"You will never see the intelligence if you never try to use the tools!" - Futurist Jim Carroll
I've long believed that you can't talk about something if you don't do that something.
I certainly can't talk about climbing Mt. Everest because I never have. My good friend Scott Kress has, and so he talks about building an effective team to accomplish such a goal. He's pretty good at it.
Which brings me to skiing and AI.
One thing I've learned by becoming a skier at the age of 40 is that you can learn a lot from people on the chair ride open - most people are friendly and eager to talk. Conversations during the minutes on the ride up can cover anything and everything, from family to career, politics to sports - and of course, snow conditions.
I've found that I can learn a lot - particularly from younger generations - about important trends and personal motivations. Many years ago, I found that they don't talk about what they do for a living - they talk about what they like to do., I worked that into other observations I was seeing at the time, and turned it into a full workforce trend,
For some reason, I ended up spending some time on the 'ride up' with several young adults who had carved out a career in IT, and our conversations turned to AI. Among the many observations I learned was this key one - most are very cognizant that artificial intelligence is rapidly changing the very nature of software coding, and while they aren't concerned about losing their jobs to an AI, they sure as heck are spending time learning how to use it to their advantage.
In other words, they are thinking that if they are busy writing computer code, they might as well use AI as an ally to help them do so, to be more efficient, master more complex tasks, and further enhance their skills. In that vein, they are learning how to find the intelligence in the AI by actively using the tools.
That's a wildly important concept that everyone should follow.
With this idea firmly planted in my mind, in addition to skiing writing a book, and doing various holiday things over. the break, I've also spent a lot of time working with AI coding tools to find its intelligence. And, it's, wow ...
Let's start here. While I'm out there, I use an app on my Apple Watch called SkiTracks - it grabs a ridiculous amount of data about each and every run - speed, distance, time, etc. The raw data looks like this - basically, a whole bunch of GPS data that summarizes what is going on every second.
At the end of the day, it takes all that data and turns it into something like this:
It's just a fun way to see your day - lots of us share this summary information through Strava, achieving some entirely inconsequential bragging rights.
But my conversations got me thinking - what if I could use an AI to write some computer code that would take all 12 days' worth of ski data, and give me an overall summary of all my skiing activity over the break? To do that, I turned to Claude.AI, which is drawing a lot of attention for being one of the best tools to write computer code.
I first set out to see if it could read the raw ski data as seen above - it could. Large language models are proving to be very powerful tools to analyze and summarize raw data like this - which has powerful implications in such fields as pharmaceutical clinical trials, with analysis of vast research data being the target.
In a matter of seconds, it was building information summaries similar to my app.
From that point came a lot of back and forth as I had it proceed to write me a program that I could run on my Mac that would take ALL of my data from the 12 days, and give me some summary information. After about 20 minutes - and remember, I'm learning as I go - I had the summary I was looking for.
The thing is, I would never know how to write the code to do this type of thing unless I had taken a course and spent countless hours learning the programming logic, That's not in my skill set - but learning how to use an AI to write code on my behalf now is. (Two observations - that's a lot of skiing! Second, I think the Top Speed might be a data error, and am going back to investigate that!)
This brings me to my Random Quote Generator. It's over at https://random.jimcarroll.com/. Without getting into a lot of details, I thought it might be interesting to have Claude.AI write two computer programs that would a) extract, from my thousands of Daily Inspiration posts, the key quote and b) build a Web site that randomly grabs one of these quotes and builds a big, bold Website with the quote.
Visit it and give it a try!
Behind the scenes, there is some pretty complex stuff.
As they say, this is pretty slick stuff.
The point of all this is that we are clearly at an inflection point with some aspects of AI. While the debate rages on as to whether AI is a friend or a foe, accurate or inaccurate, some powerful use cases are evolving. My mindset is, that if you don't take the time to find the actual 'intelligence' in artificial intelligence, you'll never really gain a deeper understanding of what is going on.
You've got to use the tools to understand the tools.
So - get going! Take on your own small, inconsequential projects, simply to explore, learn and move forward!
Futurist Jim Carroll has learned most of what he knows about technology simply by hacking around. His long-term goal is to get a humanoid robot so that he can learn how to program it. His wife will probably not be amused by this idea.