5 pages. I re-ran that draft through the AI again, but this time it stripped out way too much detail. That told me that I probably had it about as compact as I reasonably could while still retaining its meaning & impact. So with a few final tweaks, I submitted it and my boss loved the new version, immediately sending it unchanged to the big boss.

I was honestly impressed by how useful it was in repackaging my statements & information into more efficient language. I'm still very wary of using AI stuff, because I feel like alot of it is gimmicky or pointless. But used in the right way, I can see certain applications where it definitely has utility.

[riverwed070707]

We had this chat a couple months ago. I use it every day. A little bit for work, a lot for personal - travel, journal prompts and analysis, productivity and task prioritization, and just general knowledge type questions because I can dig in and get specific and ask follow up questions without having to click through 15 google search results. I also really love it for translations over google or the translate app because it gives you options for natural language response and takes context into consideration instead of just a literal word for word conversion.

[ua_guy]

Using AI is one of those things that I've committed to allowing into my life in a controllable and experimental manner because it already is/ will be a "thing". This is a pivotal moment, much like home computers were for my parents. That left a lot of people behind. I don't want to be left behind.

I'm already behind. AI is becoming part of my work, although I'm only involved in implementing the business end of it. The modeling and back end is handled by highly specialized people. We are using large language models to respond to inquiries with clinical information integrated into AI-generated letters. It's pretty amazing. We're also using a reporting tool which allows end users to tell the system what kind of data they're looking for, versus having to write a query using SQL, or use a complicated user interface to build a report from a data model.

At home, I've been using AI as a starting point for questions. Because we'd maybe like to live somewhere warm someday, we follow the real estate market where we like to vacation near Palm Springs. So I prompted AI with questions to analyze the surrounding market conditions, and the answers were pretty amazing - with references.

My goal with AI - to begin with a better, more informed starting point than a Google search for my questions. AI supercharges that. It's been really helpful.

It also helped recently as I needed to formulate some goals for my annual evaluation at work. So I had ChatGPT write me a sample resume with skills/qualifications for a job title that I'm trying to reach. I used those skills/qualifications to tailor my own goals.

Since we're doing retirement planning with a CFP, I've been asking AI questions about timelines and trajectory. It will blow your mind with how incredibly detailed it gets and the information it can incorporate into its model.

If anyone wants to kill a few minutes, start on ChatGPT.com. No signup or account is needed. Have fun. Don't give it personal information or data that shouldn't be shared publicly on the internet.


For someone starting out on this forum, you can tell ChatGPT "I need help developing a household budget because I don't know where to start." And it will take it from there.

[ua_guy]

I definitely struggle with this. I don't want to be "left behind" but I'm also not going to use AI just for the sake of using it. I oppose what it is doing to the environment and electrical grid. I oppose what it is doing to writers and artists and photographers and other creative folks. So far, I haven't personally found anything I actually need it for, nothing that would benefit me to turn to AI. I don't need it to write an email for me. I don't want it to write my ebay descriptions or generate images for me. I enjoy travel planning myself. That's part of the fun of the trips. I don't like the AI-generated search results on Google. I don't need AI to manage our portfolio, nor would I trust it to do so. I'm retired so any integration of AI into the workplace isn't relevant to me. I've had multiple highly frustrating encounters with AI customer service systems.

I'm sure that at some point, something will come along that I can use it for, but until then, I'm not feeding the beast for no good reason.

All of those are valid concerns. A lot of AI is junk, and a lot of AI is poorly implemented. AI is used for nefarious purposes. It's not eco-friendly. But it's also used in a lot of interesting and beneficial ways.

In my own adventures with it, I haven't replaced things that I would rather do myself or which would have dire consequences if AI got it wrong. It doesn't deserve that level of trust. It doesn't manage my portfolio, nor would I let it, at this stage.

I still think there's worth in developing skills around how to prompt AI to return meaningful information or do meaningful things. And also learning how/where to fact check AI and learn about its sources.

The Google AI summaries are terrible, IMO.

Case in point: We had forgotten how much bleach should be added to our RV's water tank to sanitize it. For a 120-gallon water tank, it recommended 18 gallons (!!!) of bleach. LOL. That would destroy plumbing and seals, and the hot water heater. In reality, it's more like 1/4 cup per 15 gallons. AI can give bad, if not harmful, information.