I received an email from a student recently who decided to cancel his membership, and he took the time to explain why. He was very respectful about it, and I genuinely appreciate where he's coming from. The short version is that he's been using AI more and more to help him with his Microsoft Access database. Instead of spending time searching for lessons, learning a technique, and then writing the code himself, he can explain what he wants to an AI and often have a working solution in a few minutes.
That raises a very reasonable question: if AI can help you get this stuff done, why should you keep paying for tutorials? My answer is that maybe you shouldn't. It depends entirely on what your goal is.
I'm here to teach people who want to learn how this stuff works. I teach Microsoft Access, database design, SQL, VBA, programming, and all the little nitty-gritty details that go along with them. Some of you want to become database developers. Some of you want to become programmers. Some of you are already developers and just enjoy learning new techniques. Some of you are retired but nerdy like me and just love learning this kind of stuff. You're the people I'm making these classes for.
But maybe you're an accountant who just wants a database to keep track of your clients. You don't care how an INNER JOIN works. You don't particularly want to learn VBA. You couldn't care less about normalization, recordsets, or the difference between a Long Integer and a Double. You just want the damn database to work so you can get back to accounting. There's absolutely nothing wrong with that. If you can explain what you need to an AI and it builds something that works for you, fantastic. Use it.
I look at AI a lot like the autopilot in an airplane. Modern autopilot systems are amazing. They can handle an enormous amount of the actual flying, and modern aircraft can even perform automated landings under the right conditions. If your goal is simply to get from New York to Los Angeles, you don't need to understand aerodynamics, operate the control surfaces, or know what to do when an engine quits. You just want to get there.
But if your goal is to become a pilot, that's completely different. You want to understand why the airplane flies. You want to know what all those instruments mean and what happens when something goes wrong. If the autopilot suddenly stops working, you'd better know how to fly the airplane yourself. That's what my classes are about. I'm not here just to give you a finished database. I'm here to teach you how to build one.
Another good example is foreign languages. I have zero desire to learn another spoken language. One of the things I told myself when I moved to Florida 13 years ago was that I should learn Spanish because there are a lot of Spanish-speaking people around here. Turns out, as time goes on, I realized I have no desire to really learn Spanish. I communicate perfectly well in English. I've been to Spain. I've been to Greece. I've traveled to places where English isn't the native language, and if I need to communicate with someone and we don't share a language, I've got Google Translate. It might not make me sound like a native speaker, but it'll usually get the job done.
That's exactly what I want from it because I don't want to become fluent in Spanish. I want to order dinner, ask where the bathroom is, or figure out what the guy behind the counter is trying to tell me. But if learning Spanish were something I enjoyed, or if I wanted to live in Spain and really understand the language, then Google Translate wouldn't replace learning Spanish. It would become another tool I could use while learning it. That's how I see AI and programming.
And I want to be very clear about this because I don't want anyone thinking I'm trashing AI. Quite the opposite. I use AI all the time, and it's become an extremely useful tool in my daily work, especially when I'm working with technologies that aren't my particular area of expertise.
Just recently, for example, I worked with AI to build an MCP interface for my website. I knew very little about MCP going into it. AI helped me understand what I needed and helped me put the pieces together. But I already understand Visual Basic. I understand ASP, databases, and how web services work. So when AI gives me a chunk of code, I'm not blindly copying and pasting some mysterious incantation and hoping the computer gods are pleased with my offering. I can look at it and understand what it's doing, modify it, and integrate it into my existing system. When something doesn't work, I have enough underlying knowledge to figure out why.
That's incredibly important because AI still gets things wrong. I've had AI give me code for Microsoft Access that simply doesn't work. Sometimes it mixes up VBA with another flavor of Visual Basic. Sometimes it suggests a method that Access doesn't support. Sometimes it invents a property or command that sounds perfectly reasonable but doesn't actually exist. And the fun part is that it'll often tell you this nonsense with tremendous confidence.
I know when that happens because I've been working with Access for three decades. I can look at the answer and say, "Nope. Access doesn't work that way." If you don't know Access yourself, you might not know the difference. You might spend the next hour trying to figure out why perfectly reasonable-looking code doesn't work when the real answer is that the AI made something up.
That's one of the biggest reasons I still think learning the fundamentals matters, even if you plan to use AI for most of your actual work. The better you understand the subject, the better you are at using AI. You know what questions to ask, you can give it better instructions, and you can recognize when it's heading down the wrong path. You can take 80 percent of a solution that it gives you, throw away the 20 percent that's garbage, and adapt the rest to what you're actually doing.
AI can make an experienced developer dramatically more productive, and it can certainly make a beginner more productive too. But there's a danger in relying on it completely when you don't understand what it's building for you. Eventually you're going to reach a point where something breaks, the AI can't figure it out, and you don't even know where to start looking. That's the difference between having a solution and having knowledge.
And there's nothing wrong with only wanting the solution. That's the point I really want to emphasize. Not everybody who uses Microsoft Access wants to become a database developer, and not everybody who writes a little VBA wants to become a programmer. Maybe Access is simply the tool that runs part of your business. Maybe you need a customer database, an inventory system, an invoicing database, or something to keep your office organized. If AI can help you build that without spending hundreds of hours learning database development, great!
Use the tools that make sense for what you're trying to accomplish. Let the autopilot fly the plane. Let Google Translate order dinner. Let AI write some SQL. But understand that those tools are solving a different problem from education.
If you're the kind of person who looks at a database and wonders why it works, if you enjoy figuring out how tables relate to each other, if you want to understand SQL instead of just having AI write it for you, or if you actually enjoy programming and want to know what's happening underneath the hood, that's what I'm here to teach you.
AI doesn't make learning this stuff obsolete. In a lot of ways, I think it makes that knowledge even more valuable because now you've got an incredibly powerful assistant sitting next to you. The more you know, the more useful that assistant becomes, and the better equipped you are to recognize when it has absolutely no idea what it's talking about.
So the question isn't really whether AI can build your database for you. Increasingly, the answer to that is going to be yes. The better question is whether you just want the database or whether you want to know how to build it. Both answers are perfectly valid, but they're two very different goals.
My classes are for the people who want to learn how to fly the plane.
I want to learn how to build the plain and fly it. :)
Sam Domino
@Reply 11 hours ago
Kevin Yes...Yes it is! Let's just hope Richard doesn't have to "reinitialize" it. LMAO!!!
As for AI, every new technological leap has us asking these same questions. H.G. Well's "The Time Machine", in part, asked if automation/industrialization would cause people to forget how to take care of themselves. Asimov's "The Feeling of Power" questioned if computers would lead to people not knowing/forgetting Math. In the end, we'll continue building better "tools" because they allow us to pursue whatever we dream of doing.
I'm glad someone here is old enough to get that Airplane! reference. Yes, that's the autopilot. LOL. I knew at least one of you guys would catch it.
And Sam, exactly. There are lots of science fiction stories about people forgetting how to use the technology that runs their lives. Star Trek did this in many episodes too.
Jeffrey Kraft
@Reply 11 hours ago
Using AI is like using a calculator in 3rd grade. It's better to understand how the answer is arrived at rather than just dumping what the technology told them. There will be days you don't have the calculator or fat fingered the numbers in and flat out accept the results because you don't understand the mechanics (rules and really blatently obvious when it's PEMDAS related) behind it.
I have gone to AI with some stuff because I can't recall or don't know after doing Leonard Nemoy's "In Search Of", and AI steared the ship into the wrong quadrant.
Jerry Fowler
@Reply 10 hours ago
Being retired and mostly building Access projects for myself, I don't need to know how to build the plane, I just want to know how to fly it. It is also useful to know what to do in an emergenCcy, using AI doesn't give you that. jJst a technical note, I believe the Autopilot should be in the right seat, the Capt. uses the left seat
Donald Lader
@Reply 7 hours ago
I have fully retired now and this gives me time to experiment. I am taking me working database and moving it to SQL server. That is not all. I am interfacing with AI for all external work such as reading and writing emails, running the client document system, translating several languages, completing forms, interfacing with the online court system, interfacing with the online client portal, and more. To make this interesting, I am not going to write any code myself. I am going to give the AI Agent a go at it first. This should be fun to watch. However, I will be reviewing the code and providing input after the initial try by AI. The lessons I learned here allow me to do just that. A big thank you Richard for getting me to tis place.
Kevin Yip
@Reply 6 hours ago
Many people can be self-taught, even before the age of AI. Maybe he already has a foundation on coding. When you already know, say, Python, it's not too difficult to teach yourself Javascript, and vice versa. Using AI is not exactly a gimme, as AI can be faulty, and you need to able to detect BS from AI. Also, there are plenty of books and online info for someone to self-teach.
Kevin Yeah, absolutely. Self-teaching has always been an option. Hell, that's how I learned most of what I know. When I was a kid with my TRS-80 Color Computer, if I wanted the computer to do something, I had a manual, some magazines, and a whole lot of trial and error. There was no YouTube, no Stack Overflow, and certainly no AI to ask why my 12 lines of BASIC weren't working. LOL.
I think the big difference now is speed. Someone who already has a decent foundation can use AI to dramatically shorten the time between "I don't know how to do this" and "okay, now I've got it." That's an amazing tool to have. But like you said, you still need enough knowledge to recognize when it's confidently feeding you BS. My Color Computer never hallucinated an answer for me. It just sat there and gave me a SYNTAX ERROR. :)
And really, technology has always been about speeding up the process of getting knowledge into your noodle. I'm sure the first guy who showed up with an abacus had everyone around him going, "Holy crap, look how fast that guy can count." Then came mechanical calculators, electronic calculators, computers, and eventually tools that made finding the information itself faster.
There was a time when having a set of encyclopedias at home saved you a trip to the library. Then Google let you find information in seconds. YouTube let somebody actually show you how to do something instead of making you read about it. Now AI can take all of that another step further and digest the information for you. But every one of those tools still depends on the quality of what's going into your brain. Books can be written by people who don't know what they're talking about - and yes, I've read a few. AI is no different. The faster we get at acquiring information, the more important it becomes to know enough to recognize when the information we're getting is crap.
And the same thing applies to TikTok, social media, and the rest of the Internet. These days you need a pretty well-calibrated bullshit detector. There's always some viral video telling you the lines behind airplanes are chemtrails, seed oils are going to give you Alzheimer's, or whatever nonsense is making the rounds this week. If you've got a decent foundational understanding of science, you're much better equipped to stop and say, "Wait a minute, that doesn't sound right." AI is no different. When you're using it for coding, you need enough knowledge to recognize when it's helping you solve the problem and when it's confidently leading you down the wrong rabbit hole.
Sorry, only students may add comments.
Click here for more
information on how you can set up an account.
If you are a Visitor, go ahead and post your reply as a
new comment, and we'll move it here for you
once it's approved. Be sure to use the same name and email address.