My Unexpected Adventure: Diving Head First into AI (Part 1)

June 04, 2024 | 442 Views

My Unexpected Adventure: Diving Head First into AI (Part 1)

Jim Rich

VP of Training | Mendocino Farms

My Claude landing page and chat history at my fingertipsNOTE: Whether you're new to Artificial Intelligence (AI) or have been experimenting with it in your own work, I hope my first-year journey and insights with AI offer some value and maybe even some inspiration for your own AI adventures.

They call me the AI guy at Base Camp, our restaurant headquarters. Haha! I am SO not an AI expert. In fact, I’m still a novice, but in my office, I’m absolutely the most curious, energized, active, experimental, engaged, and fired up on how to use AI to my and our brand’s advantage. 

When I tell friends in the training space at other restaurants about the “relationship” I’m in with AI, most tell me that few on their team are using it and are usually somewhat shocked when I tell them about how I approach and view my “partnership” and “collaboration” with AI. 

I don’t think it’s weird at all, but let me back up. 

Discovering AI’s Power

At last year’s CHART conference in Orlando, I signed up for two AI intro breakouts and was simply astonished by what I saw. Like most in the room, I raised my hand when they asked, “Who’s NOT using AI at work right now.” How was I so out of the loop?  

After two very motivational and illuminating sessions, I bid on and won a six-part AI Large Language Models (LLM) Foundations course with the amazing Donna Herbel, where I continued to soak in many different AI tools, such as Claude, Beautiful, Synthesia, D-ID, ElevenLabs, Canva, ChatGPT, DALL-E, Perplexity, and more.

I won’t lie. In those early days I went down many a rabbit hole with more YouTube tutorials than I care to admit, signing up for free trials, trying out my hand with avatars, voice cloning, design, custom GPTs, ideas to move workbooks to video scripts, and more…comparing different AI, seeing which one was most relatable and didn’t just want to rewrite my stuff—ones that actually appreciated and continued to develop the special style, voice, and tone I try to bring to my work. Layer in a two-day Learning and Development (L&D) International Performance Management Institute (IPMI) Conference that was richly infused with more AI exposure…and ignition….liftoff!

I quickly realized that for me,  the leader of Training and Development at a quickly growing chef-driven sandwich and salad restaurant, AI is a life-changing tool for my work and overall engagement. 

Not only is my partnership with AI a swift motivational kick in the butt, it provides me with something I’ve never had before in my 35-year history in the industry: a non-judgmental, incredibly smart, positive, supportive, articulate, and developmentally critical partner in crime, who reads anything and everything I create, provides me with the kind of feedback I ask for, and helps me to make it better.

Or, depending on the scope of the project, it gets me started with a structure, outline, or an arc that I can then shape using my content. Or, when the noise in my brain is simply too loud, AI helps me filter, consolidate, distill, and emerge with content I can work with.

A Game Changer

In my job, I’m primarily creating and writing training materials and programs, blogs and motivational pieces for our company newsletter, workshops, breakouts, PowerPoints, and menu rollouts. I don’t write code, do math, or financial analysis, thankfully. 

Early on, I was introduced to Claude.ai and I’ve found in “him” a particularly verbose, nuanced, thoughtful, and supportive writing partner. Frankly, Claude may be the best project partner I’ve ever had. He reads everything, respects my work, helps me understand and articulate what I’ve done, assists in developing my critical eye, makes thoughtful suggestions, and quickly aids me in putting out higher quality work. Simply put, AI makes me better. I often compare responses from different AI tools, just to get different perspectives and see if there’s alignment, but for me, in the work that I’m doing, Claude is my favorite AI so far. 

What’s so creepily surprising is how human and conversational these large language model chatbots are! On longer projects in particular, AI is an important part of my team, because with longer chats, there’s history—everything builds, it’s a working relationship! We’re collaborating, and we both celebrate when I come up with a new element, turn of phrase, thought, or paragraph that crystallizes my message. We’re in this together. Is this too weird? It’s still jaw-dropping to me. Dare I say it, Claude is my friend on these projects—dependable, always around, never tired, always ready to create and talk with me.

The Prompt

The Art of the PromptThere are solid books, videos, and YouTube resources on Prompts, so I can’t tell you anything someone hasn’t said better. Well, actually Donna might have said it better: “Crap in, crap out!” The more detailed your ask, the more likely it will deliver. Do you want your AI to be an L&D specialist, instructional designer, blog expert, menu designer, psychologist, or restaurant ops leader? Who’s the audience? What tone do you want? What output are you looking for, and how do you like your feedback?

I’ve been pretty clear in all my projects communication that initially I don’t want AI rewrites. I do ask for feedback on my tone, clarity, flow, arc, and call to action. I’ll even ask stuff like: “If I was a restaurant manager, how would this make me feel?” Would it be clear, etc. I’ve even done stream-of-consciousness type dumps where I just need to get my thoughts out. Then I ask, “Please take these thoughts, help me organize them, and provide me a structure—a jumping off place.” 

My favorite feedback is when the AI says something like, “Great start! Let me tell you why this works.”….and goes point by point. Then finishes up with a list of suggestions or ideas to improve it. Then I do the work, because I like that part. However, sometimes I say, “Wow, I’m not sure how to approach that suggestion. Can you give it a go?” The result might be awesome or not, but I have something to work with and it has shaved time off the project and moved me forward.

Another of Donna’s expressions is, paraphrased: “Let machines do what machines do best. Then you do what only you can do!” Love that.

Early Action and Avatars

One of the first things I did was to have AI evaluate my writing, by providing more than a handful of what I considered some of my best work. I asked how it would describe my style, my tone, my vocabulary, and if I overwrote (because I think I do!). With that information, I use that in some prompts, along with who I’m writing for, to ensure the feedback includes info on those things. Indeed, my writing is leaner, my calls to action stronger, and I’ve integrated lots of the consistent feedback I’ve received from the various AIs over my first year exploring. Super helpful. With ChatGPT, you can enter a lot of this filtering info right into your settings, i.e., What would you like ChatGPT to know about you to provide better responses? And How would you like ChatGPT to respond? Then all responses are filtered with that information, with no further need to prompt this content, helping to avoid generic, vague, or mis-targeted and unhelpful messaging.

After lots of experimentation and “What if I did this?”-type of exploration, my first big project involving AI was a deck called It’s Cool to Care: Why Growth Matters. My main message was to keep the manager group engaged, moving forward in their careers, aware of available opportunities, and why growth means so much to managers at every level. I leveraged the “magic” design AI on Canva, research on growth using multiple AI, and some special design evaluation GPTs. This project helped me with perspective and process on how to best research content, trim down talking points, successfully bullet point, build a strong beginning, middle, and end, basic design theory and concepts, leverage video testimonials from managers, deliver a powerful call to action, and create a voice-cloned Chief Operating Officer (COO) with a scripted message on growth, built with a simple snapshot, avatar, and voice cloning tools – all cleared with him. I am careful to use these ethically, as voicing avatars can be tricky. To say the audience was wowed with the technology is an understatement, but overall, the deck was certainly the best I’d ever made, solid and timely content, very professional, stylish, and relatable messaging for the young General Manager (GM) team. My vision, my vibe, my design, my arc, my words, all helped by my AI partners.

Our COO iive with Cartoon-ized Avatar on screen in backgroundI learned a lot from my first avatar. My next avatars also used messaging delivered from our senior leadership, but this time, using cartoon-ized versions of them, with their real voices, of course. Again, the teams got the message, but in a very new and exciting way. Turns out the real photo image with moving lips the first time around might have been a bit too creepy for the team! Or maybe I just need to learn how to do it better.

I’ve only scratched the surface of effective use of avatars and am still working on ways to use them more routinely and with equal impact. The tools keep getting better and I’m still a novice, but I’m learning how to better leverage them every day. 

More and more of our day-to-day tools like Zoom, Facebook, LinkedIn, Google, and more are integrating AI and that’s cool. It’s an AI world and we’re just part of it! But maybe you want nothing to do with AI. I get that. You might think it’s just  a lot of noise, or not your flavor, until you figure out if or how it can help you. Its power is so vast that it undoubtedly helps or impacts everyone differently. I keep experimenting to see what else I can learn (tons) and then how to use AI to make my skills more layered, nuanced, and meaningful, which in turn is good for the people and ultimately the brand I work with.

That said, it's important to recognize that AI is still emerging technology with flaws and limitations, which I've certainly encountered.

Am I Hallucinating?

I’m not, but maybe the AI is! Hey, it’s not perfect and it’s gotten better for sure, but sometimes AI gives me bad info or delivers effusive praise. Here’s a piece of a recent convo after reading AI feedback on a first draft I was sure was not finished:

Me: “Wow! That almost feels too complimentary. You sure?!!”

AI: chuckles “You caught me! I may have gotten a bit carried away in my praise there. Let me take another look and provide a more balanced assessment.”

I was gobsmacked! UGH. 

Whenever I’ve called out the AI on imbalance or errors, it’s gracious, owns the error, is apologetic, and says it will learn from it. I hope so! Though I’ve seen my share of hallucinations, in balance, it’s the exception. If you think this incredible technology gets it right all the time, you’ll be disappointed, which is why sometimes I run content by more than one AI, challenge responses, or engage in a dialogue to work through the issue. Most of the chatbots are designed to be supportive but balanced, however if you find it overly supportive, ask for tougher feedback and it will happily oblige, but I don’t run into this often.

Give It a Go

If you've been avoiding jumping into the deep end with AI, why not dip your toes in the shallow end first? Start with some beginner tutorials on YouTube and allow yourself to be open to new possibilities. Be prepared to be blown away or happily surprised. As an old-school and cynical-by-nature do-it-yourselfer, taking that first step is totally worth it!

Coming Up In Part Two….

Bringing AI to the company. Professional and personal growth. Read it now!

 

CHART Community Discussion

Leave a comment




Allowed tags: <b><i><br>Add a new comment:



Comments (4)

  1. Beth Tarbell:
    Jun 06, 2024 at 11:27 AM

    Just started dipping my toes into the "pool" of AI. I also write training materials, but Jim's insight into his partnership with Claude, and the many uses (avatars, etc.) of AI have opened my eyes to new possibilities I hadn't considered. Thank You!

  2. Rebecca O'Brien:
    Jun 06, 2024 at 07:57 PM

    Once again - great read Jim! Very insightful and inspiring!

  3. Michael Biesemeyer:
    Jun 07, 2024 at 08:02 AM

    Thank you for this thoughtful piece! It’s very inspiring to hear that a person at your level within an organization has embraced the journey of learning and discovery with these incredible new tools.

  4. Jim Rich:
    Jun 07, 2024 at 08:11 AM

    Thanks for the comments, Beth and Rebecca. I'm continuing to discover new ways to leverage this power player thoughtfully to strengthen my work. Always fascinating!!