Learn AI With Me, A 30-Day Sprint
I'm an AI Skeptic. I'm starting a 30-day sprint to get smart quick. Join me!
It’s safe to say that I am not an early adopter of AI. I have to admit that I'm pretty resistant, actually. Count me in the ranks of the skeptics. That said, in the last few months it’s become increasingly clear that I can’t afford to ignore AI. Instead, I need to dive in and build a more informed opinion.
I’ll get into my skepticism in a another post. For now though, let me just describe my project, and invite you to join me on this journey.
Learn AI With Me: A 30-day sprint to get smart quick.
For the next 30 days, my plan is to spend one hour a day exploring AI and sharing my what I learn with you. (Not incidentally, I’ll be sharing my ignorance too, so please bear with me.)
I’m planning to give myself a crash course in this current generation of emerging AI tools. As I said, I’m pretty close to being a total novice at all of this. I’m an experienced designer, product person, and author—I’ve been doing this work for 35 years!—but as happens when you gain seniority, you lose touch with emerging tools.
I imagine that this series might attract three types of readers:
You might be an AI novice too. If you are, I invite you to learn along with me and share what you’re learning too.
You might be a recent adopter—if so, I welcome your advice, especially about your learning journey.
You might be an expert—in which case, you should probably be paying attention to the struggles of the novices. These are the struggles that can illuminate your path forward—and hopefully help you make better AI products.
My Plan
I’m planning on posting every day—either a full-fledged Substack post, or perhaps a smaller observation shared via Notes or in Chat. (I don’t want to overwhelm your inbox.) I hope you’ll join me, and please do share this with anyone you think is interested. I’ve enabled subscriber chat, and I’d love to hear from you there.
Why I’ve been resistant to AI
I thought it might be useful in this first post to be transparent about my skepticism. My plan is to check back in on this from time to time, and see if my feelings are changing as I learn.
I wrote above that I’ve been resistant to AI. Part of the reason is that my early explorations with ChatGPT yielded such unremarkable results. I mean, it was remarkable that AI functioned at all. It was remarkable to see a computer generate such plausible text. What was unremarkable was the quality of the result. The language was dead and lifeless. The content was uninspired. I couldn’t think of a single thing that it was doing that I found valuable or satisfying. I guess that you could call this my quality objection.
Of course, there are some substantial downsides to AI that have little to do with my aesthetic reservations. Objectively AI uses a vast amount of power to generate these unsatisfying results. So we’re burning up the planet and filling it with low-quality work all at the same time. Let’s call this the environmental objection.
Finally, there are countless ethical objections to deal with. The results aren’t just unremarkable by aesthetic standards. They are causing harm right now, and there is the potential for even more significant harm to follow. LLMs amplify biases represented in the data they are trained on, and given that they are trained on the Internet, a data set that is filled with biases, it’s hard to escape the conclusion that you’re just gonna get biased results
An Old Dog Learns New Tricks
In addition to the three objections above, there’s a personal resistance to talk about.
In the 90s, when I began my career in tech, I was a young man and eager to learn new things. One source of this feeling was that at that point in my life, I hardly knew anything, so everything felt new and wonderful. Nothing I was learning threatened the way I did things before.
Now I am most decidedly not a young man, and I’ve learned how to do a whole bunch of things—some of them pretty well, thank you very much. There’s a natural hesitation at this stage of life to learn new things and to learn to do old things in new ways. Inevitably doing things in new ways is just gonna make you worse at those things, for a little while anyway. It sucks to feel bad at things that you used to feel good at.
I don’t think this is a useful set of feelings, so I resist them. Hence this project. But I suspect that there are lots of folks out there like me—people with hard-won skills, expertise, and seniority who feel similar hesitation about AI. If you’re reading this—and you’ve made it this far, I encourage you to join me.
Oh and if you’re a young expert, share what you know. And better yet, pay attention to the folks out here struggling. Those struggles have the ability to guide your path, and if you think about them the right way, will create opportunities for you to create products, services, and solutions that could unlock your future.
OK, enough. Join me, and I’ll see you tomorrow.
"It sucks to feel bad at things that you used to feel good at."
Sometimes it’s good to start things from scratch, because you face the steepest learning curve at the beginning of everything. Mastery requires not only time and repetition, but also the ability to handle local minimums that make you question whether you should keep going on in your old behavior, pivot to a new (in this case your AI journey) or stop.
One thing I’ve learned is how powerful context can be.
One thing you might find useful is letting AI know your strategic context and your goals / hypothesis ( I essentially feed it with the relevant lean strategy canvas and product canvas ). In ChatGPT for example the project instructions is a great way to do that. But you can also just add it to the prompt or attach a file.