About Me
Last updated:
Me
I’m Michael Kelly. I have over 13 years of experience in IT. Twelve Years as a very technical Business Analyst, another year as a Business Intelligence Specialist.
Skills I’ve honed over that time:
Communication
I’m comfortable being very technical and talking with developers about how we need something to work or talking high level with a business user to understand their needs.
- Active Listening - When discussing something with a business user, I’m working to fully understand what the request is and getting to the why of the request. I find this ultimately leads to the best request that the business user is happy with.
Process Design
- Systems Thinking - One thing I’m always trying to do is fit something into the big picture as I understand it. So when learning about a new process, I’m trying to think about how it fits into the rest of what I know about the framework. I’m using analogical reasoning to try to conceptually understand what’s happening.
Technical Expertise
- SQL - When it comes to querying databases, this is something that I became very proficient at. SQL was always helpful in getting data together to examine certain issues more closely.
This site contains blog posts about topics I learn about in programming as well as some posts about various IT consulting projects I take on. You’ll also find posts about my principles/philosphy when it comes to consulting. I’ve posted a high level overview of my consulting principles here.
Programming has become a much greater focus in the past few years. The majority of my programming was originally in Python with a focus in Data Analytics. From there, my focus shifted to game development. At first I was working in Godot, but I discovered Ruby in 20204 and started using DragonRuby for game development.
Ruby has been awesome. I can work on a game in DragonRuby then flip over to a Ruby on Rails web development project. If I need to do a little scripting, Ruby is great for that as well.
My current thoughts on AI
TL;DR I see AI as something I can use as a second ‘set of eyes’ or something I can ask specific questions to when I can’t find a good answer by searching. If I’m writing something, AI might be used to review grammar or improve clarity.
The rest
To me, AI can be very good at certain things. I have long term concerns for a few reasons, but the two main ones are:
- Training data - AI has been trained on so many things, both open source and copywritten material. AI is now producing so much content that future AI will be trained using content AI has generated. Many times AI content is what people would call “AI Slop.” Who knows what will happen as AI is fed content from previous AI generations.
- Cost - Right now AI cost is being subsidized, quite heavily from my understanding, so what happens if Claude/GPT/etc. go from $20 a month to double, triple, or even 5x the price?
I think AI’s best role is as almost an assistant, but not in the way of something like Clawbot. For example, instead of using AI to write a first draft, AI is instead helping review the writing and improve clarity. It’s likely I won’t be running most of this site through AI as this is my site, not an AI bot’s site.
For programming, once again I find it most effective to have a chat with a bot for specific questions I have. My main goal with programming is to learn, so I have to code by hand, I have to make mistakes, sometimes I have to ‘bang my head against a wall’ for a bit to get a concept down. To me, the struggle is where learning happens.
If the end product of the code is the result, AI should be used sparingly in the beginning. I did use AI to help with the inital setup of this site as JavaScript is not an area I have dug as deeply into. The main focus is the content itself. So for a simple static site like this, a bit of AI can go a long way. I suspect as I get better at JS, I’ll likely need to do some refactoring to tighten things up.