- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
re: AI in the Workplace - Open Discussion
Posted on 5/15/26 at 9:22 am to AaronDeTiger
Posted on 5/15/26 at 9:22 am to AaronDeTiger
I’ve always generated HTML for my projects to keep actual project status and findings. So maybe I was just ahead of the curve a little (but I find it hard to believe, I’m just guessing I don’t understand).
I’ve been creating dashboards for each project I work on to keep all change logs, statuses, bugs/enhancement requests etc. because my brain is not reading through markdown files over and over.
Markdowns for Claude. HTMLs for me.
I’ve been creating dashboards for each project I work on to keep all change logs, statuses, bugs/enhancement requests etc. because my brain is not reading through markdown files over and over.
Markdowns for Claude. HTMLs for me.
This post was edited on 5/15/26 at 9:25 am
Posted on 5/15/26 at 9:27 am to AaronDeTiger
quote:
its also a different way of working with AI. Instead of editing or reading long markdown files, the AI can build interactive HTML docs for you. You can copy and paste different outputs via the HTML as well. I'm still wrapping my head around all of it too.
It's usually just a combo of html, css, and javascript to do the interactivity pieces and heavy lifting. It's really just sort of a minimalist web app when it comes down to it.
Posted on 5/15/26 at 9:58 am to Chromdome35
quote:
Sounds like a perfect use case.
I watched a YouTube video a while back on where an accountant was using Claude to analyze client bank statements and upload to QuickBooks. Was pretty impressive.
claude also does a much better job of analyzing quickbooks data than quickbooks. quickbooks reporting is weak and cumbersome.
Posted on 5/15/26 at 2:30 pm to Lazy But Talented
Director in a small department on the operations side of a Healthcare Technology company. Like so many, our leadership has really been pushing AI for about two years now. In 2025, we couldn’t get away from it. In 2026–after basically driving one of our top talent Engineering VPs absolutely insane and causing him to leave—they have dialed back the pressure noticeably. The tools they have endorsed are trash IMO. Amazon Q Business and MS Co-Pilot. I’ve gotten much better results from ChatGPT but this year they’ve started cracking down on its usage for anything that involves data. I get it, but I’m extremely frustrated by people who don’t have to use AI for jack shite (i.e., the CEO) telling us how to use it.
Rant complete.
Rant complete.
This post was edited on 5/15/26 at 2:33 pm
Posted on 5/15/26 at 2:46 pm to Lazy But Talented
Claude can generate beautiful HTML based dashboards. I recently built a dashboard for one of my area's of responsibility by dropping in 400K rows of salesforce data and having it slice, dice, and visualize the data. What would have taken me a couple of days due to the volume of analysis took a couple of hours of work with claude.
The initial generation was good, but there were lots of things that needed to be fixed before it could be released into the wild. The prompt/response thread was long.
Claude will get you 80% there in a matter of minutes, closing the remaining 20% takes a while. However, due to being able to reuse the dashboard, "Update the dashboard to reflect June 2026 activity using the new data file" only takes a few minutes once you have the dashboard created. So I view the time spent designing it as an investment that will amortize across the life of the dashboard.
The initial generation was good, but there were lots of things that needed to be fixed before it could be released into the wild. The prompt/response thread was long.
Claude will get you 80% there in a matter of minutes, closing the remaining 20% takes a while. However, due to being able to reuse the dashboard, "Update the dashboard to reflect June 2026 activity using the new data file" only takes a few minutes once you have the dashboard created. So I view the time spent designing it as an investment that will amortize across the life of the dashboard.
This post was edited on 5/15/26 at 2:49 pm
Posted on 5/16/26 at 3:46 pm to AaronDeTiger
I’ve started using it to help inform strategy. Take a complex problem statement and then have AI write a wildly complex and layered prompt for itself and then execute the prompt.
It produces work that would have taken months in 30 mins. Big acceleration to coming up with creative ideas.
It produces work that would have taken months in 30 mins. Big acceleration to coming up with creative ideas.
Posted on 5/17/26 at 5:24 pm to lynxcat
quote:
It produces work that would have taken months in 30 mins. Big acceleration to coming up with creative ideas.
I coach my team that, yes, AI can produce some incredible work; however, if you don't take the time to consume it and validate it so you understand and make the final decision, you are working yourself out of a job in the long run. Companies don't need people who just pass AI work off without adding to it, they need people who can use AI to make themselves better.
I hate people handing me stuff that's obviously AI-generated, and they don't know a darn thing about what it says. Some people think that if AI generated it, it must be good. AI can be an incredible tool that makes you much more efficient, but you have to remain in control and be the human part of the human-in-the-loop.
Posted on 5/19/26 at 10:44 pm to Chromdome35
I agree. AI has pretty big gaps but it’s been a great accelerator for me.
For example, if I want to understand the top events in a city that might be a fit for a brand across the year with minimum of 10 events evaluated per month and scored based of a series of factors…I now have 120 ideas for a single city and it’s a replaceable process for fill in the blank location.
That work pre-AI is a massive workload that is now a brief exercise.
For example, if I want to understand the top events in a city that might be a fit for a brand across the year with minimum of 10 events evaluated per month and scored based of a series of factors…I now have 120 ideas for a single city and it’s a replaceable process for fill in the blank location.
That work pre-AI is a massive workload that is now a brief exercise.
Posted on 5/19/26 at 10:47 pm to Chromdome35
quote:
if you don't take the time to consume it and validate it so you understand and make the final decision, you are working yourself out of a job in the long run.
I don’t understand the framing. Fear tactics of working yourself out of a job is a poor way to influence or motivate.
Posted on 5/20/26 at 6:45 am to Lazy But Talented
cyber security
FML. that is all
FML. that is all
Posted on 5/20/26 at 7:52 am to lynxcat
quote:
I don’t understand the framing. Fear tactics of working yourself out of a job is a poor way to influence or motivate.
Don't use it solely as an "answer machine" where it does all the thinking and work for you, and then you just regurgitate the results verbatim without even validating them. I could see how the idea that "Well if ai can do this without you, why do I even need you?" could sneak into a manager's head then. Instead use it to enhance and supplement your own work and knowledge, not just as a magic oracle/answer machine.
Posted on 5/20/26 at 8:03 am to jdd48
Fully agree. I’ve found the power to be in the iterative engagement with AI as a tool to refine rapidly. Output and re-prompt will get you a very strong output and it takes some critical thinking to drive that outcome (versus this magic first response hitting the mark).
Posted on 5/20/26 at 8:28 am to lynxcat
quote:
Fully agree. I’ve found the power to be in the iterative engagement with AI as a tool to refine rapidly. Output and re-prompt will get you a very strong output and it takes some critical thinking to drive that outcome (versus this magic first response hitting the mark).
Yea agreed as well. The way I approach it is "Could I still arrive at the same result, even if it is a (much) slower process, if AI suddenly becomes unavailable?" IMO if the answer is no, then someone needs to reconsider how they are using AI. I heard of two cases recently that are great examples of overall "brainrot" being caused by AI. One was a sysadmin on reddit that said he was being badgered non-stop by a dev after they got ransomwared about when chatgpt would be available again. The dev claimed he was simply unable to do his job without it. The other was a manager retelling a story of interviewing a recent grad and asking how comfortable they are working in an IDE. They actually told the interviewer "I am not entirely sure what IDE even stands for... but give me a Claude subscription and you'll be amazed by what I can do."
Posted on 5/20/26 at 10:47 am to Lazy But Talented
I've been playing with HTML some. Here's an itinerary for my sister and her friend's road trip back to her friend's house in Rhode Island. They wanted options for lunch stops, dinner + drinks afterwards, hotels under $220/night.
They both had different ideas on driving durations/stops so I put all their options. Has different buttons for navigation, print, google maps, clear picks, expand/collapse all.
Click on a day, then you can click on the lunch/dinner/lodging options and select them.
github
They both had different ideas on driving durations/stops so I put all their options. Has different buttons for navigation, print, google maps, clear picks, expand/collapse all.
Click on a day, then you can click on the lunch/dinner/lodging options and select them.
github
This post was edited on 5/20/26 at 10:51 am
Posted on 5/20/26 at 11:05 am to AaronDeTiger
Very cool. I do something similar with my trips haha. Helps keep me organized with packing for the kids and myself along with itinerary management.
Posted on 5/20/26 at 12:13 pm to jdd48
quote:
Don't use it solely as an "answer machine" where it does all the thinking and work for you, and then you just regurgitate the results verbatim without even validating them. I could see how the idea that "Well if ai can do this without you, why do I even need you?" could sneak into a manager's head then. Instead use it to enhance and supplement your own work and knowledge, not just as a magic oracle/answer machine.
This
Posted on 5/20/26 at 3:47 pm to AaronDeTiger
15 years of programming and DevOps here. Claude Code has impressed me for system design. Moreover, optimizing code testing and deployment pipelines. Tools like DeepCode help catch bugs early on, and services like AWS CodeGuru provide insights for performance tweaks. It's like having an extra set of eyes that never tires. Makes life a tad easier when juggling multiple projects from a Senior SDE point of view.
Posted on 5/21/26 at 12:30 am to Lazy But Talented
Tech Lead, we've had your typical automation, CI/CD, some AWS stuff; not bleeding edge but not stone age either.
We have gone all in on AI with Claude and Claude Code and building a process around how and when to use it, which tools to use where, how to set up the markdown files depending on who you are, etc...
It has sped things up on greenfield projects for sure. It helps on existing ones as well, but it'll send you on a wild goose chase if you're not careful.
There is definite pressure and the fear of missing out is well known.
My biggest issue is that it still has that sycophantic "personality" and tends to tell you what it thinks you want an answer to be and you have to badger it like a police detective to make sure it's not bullshitting - sorry, hallucinating, although instructions/skills can mitigate that to a point.
It's not fun when someone throws pages of an AI analysis and I have to say why it's wrong. I try to explain the context that is missing, but I feel like I'm doing that more often lately.
We have gone all in on AI with Claude and Claude Code and building a process around how and when to use it, which tools to use where, how to set up the markdown files depending on who you are, etc...
It has sped things up on greenfield projects for sure. It helps on existing ones as well, but it'll send you on a wild goose chase if you're not careful.
There is definite pressure and the fear of missing out is well known.
My biggest issue is that it still has that sycophantic "personality" and tends to tell you what it thinks you want an answer to be and you have to badger it like a police detective to make sure it's not bullshitting - sorry, hallucinating, although instructions/skills can mitigate that to a point.
It's not fun when someone throws pages of an AI analysis and I have to say why it's wrong. I try to explain the context that is missing, but I feel like I'm doing that more often lately.
Posted on 5/21/26 at 1:44 am to GetMeOutOfHere
FP&A manager
One thing I’ve really enjoyed in my FP&A role lately is using AI to improve my thinking process and logic development.
For example, I’m currently helping implement a BI tool at our company and rolling out dashboards across our teams. We’re giving people access to data they’ve never really had before (at least not this quickly or without a ton of manual work) and allowing them to analyze it in ways we never could previously.
What’s interesting is that I came into this with essentially zero IT background. But through using AI as a learning tool, I’ve become fairly comfortable with SQL, Python in Jupyter notebooks, and building logic within datasets to drive the right outputs and reporting.
The biggest benefit has honestly been the learning aspect. I don’t just ask AI for answers, I ask why it arrived at those answers, how the logic works, and what could be improved. It’s accelerated my learning curve tremendously.
I’m fortunate that while we still have deadlines and expectations, they’re realistic, and I’m given the freedom to build processes and help improve the company. It’s been a great balance and honestly a win-win.
One thing I’ve really enjoyed in my FP&A role lately is using AI to improve my thinking process and logic development.
For example, I’m currently helping implement a BI tool at our company and rolling out dashboards across our teams. We’re giving people access to data they’ve never really had before (at least not this quickly or without a ton of manual work) and allowing them to analyze it in ways we never could previously.
What’s interesting is that I came into this with essentially zero IT background. But through using AI as a learning tool, I’ve become fairly comfortable with SQL, Python in Jupyter notebooks, and building logic within datasets to drive the right outputs and reporting.
The biggest benefit has honestly been the learning aspect. I don’t just ask AI for answers, I ask why it arrived at those answers, how the logic works, and what could be improved. It’s accelerated my learning curve tremendously.
I’m fortunate that while we still have deadlines and expectations, they’re realistic, and I’m given the freedom to build processes and help improve the company. It’s been a great balance and honestly a win-win.
This post was edited on 5/21/26 at 1:48 am
Posted on 5/21/26 at 9:28 am to gmrkr5
quote:
cyber security
you don't like Charlotte ?
Popular
Back to top


1



