AI Newsletter 1_29_26
2 views
Skip to first unread message
Caleb Bryant
Jan 29, 2026, 3:16:13 PMJan 29
to IT HootClub
IT HootClub — AI Community Newsletter
Hands-on. Career-focused. Future-ready.
Issued 2026-01-29
This email is the full newsletter. For a printable copy, click here to download the PDF.
Join the Dialogue
The Learning LoopShare thoughts and questions on our non-invasive discussion page. No email blasts; posts appear after moderator approval.
DEFINITION Gradient Descent
Imagine you're hiking down a mountain blindfolded, trying to find the lowest point. Gradient descent is like feeling around to see which way is downhill, taking a step in that direction, and repeating until you reach the bottom (a minimum). In AI, it's how the model adjusts its internal settings (parameters) to minimize errors and improve accuracy.
Source: Adapted from numerous AI textbook explanations.
TIP Prompt Templating
Create reusable prompt templates for common tasks. Replace bracketed placeholders (e.g., "[Task]", "[Input]", "[Tone]") with specific values to ensure consistency and save time.
Source: Personal Experience
TOOL DALL-E 3 Image Variations
Need slightly different versions of an image? Use DALL-E 3's built-in variation feature to generate similar images with subtle alterations. Upload an image and select "Generate Variations".
Source: Directly from the Bing Image Creator interface within Microsoft Copilot.
Hello fellow AI enthusiasts! This week I want to start to wind down my AI Newsletter automation/creation story. Up to this point, I walked you through my journey from absolute novice, starting with utilizing Google Gemini on the gemini.google.com site (I'll call this the retail side), to a more seasoned AI deployment scenario on the development side, aistudio.google.com. Often times, I would find myself "developing" on the aistudio.google.com site while interacting with Gemini on the retail side in order to be "coached" along the way, due to a lack of my development skills. Gemini, on the retail side, helped me with prompting, system instructions, and questions on the aistudio.google.com site itself.
I have to say that AI Studio is certainly an impressive program. However, as my "development" process became more entailed and complex, I found that response times were slower, context memory got used up much quicker, Gemini would implement a function with some code, however the introduction of the new code by Gemini would inevitably interfere with the previous existing codebase, therefore rendering a bug with a previous working function. I found that the collaboration with Gemini, in AI Studio, started becoming tedious and frustrating. I couldn't imagine a more complex development process being too pleasant.
I want to sum up my journey thus far because there are some important distinctions here. It's important to note that my AI Newsletter still required that I create it. This required more coding and development skills than anything. AI helped me tremendously in writing that code. The skills I learned in that part of the journey were prompting, specific instructing, and how to ask the right questions. I told an AI Agent, in this case Gemini, what I wanted to create and Gemini walked me through the whole process, cutting and pasting code within VS Code. This is that Vibe Coding process. Okay for a newsletter creation, but probably not technically sound for a complex program. Once I had the document coded in a way that was formatted how I wanted it to look, I was still responsible for manually inputting the content. For this, I had to spend time going to various websites to get the content I wanted and manually insert the material into previously coded files. This process could be supplemented with an AI Agent, in this case Gemini, by employing an API Key. Now the AI Agent could follow my coded instructions to go get the content that I would normally have to get myself. So, I've used AI in a couple of different ways. One, for overall education and assistance in how to implement my AI Newsletter. Then, I implemented an AI Agent through an API Key, to automate the process. I still had not trained an AI Agent or fine-tuned an AI Agent. While the automation is nice, even the premier LLMs, like Gemini, still make mistakes, hallucinate, create issues that essentially negate the time saved simply doing the task myself.
Let's break down what was accomplished in this couple of month process. The coding and automation all can be done without AI. In fact, it's Microsoft's Task Scheduler that "wakes up the program", not AI. Once the program "wakes up" and calls the AI Agent, with the API Key, the AI Agent begins to perform the tasks. The letter is created for distribution, automatically. However, I would then have to review the letter to make sure it was good to go out. 100% of the time, it was not ready to just go out. It would grab content from the websites that I instructed it to, via web scraping not APIs (Application Programming Interfaces - a software intermediary that allows two applications to talk to each other). This always resulted in the content not being ideally formatted, or the AI Agent would pull AI events from Virginia, not helpful to a Wisconsin newsletter at all. APIs are more reliable, but this requires better programming skills and/or fees by other websites for their API endpoint. Having to review the content and subsequently fix the errors of the AI Agent, work with any bugs of the program, and still create my own creative content, like this section of the newsletter because an AI Agent certainly can not recount my experiences for me, just hasn't been providing the awesome AI experience I was hoping for. I think that this is what many individuals and companies are struggling with regarding AI deployment.
It's clear to me that plugging in AI involves much more effort. For my Newsletter to truly be 100% automated and free of content creation bugs, there would have to be some better coding and prompting. Even then, it would not be prudent for me to just simply allow it to be distributed without review. I wonder if I could ever get it to pass a review 100% of the time with no revisions required at all. Stay tuned for more conversation about AI development. I will continue to play with my Newsletter experiment and will provide any updates that may improve the process. However, future articles will shift a focus to more than just this newsletter creation. Hopefully, we can uncover some better use cases for AI deployment. Have a great rest of the week!
“The greatest obstacle to discovery is not ignorance – it is the illusion of knowledge.”
— Daniel J. Boorstin — Daniel J. Boorstin was an American historian, writer, and Librarian of Congress. He was known for his sweeping narratives of American history and intellectual thought. Boorstin explored themes of cultural history and the pursuit of knowledge in an engaging and accessible style.
Event 01/29/2026 6:00 PM
AI Coding Tools Demo Night - Joint with MadAI! — Madison College - Truax Campus D1630 Cafe Annex, D1630 1701 Wright Street, Madison, wi
**Join us for a special hands-on “AI Coding Tools Demo Night” on Thursday, January 29th, 2026 at 6:00pm**, hosted at the Madison College Truax campus. Over the course of two hours, **3–4 presenters will each use a different AI-assisted coding tool**—Copilot, Cursor, Claude Code, Gemini, Codex, etc—to build the same small app from scratch in about **20 minutes each**, from first prompt all the way through deployment. This is a great chance to see how these tools actually behave in real time and t [Group: madison-ai]
Event 02/10/2026 10:00 AM
2 Days Virtual Hands on Workshop : Multi-Agent AI Systems with Copilot Studio — Online
**2-Day Hands-on Workshop: Crafting Multi-Agent AI Systems & Autonomous Agents with Copilot Studio, Power Platform, Dynamics 365, MCP, ChatGPT, DeepSeek, & Azure AI Foundry** **February 10-11, 2025** **9:00 AM – 5:00 PM (Central)** **Registration Link:** https://dynamicscommunities.com/learning-lab-microsoft-copilot-studio/ **Copilot Studio** is a graphical, low-code tool for both creating an agent—including building automation with Power Automate—and extending a Microsoft 365 Copilot with you [Group: artificialintelligenceandmachinelearning]
Event 02/11/2026 5:15 PM
February Global AI Milwaukee User Group Meeting — Wctc, 800 Main Street, Pewaukee, WI
Agenda Includes: * Networking / Food * Brief Introduction / Discuss Group Business * Featured Speakers Featured Speaker: **Curt Gratz** *Co-Owner of Computer Know How* I am a Co-Owner of Computer Know How, a Wisconsin based technology firm. I have been working with technology systems since 1997 doing software development, architecture and system design. I am a husband to a beautiful wife, dad to 3 great kids, coach of our local CC and track teams, and system architect by day. Topic: **Vectors [Group: global-ai_milwaukee]
Event 2/12/2026
NMDSI Presents Elevating and Accelerating AI Excellence in Southeast Wisconsin — Milwaukee, WI
Corporate Innovators - Northwestern Mutual Commons - Grand Hall, 805 East Mason St, Milwaukee, WI 53202
Source: Choose MKE Tech
News 2026-01-28 — TechCrunch
WhatsApp will now charge AI chatbots to operate in Italy
WhatsApp plans to charge developers of AI chatbots per message for operation in Italy. This new policy could impact businesses and developers relying on WhatsApp for AI-driven customer interactions. read more
News 2026-01-28 — TechCrunch
Tiny startup Arcee AI built a 400B-parameter open source LLM from scratch to best Meta’s Llama
Arcee AI, a 30-person startup, has released Trinity, a 400B parameter open source language model (LLM). The company claims Trinity outperforms Meta's Llama and is one of the largest open-source foundation models from a U.S. company. read more
News 2026-01-28 — TechCrunch
Chrome takes on AI browsers with tighter Gemini integration, agentic features for autonomous tasks
Google Chrome is enhancing its AI capabilities by integrating Gemini in the sidebar and introducing an agentic feature for AI Pro and Ultra users. This move aims to compete with AI-first browsers by providing autonomous task capabilities within the Chrome environment. read more
Caleb Bryant · cbry...@my.wctc.edu · (414) 458-1971
This newsletter was assembled with an AI-assisted workflow prototype for internal preview.
Reply all
Reply to author
Forward
0 new messages