Once again I've stumbled across another company that's trying to move people out of Access and framing it as an Access problem. This time it's a blog post from a company called GAPVelocity AI, with a headline that reads Why ChatGPT Can't Fix Your Access Database.
Right out of the gate, you can see where this is going. The article walks through a few familiar talking points. It says Access databases are locked away in binary formats that AI can't fully understand. It talks about how AI can get you maybe 70% of the way through a migration before hitting a wall. And it frames Access as a "monolithic" system that has to be broken apart and rebuilt into a modern web architecture.
Now, on the surface, some of that sounds reasonable. Yes, migrating any complex application to a completely different platform is hard. Yes, AI is not magic and it does not perfectly reconstruct entire systems from partial information. But here's the part they don't emphasize enough. That problem is not unique to Access. That is a universal truth for software migration, whether you're coming from VB6, Java, PowerBuilder, or anything else. What they're really describing is the difficulty of doing a full architectural rewrite, not some fatal flaw in Access itself.
So who are these guys? GAPVelocity AI is part of a larger consulting firm called Growth Acceleration Partners. Their business is modernization. In plain English, they get paid to take "legacy" systems and move them into modern web and cloud platforms like .NET and Blazor. They offer a mix of automated tools and consulting services to help companies convert applications, including Microsoft Access, into something else. That is their product. That is how they make money.
And once you understand that, the tone of the article makes a lot more sense. It's not a neutral, unbiased technical discussion. It's a sales pitch. The message is essentially: your Access system is old, AI can't save you, and you need a specialized tool and a consulting engagement to get out of it. That doesn't mean their tools are useless. For large enterprises that are required to move everything to the web, those kinds of services can absolutely have value. But it does mean you have to read their claims through the lens of their business model.
What gets lost in all of this is what Access actually is and what it does well. Access is not just a "database." It is a rapid application development environment that lets you build complete business solutions quickly and efficiently. It has supported split front-end and back-end architectures for decades. It works beautifully as a front end to SQL Server. And for small to mid-sized businesses, and even many enterprise scenarios, it remains one of the fastest and most cost-effective ways to build and maintain internal applications.
And then there's the AI angle. The idea that ChatGPT "can't work with Access" is simply not true in practice. I use it every day. It writes SQL, it helps build and debug VBA, and it can assist with UI changes when you export form and report objects to text. If you know how to use the tools, AI becomes an enhancement to Access development, not a replacement for it. The limitation isn't the platform. It's whether the developer understands how to leverage it.
At the end of the day, articles like this tend to follow a familiar pattern. They highlight real challenges, frame them as problems with the platform, and then position a paid solution as the only way forward. But Access isn't broken. It doesn't need to be "rescued." In many cases, it just needs to be used correctly, maintained properly, and augmented with modern tools where appropriate. That's what a lot of us have been doing successfully for years.
LLAP
RR
P.S. Here's a copy of what I left in the comments form on that page. We'll see if they actually publish it:
This article mixes a valid point about AI-assisted migration with a very outdated and misleading characterization of Microsoft Access itself. Access isn't "just an old database" or a relic that needs to be replaced. It's a rapid application development platform that combines a relational database engine, UI framework, and business logic in a way that is incredibly effective for small and mid-sized business applications, and even in large enterprise solutions it's a great front end to an SQL Server solution. For many use cases, that tight integration is a feature, not a flaw. Calling it "monolithic" ignores the fact that Access has supported split front-end/back-end architectures, SQL Server backends, and scalable multi-user deployments for decades.
It's also simply incorrect to suggest that AI tools like ChatGPT can't meaningfully work with Access. I use ChatGPT with Access databases every day. It excels at writing and troubleshooting SQL, generating and refining VBA, and even assisting with UI changes. Access objects like forms and reports can be exported to text, modified, and re-imported, which makes them perfectly accessible to AI-driven workflows. If someone isn't aware that this is possible, they are not up to date on how modern Access development is actually being done.
The real issue described here isn't that Access is obsolete, it's that fully automated, one-click migrations to completely different architectures are inherently complex. That would be true for any platform, not just Access. Of course AI can't perfectly reconstruct an entire application from partial context. That's not a limitation of Access, it's a limitation of the migration approach. The conclusion that you need a proprietary paid tool to "fix" this problem says more about the author's knowledge than the technology.
And it's worth noting that criticism like this almost always comes from companies trying to sell a replacement platform. That perspective tends to ignore what Access is actually good at: rapid development, tight integration, and cost-effective solutions that work. Access doesn't need to be replaced in many cases. It needs to be understood, used correctly, and, where appropriate, enhanced with modern tools like AI.
That's exactly what many of us are already doing successfully every day.
No comments:
Post a Comment