Body:
I decided to start to capture some of my thoughts and idea and move away from posting exclusively on LinkedIn. There are several reasons for this and I be exploring them in various ways as I post more on this site. This post, however, is mainly about the technical aspects of the blog. Boring geek stuff I suppose ... But, I thought it would be helpful to capture my journey as what I thought would only take a couple of hours but ended up taking a lot longer.
That was one reason to create the blog. To learn how to create a blog from scratch using Ruby on Rails.
To be fair, I did similar things in the past but I drifted away from my Rails tinkering. Instead of paying for an existing blog service such as Ghost.org, substack or Medium - I wanted to create a "Hello World" blog using Rails 8 from scratch. This was as a way to re-engage with Rails and to hone my technical skills as I strive for the larger goal which is to evaluate AI.
Before getting to AI a few comments about Rails.
I believe AI will be a good thing for Rails, but I also believe that the Rails is missing an opportunity and should fully embrace AI in a much bigger way. I had a decision to make:
Learn Python and all the frameworks that came with such as Flask and Django or I could double down on Rails.
Again, Rails 8 captured a bit of a buzz and I had to see for my self. I watched a few comparison videos exploring Rails alternatives and I was re-sold on the power of Rails and Ruby. Particularly - I love the classic Full Stack Development mindset that just comes more naturally with Rails.
I like having to skip the Environment step that comes with Python web development. I also feel like you get a lot of practical but powerful functionality with Rails right from the start. Full automated testing, JSON, Javascript, SQL... and much more. Yes, all of this is possible with Python Flask/Django, but I simply landed on Rails.
Why Rails for AI?
For all the hard core math - Yes, Python is king. But to connect everything together and to explore many configurations of AI such as Agents, voice as the front end, API calls, multiple databases, and more --- I just feel like Rails makes all of these things easier and more robust. The idea of "Rails as an Agent" interacting with other elements like a microservice ... it just all makes more sense to me.
This is another reason for the blog. The blog will evolve and will connect and interact with other Rails applications. I want to build Hello Worlds around AI and Optimization which I will display here and write about.
Again ... the idea is for me to take a hands on approach and to decide for myself what is AI Hype and what is legit.
This blog signals the jumping off point of my AI journey.
Here are more Technical Details of the Anatomy of this blog:
That was one reason to create the blog. To learn how to create a blog from scratch using Ruby on Rails.
To be fair, I did similar things in the past but I drifted away from my Rails tinkering. Instead of paying for an existing blog service such as Ghost.org, substack or Medium - I wanted to create a "Hello World" blog using Rails 8 from scratch. This was as a way to re-engage with Rails and to hone my technical skills as I strive for the larger goal which is to evaluate AI.
Before getting to AI a few comments about Rails.
I believe AI will be a good thing for Rails, but I also believe that the Rails is missing an opportunity and should fully embrace AI in a much bigger way. I had a decision to make:
Learn Python and all the frameworks that came with such as Flask and Django or I could double down on Rails.
Again, Rails 8 captured a bit of a buzz and I had to see for my self. I watched a few comparison videos exploring Rails alternatives and I was re-sold on the power of Rails and Ruby. Particularly - I love the classic Full Stack Development mindset that just comes more naturally with Rails.
I like having to skip the Environment step that comes with Python web development. I also feel like you get a lot of practical but powerful functionality with Rails right from the start. Full automated testing, JSON, Javascript, SQL... and much more. Yes, all of this is possible with Python Flask/Django, but I simply landed on Rails.
Why Rails for AI?
For all the hard core math - Yes, Python is king. But to connect everything together and to explore many configurations of AI such as Agents, voice as the front end, API calls, multiple databases, and more --- I just feel like Rails makes all of these things easier and more robust. The idea of "Rails as an Agent" interacting with other elements like a microservice ... it just all makes more sense to me.
This is another reason for the blog. The blog will evolve and will connect and interact with other Rails applications. I want to build Hello Worlds around AI and Optimization which I will display here and write about.
Again ... the idea is for me to take a hands on approach and to decide for myself what is AI Hype and what is legit.
This blog signals the jumping off point of my AI journey.
Here are more Technical Details of the Anatomy of this blog:
- Rails 8 Database = Postgresql Ruby 3.3.0
- Host = Heroku Domain
- Registration = Namecheap.com
- CSS = csszero (https://github.com/lazaronixon/css-zero)
- Initial inspiration by DHH (https://www.youtube.com/watch?v=X_Hw9P1iZfQ). This is a thirty min video by the man himself DHH. If you don't know who he is all you need to do is search DHH ruby or DHH rails and you'll know soon enough.
I also used a lot of AI to create the blog. A portion of AI was used to "do the dirty work" of coding and I also wanted to evaluate the potency of AI to help speed up the process of building a blog using rails.
My conclusion is -- AI is very impressive cutting my time by 80%.
I used Anthropic Claude through VSCode and the Roo add-on. I also used the basic plan on Open AI. To get this point - I burned about $10.00 USD in API AI Tokens / Credits. I'll be sharing more details with all of this and might just be updating this Post which is unorthodox vs me creating a bunch of new posts. We'll see.
My Style
It will vary -- in the age of AI I almost think typos and errors will be cool in some regards. Do I really want an idea to go from a human to AI then to me? Or worse - a human creating a bot to create content for me to read to make them money? Why waste my time? Or, from your bot to my bot - taking the humans out of the loop completely? What fun is that and no thanks. So style and even a few mistakes will signal that I am not wasting your time (or mine) by using AI to write posts for this blog.
Ok, I digress but that is the whole point of the blog - to share some raw ideas and thoughts. I'll likely return to this post and clean things up and to make updates... when I do -- I'll keep track of the edits at the bottom of the post.
The primary goal of writing this specific post is to test the core blog functionality "on prod".
So, If you're reading this, then the test passed and mission accomplished!
Thanks for taking the time to read this "real human post" and for being a part of this journey with me!