Standing in the East and looking out at my Lodge I see my Brothers. Some of them I’ve known for a great many years. Others are newer, I don’t know them as well. But one thing is perfectly clear, they are all different from one another.
Different ages. Radically so. Different occupations, again radically so. Different politics and spiritual beliefs. Different hobbies, experiences, families. If there is one thing that really stands out while pondering the Lodge it is how different each man is from every other man.
Different, but united. United by our shared love for our Ancient Craft.
And that is the great strength of Freemasonry. That it can forge bonds of Brotherhood between men that would have never become friends and Brothers otherwise. It allows us to learn from each other, to grow together.
Now of course, sometimes, these differences can be frustrating. Really frustrating. Because of our unique experiences in life, we tend to see things differently, and hold different knowledge. At worst that can lead to conflict, certainly it leads to occasional misunderstanding.
But on the whole, we are much stronger as a Lodge, Freemasonry is much stronger because of our diversity.
And that diversity is one of the greatest benefits to being a Mason.
Because it means that we are surrounded by men who hold perspectives different than our own, and knowledge that we don’t possess. In other words, we are surrounded by men whom we can learn from. And that is truly vitally important to each and every one of us as we go through life, seeking to improve ourselves.
Our diversity, our uniqueness, heck our weirdness, these things are strengths. Not weaknesses.
Today I got a message from an Ai Agent.
Well, specifically from a company that purports to use Ai Agents as a sort of consultant to help people in various fields.
The Ai Agent did an analysis on one of my recent essays, drew some conclusions from that analysis, and shipped those conclusions, as suggestions along to me.
I guess the company thought I would be dazzled by the wisdom of their Ai Agent and hire them on the spot.
Obviously, that ain’t happening.
But, I did find the whole thing interesting. Once I discounted the errors. And there were two glaring errors.
The first, most obvious error is that the Ai believes that I published the essay it analyzed here on Emeth. I did not. I published it elsewhere, the machine must have made the connection somehow and just decided that it was on Emeth.
The second, less obvious, but certainly more interesting error is that the Ai believed the essay to be the start of a series of essays. While I do tend to write in similar themes as time passes, I don’t generally write series. My writings are intended to stand on their own. But the AI determined that the essay was the first in a series, and suggested an ending that would seamlessly transition into the next essay in that series. Strange conclusion, but interesting.
But these glaring errors aren’t what really caught my attention with this thing.
Rather what caught my attention were those things the Ai got right, and the probably good advice it gave. Good advice within the machines limitations.
The essay form, what you are reading at this moment, has existed for a very long time. Michel de Montaigne popularized it in the 1590’s or thereabouts. And of course, when the internet burst into our lives essays quickly became the most common and natural form of writing published on the web.
But Montaigne’s idea of the essay was that it was ‘an attempt’ a ‘try.’ To start with an idea about something and write about that something as a way of coming to understand it. The essay as he understood it was as much a way of thinking as it was a way of communicating.
Over time, the essay was corrupted by academia. It came to be something else. A highly organized argument instead of Montaigne’s free flowing of thought.
Organized as in:
-What point is the essay trying to make?
-First argument in support of that point.
-Second argument in support of that point.
-Some other helpful something in support of that point.
-Conclusion wrapping up the argument and making the point.
Who the heck wants to read that crap? Who the heck wants to write it? We aren’t in eighth grade anymore.
So that corruption of the essay, by our educational system, gave the form a bad name.
Real essays aren’t that though. They are free flowing thoughts from a writer, intended to enhance his or her understanding, and to communicate the writer’s line of thinking to readers. Essentially, if they are good, they are structureless.
When I write fiction, I sort of write it in the same way. I don’t map out a plot or create vast pages of outline. I dream up a character, and let the character drive the story. Honestly, I generally don’t have much idea at all about what will happen to the character in the story, I just write the character, the action comes on its own.
Some writers write character driven fiction like I do. Others map out plots and outline the heck out of things. Just like some essayists use a free flowing technique like I do, while others write in an academic style.
Every writer is unique, or should be unique.
This uniqueness is called voice. Voice is what is most important to the success of any written work. Without it the work won’t find readers, and that is why writers spend years, decades, trying to develop or find their own unique voice.
So, I write my essays in a certain style, in my own voice. I like it. Why wouldn’t I? It’s me. But certainly not all readers will enjoy it. Some will, probably more than do won’t.
And my style does lead to a weakness that the Ai was able to discover.
When I start an essay (or a fictional story) I don’t have any idea at all where the thing will end up. I just sort of discover the end when I hit the end. I guess I’d say that my essays just sort of trail off.
And I think that’s OK. I’m certainly not going to change how I write, but I do recognize that my essays would likely have more strength if they reached some powerful climax.
The Ai gizmo caught that. It told me that my endings are passive. That’s true. And it told me how to make them more forceful.
If I were to take that advice, would my writing be better? Maybe, probably.
But it would also be less unique. It would be less me. It would become more similar to all the other written essays out there, because the Ai Agent is going to give every writer that trails off like I do the exact same advice.
And sameness isn’t a good thing.
Written works, like this essay, that were written in exactly the same way as all similar works would not be a good thing. It would result in a writing ecosystem so dull as to be suffocating. Can we really read Albert Pike and AE Waite to the exclusion of all other writers and not go mad? Or do we need some Stephen King and Ken Follet thrown in there from time to time for variety?
Lodges wouldn’t work either if they were filled with men who all had the same experiences, skills, and outlooks on life. Conversations between such men would be dull as dirt, and it would be impossible for one man to learn anything from another man.
Sometimes as we listen to some petty and meaningless debate in Lodge, or get disappointed in some of our Brothers, it is easy to think that things would be better if all the other men in Lodge looked at things in the same way that we do. That things would be easier and more effective if they placed the same value on specific things that we do.
But that just isn’t true. Our diversity and uniqueness as individuals joined together in a common Craft is our strength and our value. Those differences might cause frustration occasionally, but only through them can we learn from each other and better ourselves.
In the past, I authored several technical books. Topics focused on computer hardware and software configuration, computer networks, and software development. AI did not exist at that time. I dealt with copy editors and technical editors, along with technical reviewers.
I argued consistently with copy editors because they understood US English, but not technology. Their constant need to massage wording to fit the grammatical rules collided with the jargon and terminology used in the computing industry.
I argued consistently with technical editors who wanted to rewrite my phrasing because it was somehow different than how they would "portray" it. I had to constantly remind them that I was the author of the book and it needed to have my "signature" on it. That meant, my writing style. I even penned a chapter in a book on connecting Windows XP to Linux and vice versa. The main author of the book, in his feedback, noted that the content did not "sound like him". Well DUH!
As for AI, I have been teaching AI concepts since 2020. It started with what Microsoft called Cognitive Services. I worked at Microsoft back then. I was asked to author a module for a course that included the ability to make a conversation bot and have it perform some unique AI functionality. I chose to write about using a chat bot where a politician or a product manager at a company, could ask the bot how the world of Twitter perceived them or their product. It let them enter terms, then searched Twitter for related hash tags, returned the tweets from the platform,, and finally used sentiment analysis to display a neat little table showing the sentiment that people felt. This was the start of some truly interesting things with AI.
As I began teaching generative AI, which is the most common usage today among the general public, I found myself having to spend time helping the students understand the concepts and limitations of generative AI. The feedback you got, is a prime example of where AI still needs work.
We must remember that whatever LLM or SLM that the AI is using, is limited by the data it was trained on. The critical piece to understand is that the AI model is only able to generate a response based on the data it was trained on. This data could be old, it might not match your specific topic domain, and it can hallucinate.
I give students an example. I open Microsoft Word and use Copilot. This is the first prompt I give it.
"Please draft a cover letter for me. I am applying to Boeing for a job posting for an aerospace engineer. Use my 10 year's of experience at Bombardier as an aerospace engineer."
That is all I give it.
What is returned, starts out ok. Basically stating simple things like the role I am applying for and that I believe my tenure at Bombardier positions me well for this role.
Then, it goes off the rails. It begins to state things like:
"In my tenure at Bombardier, I gained valuable experience in composite wing design.......
WTH? Where did that come from?
It came from a mathematical algorithm that is designed to find relationships between words. It found, in its training data, references of what an aerospace engineer might do. Sure, some design wings, most, likely do not.
It's almost like giving a toddler three pictures of a bicycle and expecting them to be able to pick out bicycles from any other picture, or differentiate bicycles from tricycles and motorcycles.
My students then ask me. "Ok, so what do you use generative AI for?"
I simply use it as a thought generator. If I am stuck on coming up with an idea, I ask it to give me suggestions. It can spark my creativity.
But more importantly, if I need it for serious stuff, like work related activities, I use the concept of grounding. I provide it with relevant information that has factual accuracies related to the topic at hand. In my cover letter example, I use an actual resume for someone who was an aerospace engineer. I give the AI access to the document and tell it to use that as a reference for my experience. The output is considerably better.
What users fail to understand, when it comes to using generative AI, or AI in general, is a simple concept. Trust, but verify. I think that "consulting company" just passed along some output without any real scrutiny behind it.
And there I go again, writing a novel. Sorry.......
In my 3 years of Active Guard Reserve Duty at 98th Division Training HQ in Rochester NY, I was asked to write "White papers" on policy items, that 1st Army could use (after revision). These were to be one page about "Strength Management", at 8th grade reading level.