ASK Tom - learning AI tools for work
A question from an operations manager in construction excited about his AI journey
ASK TOM:
Q: As an operations manager in construction, I'm genuinely excited about my AI journey and the massive potential I see for improving project tracking and management. I'm currently exploring no-code automation platforms like Zapier and Make, which offer immediate practical applications. To deepen my understanding, I've also enrolled in a programming fundamentals course. While no-code solutions are valuable, I want to understand the coded alternatives to make informed decisions about our tech stack.
One challenge I've encountered is filtering through social media content about AI. There's an overwhelming amount of information out there, and it can be difficult to distinguish between genuine, practical insights and content designed purely for growing followers. I'd love to hear how others navigate this and find reliable resources for learning about AI in business applications.
—
A: Thank you for the thoughtfully written question and your commitment to deepening your career expertise. Excitement and curiosity are key!
I'm going to break this down into two parts and attempt to give you some guidelines on how I would approach these two topics.
Kudos to you for dedicating time to a programming fundamentals course. I assume this is an introduction to Python. Once you know how to load up a simple editor and copy-paste code for execution, you'll be on your way to deepening your understanding of how to write simple but useful apps and string various other AI tools together for various AI automation purposes.
I've noticed that many beginners are using Cursor.com to write software. This AI-powered code editor is a great tool for learning, as it provides code suggestions, explanations, and even generates code based on natural language prompts. It essentially acts as a coding tutor, guiding users as they develop their skills. Recommended.
These days, with AI tools like ChatGPT and Gemini, you can get pretty darn quick solutions to solving AI automation problems but: I might add you will benefit from making a human contact for programming as you learn.
Consider connecting with other programmers in your journey – they often enjoy mentoring because it deepens their own understanding. Having an experienced developer to turn to can be invaluable, especially since AI-generated code isn't always perfect. A mentor can help you develop crucial skills like debugging and understanding best practices that go beyond just getting code to run.
While AI tools are fantastic for learning, particularly for getting explanations and troubleshooting help, they work best when combined with human expertise and real-world programming experience. This balanced approach will help you build a stronger foundation in programming.
Bottom line: Start small. Make something work, even if it's just a test or throwaway project for a code or no-code mini-project. Add complexity as time permits. I typically just start with something personal that I want that has no bearing on anyone else's needs. The typical student projects rarely interest me. That way, I stay engaged in learning the tool. In fact, I'm doing that now with n8n.io, which I'll mention again below.
Exploring no-code automation tools
The first two tools you mentioned are a good start — Zapier and Make. I've tested them some in the past but have not landed on either for my automation needs. They can be fun for testing and running simple automations and complex ones if you decide to go deeper.
My advice is to tinker with these and feel okay with dropping the tool if it's just not working for you. Each automation tool has a different user interface and training philosophy. I've found I'll get frustrated mostly because of the way they explain things. The tool might be fine, but I'm just not connecting with it. So, I move on.
I've been testing n8n.io because there is a free version for Mac, Windows, or Linux for local hosting. It is a popular workflow automation platform for individuals or technical teams. This tool gives me full control over testing automation ideas. I can install n8n on a local machine. This enables me to start and stop automations while testing and learning. And the user support for the tool is amazing with hundreds of automations freely shared. There is a paid version you may prefer that hosts n8n.io for you allowing you to start using it immediately. Try it for a month or two and turn it it’s not your thing. That’s how I approach testing new tools assuming they offer a month to month.
Note that local installation n8n.io does require some effort. As with all of these tools, use YouTube to find people you understand to guide you through initial installation and use. YouTubers are like cooks; everyone knows how to make hamburgers, but some people you like better than others. Skip the ones that overhype stuff. BTW, I pay for the ad-free version of YouTube so my train of thought is not destroyed every time I'm attempting to learn something new. Highly recommended.
These efforts will help you understand alternatives to make informed decisions about your tech stack. My personal history is in IT management, so keep in mind you may need to gain some cooperation or agreement from your IT folks as you start endeavoring more deeply on automation projects. It is possible they will be interested in learning what you learn too.
Regarding your challenge in filtering through social media content about AI
To begin, I share your frustration of weeding through content designed purely for growing followers, sigh. Given you're an expert in your field, here is what I suggest starting with to find authors and newsletters that resonate with your needs.
Go to
and consider using a prompt like this:
"Which specialized construction tech journals have a focus on AI topics. Before you answer, ask me questions to help you refine your search."
To expand on this prompt for refinement, look at this tip from Mr. Viral.
This prompt will help Gemini know better how best to answer you versus providing something generic which can be hit or miss. I highly recommend doing this.
Here is what the results look like:
There is a new tool for searching Subreddits. It’s at:
GummySearch offers a free plan that allows for a limited number of 50 keyword searches. You do not need to put a credit card in to use it.
For testing purposes, I used ‘construction tech’ to see what came up. Your results will vary, and as I mentioned in step 1, the results you find may suggest ideas for better searching here.
I would look at writers on Substack. I’m constantly amazed at the amount of great content there. There are no pop ups or ads, just authors writing great content. Much of the content is free. Take what you’re learning in the way of good keywords and phrases from the previous two steps and experiment with searches at Substack.com.
Thank you again for taking the time to ask this great question. Stay in contact as you continue on your journey of exploring AI.
Have something you want to ASK TOM - put your thoughts here. You can do it anonymously or as others have done, leave your name and email optionally.
tp
This is a great post. I second cursor it’s really great as a copilot to learn how to code and improve existing code. You can ask it to write code for you. But a better use is to ask it to explain code you don’t understand or to suggest ways to improve any code you have written.
I’ve found Claude to be the best at coding but other people prefer other ai’s
Thanks for taking the time to answer my question Tom! This is very helpful, I will definitely look into this recommendations.