It’s good to make a blogpost every now and again just to remember how to do it…
Here at iamJoshKnox blog, we LOVE reader feedback. Last week, Jared reached out with a question about using the excel add-in to return an excel list filtered for open sales orders.
What a great excuse for a video (and blogpost)! Watch the video above (or send your AI-bot to watch and comment for you, that’s your decision).
Step-by-step guide on how to filter and retrieve open sales orders from Dynamics 365 using the Excel add-in, in response to Jared’s inquiry. Addresses common issues and offers solutions for effectively managing sales order data.
I was recently in Chicago for DynamicsCon, a top-tier community event for users and implementers of Microsoft’s Dynamics platform–Like Me!
The conference was an excellent chance to catch-up with old friends, make new ones, and learn about the latest happenings in the world of Dynamics (If you’re curious, there are a lot of happenings).
DynamicsCon was also an opportunity to learn about Chicago.
I took in a Cubs game at Wrigley with some friends.
My Uber driver suggested the best way to learn about Chicago is to take an architecture boat tour–boats ferry you around the Chicago river and they tell you about the buildings. I didn’t have time for a boat tour, but his recommendation primed me to think about the architecture and design of the Chicago skyline. As I walked around Chicago, I snapped pictures of buildings that interested me.
Problem: I didn’t have an architectural tour guide to tell me about what I was seeing. Solution: I built one.
–INTRODUCING– Willis: A Chicago Architectural Tour Guide
Sears Tower will ALWAYS be Sears Tower to me, but I named this bot Willis because it seemed like an appropriate cousin to Watson. Simply upload a photo of your Chicago structure and Willis will tell you a bit about what your looking at and how it came to be.
An advantage over an architectural boat tour: Willis can also go on land.
I’ve written other posts on how to create a custom GPT. If you are interested, these are Willis’ configurations:
Without further ado…
A Chicago Architectural Tour, Guided By Willis
Pictures by me (All Text by Willis):
Alright! You’ve captured some amazing shots of the Merchandise Mart—or as us locals call it, “The Mart.”
🏢 Name: The Merchandise Mart 📍 Location: 222 Merchandise Mart Plaza, Chicago, IL 60654 🗓️ Year Built: 1930, during the era of Art Deco’s peak style (think Al Capone and the start of the Great Depression!) 🎨 Notable Feature: The building’s imposing size—it covers over 4 million square feet—earned it the title of the world’s largest building when it opened. That’s like 22 football fields of floorspace! 🏆 Historical Fact: Originally owned by Marshall Field & Co., it was later bought by the Kennedy family in 1945. The Mart also played host to radio station studios in the 1930s and, more recently, has become a major hub for design showrooms and tech startups.
Fun tidbit—when the Chicago Bears last won the Super Bowl in 1985, The Mart was already well-established as a local icon!
The building overlooks the Chicago River and is known for its striking setbacks and vertical emphasis—a classic example of Art Deco style.
It’s a new year, OpenAI released a new model, and we might be at the dawn of Average General Intelligence (AGI) – a world with AI systems capable of understanding, learning, and applying intelligence across a wide range of tasks at human levels or beyond.
What are the implications? What will an AGI future look like?
Dario Amodei, CEO of Anthropic, recently wrote Machines of Loving Grace speculating on 5 domains that will be radically transformed by powerful AI. It’s essential reading on the topic.
I have my own hunch: the future’s already here, it’s just unevenly distributed.
“A simple way to forecast the future is to look at what rich people have today; middle-income people will have something equivalent in 10 years, and poor people will have it in an additional decade.”
This is a clear pattern looking back through history:
Automobiles
Karl Benz invents the first automobile in 1885. It’s a luxury good.
In 1908, Henry Ford brings his model T to the masses.
Today, you can get in Twitter fights for suggesting car ownership is anything other than an essential right.
Instead of using ChatGPT through the web browser, you can use Microsoft Edge to access it as a desktop web app. I find this to be useful for frequently used websites because the application is always available when I pin it to my toolbar instead of getting lost among all the other browser tabs I may have open.
OpenAI (the ChatGPT company) is working on a Windows desktop app. It’s currently in trial mode and only available to ChatGPT plus customers.
Recently, I was loading an updated file into a SQL table. Before proceeding, I wanted to compare it with the original SQL table to identify data differences. This can be time consuming, but I knew just who to turn to for help: my friend Chatty-G.
The Prompt:
Both tables had the same column structure, making the comparison straightforward. Using SQL Server Management Studio (SSMS), I right-clicked on my SQL table and selected “Top 1000 rows” to easily get and copy the column names from my table.
I then pasted the select statement from my first table, and the name of the second table into ChatGPT and explained what I wanted in 2 poorly worded sentences:
The Performance:
Then, within 30 seconds, ChatGPT returned the ~100 line comparison CTE below:
…
The Process:
I copied the code into SSMS and executed it. No errors!
The Postscript:
The CTE here is not particularly complex. However, ChatGPT generated error-free code faster than I could type it. And if I had typed it, there certainly would have been an error to two to debug. This task is a good candidate for AI because it is relatively straightforward to verify the accuracy of the answer provided. I can easily see this causing trouble if used to “solve” problems that are not easily verified.
There are higher level questions I could ask about the data. Is this the way I want to verify the changes? Maybe I should return mismatched values instead of a binary identifier. If I needed that, I could update the structure of the code myself or better prompt ChatGPT in that direction.
A last thought is that this was a two-step process, but I can see the compounding advantages coming as Copilot and other AI companions get baked into the core of IT and Data products.
In this video, I configure a custom GPT to generate dad jokes. I demonstrate how prompting a custom GPT can enhance standard ChatGPT’s ability to generate and express humor.
It’s a simple premise…the main takeaway I want to give you is that this is actually very easy to configure. If you have an idea, it does not take a lot of special knowledge to build a prototype.
Dad[AI]Base Use Demo:
Want to test Dad[AI]Base’s humor yourself? Try it here!
