Robert Newcombe ⚗️’s Post

I’m happy to share that I’m starting a new position as Part-Time Instructor at Dalhousie University! Along with Scott Flemming, I'll be teaching Industrial Engineering students how to effectively, responsibly, and safely apply generative AI tools in their careers (IENG 4598: Special Topics in Industrial Engineering II). I have fond memories of the industrial engineering computer lab and am excited to be leading the course labs this winter! What's one thing you believe 5th-year industrial engineering students should learn, experiment with, or build using generative AI?

We kicked off the first live sessions of the Digital Nova Scotia AI Microcredential Course yesterday! One simple concept I shared earlier resonated with many and sparked some ideas... I call it the AI Use Case Venn Diagram: → AI can do a LOT (and seemingly more every day). → What we know AI can do is just a fraction of its full potential. → Depending on your role, there may be only a small overlap between what you do and what AI can do. When you boil it down, there’s really only one thing to focus on: increasing your knowledge of what AI can do for you. Here’s a simple step-by-step framework to achieve that: 1. Identify a task that bugs you (either time-consuming, mind-numbing, and/or energy-zapping). 2. Research (using AI) how AI can assist with that task. 3. Experiment with it. 4. Repeat. Let me know in the comments what tasks you've outsourced to AI!

Don't know where to start with AI? Forget about the technology—understand your inefficiencies. Here are 16 wastes that may exist in your work and how you can use AI to help: 8 Wastes from Lean (TIMWOODS): 1. Transportation (Moving items or information)   → ChatGPT can streamline communication and information sharing, by centralizing information in custom GPTs. 2. Inventory (Items or information that customer has not received)   → Generative AI can help modify documents whenever is necessary, so you don't have to keep 10 different versions of the same document (e.g., cover letters, resumes, etc.). 3. Motion (Excessive movement within workspace)   → By automating repetitive tasks with tools like Zapier connecting to Generative AI APIs, it can reduce switching between different tools and documents. 4. Waiting (Waiting for information or items to arrive)   → Generative AI can provide instant responses to questions or status updates, reducing downtime when waiting for key information or decisions. 5. Overprocessing (Doing more work than necessary)   → Generative AI can draft, review, and summarize documents. If AI can help, and you're not using it, you're overprocessing. 6. Overproduction (Doing work before it is needed)   → With the speed and effectiveness of generative AI, you can do the work when it's needed and not too far in advance. 7. Defects (Mistakes and errors that need to be reworked)   → Generative AI can find mistakes and errors and offer solutions. 8. Skills (Not using workers to their fullest abilities)   → Generative AI can free up employees from routine or menial tasks, allowing them to focus on high-value work that leverages their full potential and expertise. 8 Wastes for Knowledge Workers (HIVEMIND): 1. Haphazard (Lacking strategy and reacting to requests)   → Generative AI can help maintain focus by assisting with structured planning, setting priorities, and offering strategic guidance, reducing reactive decision-making. 2. Interruptions (Stopping or slowing the progress of work)   → Drafting emails and communication in ChatGPT can help keep you away from tools that have constant notifications. 3. Veer Contexts (Switching between work activities)   → Tools like ChatGPT can write, review, edit, summarize, analyze data, create images, and more, so you don't need to go anywhere else. 4. Exploring (Searching for information)   → Generative AI can quickly retrieve information from a variety of sources or documents, eliminating the need for manual searching and research. 5. Multi-tasking (Splitting attention between activities)   → AI tools like Fireflies.AI can take meeting notes so you can focus on what's being said in the meeting. (see comments for the rest of the list...)

Excited to share my insights on using AI effectively without sharing sensitive information a month from today! I've also been working closely with Fred Allen and Mbongeni Ndlovu the last few months on an AI microcredential course... can't wait to see what they've got in store for the Impact of AI panel!

Why aren't organizations seeing productivity gains with AI when employees are? Inspired by Ethan Mollick's latest article (link in the comments), I’ve refined my mental model on AI deployment within organizations. There are three key groups in an organization: 1. The Leaders These are the people with the power to allocate time and resources, from supervisors and managers to the CEO. They provide: → Empowerment and Motivation: They encourage employees to turn to AI first to solve problems and help clear roadblocks. → Psychological Safety: They foster a culture where innovation, failure, and sharing are encouraged, ensuring no layoffs due to AI advancements. → Time and Resources: They allow time for experimentation with AI, and provide access to training, internal products, and capital. 2. The Crowd This is everyone in the organization, each with the ability to use AI to improve their work. They contribute: → Pain Points and Challenges: Less tech-savvy employees share the problems and challenges they face. → Ideas and Insights: More tech-savvy employees offer ideas on how AI can assist and share insights into its impact. → Experiments and Results: Employees using AI share their results across the organization, a concept known as "Yokoten" (sharing laterally) in Lean. 3. The Lab This team is passionate about AI and includes subject matter experts, technologists, and non-technologists. They deliver: → Prompts and Products: Based on feedback from The Crowd, they create a library of helpful prompts and deploy products like GPTs, Claude’s Projects, or Google’s Gems. → Guidelines and Benchmarks: They evaluate AI models for business-critical tasks, develop legal and ethical guidelines, and set guardrails. → Training and Material: They produce how-to videos, SOPs, or materials (like custom GPTs) to share knowledge and standards across the organization. If any of these groups lacks structure or clarity in their roles, AI improvements will fall flat. P.S. I have openings in my calendar next week before a busy month of keynotes, training sessions, and projects. Send me a message if you'd like to discuss how this model could work for your organization! P.P.S. There's something about writing ideas out with pen and paper that cannot (currently) be replicated by technology. Agree?

Note to self: Entrepreneurship is a long game (a very long game). 10 years ago I signed up for the Startup 100 program. The goal of the program was simple: Accept 100 participants across the province to create 100 new ventures in 100 days. It had nothing to do with consulting, training, coaching, operational excellence, or artificial intelligence. But it helped solidify my identity as an entrepreneur. It was another vote for the type of person I wanted to become. I didn't create a new venture in 100 days, but I did create one in 100 months! P.S. I stumbled across this as I was cleaning my home office and completely forgot about it. It might have to go in a frame and on the wall now! Jared Perry

Advanced voice mode is here for paying ChatGPT users! I’m dropping my car off for an appointment this morning and then walking to meet Jared Perry and others for coffee. Plenty of time to give this a spin! For those without the paid version, is there anything you're curious about regarding advanced voice mode?

Our AI microcredential course for 90 Nova Scotians launches next week—want a sneak peek? My favorite model we're sharing across all three streams is the AI-First Flywheel. The concept of a flywheel is simple: each step follows naturally from the last, and as you go through the steps, momentum builds. It's slow and challenging at first, but the more you repeat the process, the more effortless it becomes, and the better the results. Here's how the AI-First Flywheel works: 1. Identify Your AI Use Cases Two elements here: understand what AI can do, but more importantly, know what you do. Where those overlap is where you should start. 2. Select Top AI Use Case  This will differ for everyone, but I recommend choosing the use case that solves the most personal pain and is easiest to implement and maintain. 3. Pilot New Tool and Workflow Once the use case is selected, pilot it. Treat it as a test, so adjustments can be made as you see how it works in practice. 4. Standardize New AI Workflow After the pilot, apply the lessons learned to standardize the workflow. Ensure it's sustainable and can be rolled out across teams or departments.  "An improvement isn't complete until it is shared." — Ryan Tierney 5. Measure Benefits and Inspire Others Once standardized, measure and share the benefits. People will be inspired when they see the results and how AI can save time or improve work quality. 6. Create More Time to Play with AI The time saved from AI should feed back into the flywheel, allowing you to explore new use cases. I use the word "play" intentionally—once AI feels like a chore, interest will fade. There’s more nuance in the course, but this model is designed to be used by anyone, anywhere, anytime. I’d love to hear your thoughts and feedback!

I've upgraded my workspace to a closet! 👀 My previous workspace was cluttered and uninspiring. The backdrop was an empty whiteboard, the lighting was poor, and the side wall featured a whiteboard that had fallen off, with scuffed paint covered by sticky notes. The main reason for the shift was that I wasn't getting enough deep work done. I'm not blaming the workspace—it's entirely my fault—but sometimes a new environment can spark better habits. This new workspace is literally a closet, and I love it. It fits my electronic standing desk perfectly; it wouldn’t even fit through the doorway, so I had to take it apart and reassemble it inside the closet. It has a built-in shelf for everything I need within reach and is free from distractions. The biggest downside is that I have to crawl under the desk to get to my computer. In a way, I love it. The crawl helps separate deep, focused work from distractions. Once I crawl in, I'm locked in. Next on my list: → Adding some plants to brighten it up a bit → Tidying up the cords → Adding some pictures/posters to the walls → Soundproofing (since it sounds like I'm in a closet). I'm also considering replacing the door with a bookshelf for a cool, secret entrance. I'm open to all sorts of ideas! Let me know in the comments what you recommend.

What do 160 women from Women's Institutes across Canada want to know about AI? Lots! But first, do you remember the first time you used ChatGPT? Or the first time it saved you hours of work? For me, it was a project where I needed to create a table of all the middle schools, junior highs, and high schools in the province and match them with their county. The list didn’t exist (or if it did, my client didn’t have it, and I couldn’t find it on Google). I asked ChatGPT, and it created the table in seconds (with a few hallucinations). That was the moment I knew it would change the way I work forever. Fast forward 20 months, and I got to experience that feeling again. Most of the women in the room had never heard of ChatGPT before. Less than 10 had actually used it. I demoed the basic functionality available for free, and I got cheers, applause, and gasps. Here are the use cases that interested them most based on their responses and questions: → Writing a newsletter page from a couple of sentences → Writing emails → Creating images for posters → Creating images for the church bulletin (this was going to save one woman hours every week) → Using text-to-speech and voice prompts → Summarizing documents (they're manually going through a lot of archive material, which takes time) → Asking questions about information in a document → Analyzing, summarizing, and visualizing data (without ever having to open an Excel worksheet again) → Taking screenshots and pictures with their phones to troubleshoot technology → Possibly using it to help with the NY Times Connections puzzle* Takeaway: If you haven’t introduced an AI tool like ChatGPT to family members yet, do it! You’d be surprised how much it can help and the creative use cases they come up with. Let me know in the comments how your friends/family are using AI to make their lives easier. P.S. Thanks to Mila MacLean P. Eng for the introduction! It was a pleasure speaking at the 23rd National Federated Women’s Institutes of Canada Convention yesterday morning. And special thanks to Jaysinh Rayjada for joining me on the trip and capturing great pictures and videos! P.P.S. *This wasn’t confirmed, but a lot of eyes lit up when I demoed o1’s functionality to solve puzzles.