She is talking pure nonsense
MS Access engine is the 7th most popular database engine in the world
https://db-engines.com/en/ranking
and it looks from that source that it is actually getting more popular.
Visual Basic ranks amongst the top 20 most popular languates in the world
http://redmonk.com/sogrady/2017/06/08/language-rankings-6-17/ and that study looks like it is measuring popularity rather than usage. Usage is probably higher.
I would suspect there are more VBA developers than any other language in the world with nearly every IT support person able to do basic VBA support and many are very good. VBA along with C# have absolutely massive number of forums from which you can obtain support. Utter access and this one included. The experience in VBA and C# around the world probably completely outstrips most other languages put together. The language is stable as hell libraries rarely change and it is likely you could open an Access 97 database on a 2018 machine and it will probably 90% run. Holy hell your lucky if a web framework lasts a year. If a web framework changes it often requires almost a complete Application Re-write. The only other language as stable is probably SQL and that's the second pillar of Access anyway.
Microsoft is one of the major players in the IT industry and Access has an unbroken track record of support for 2 decades. It doesn't look like they are going bust.
Access is the most popular desktop database in the world.
Yes its not easy to pick up someone elses project and run with it. But if someone writes a book or writes a document or creates an excel spreadsheet or god forbid trys to create an application (in whatever IDE) no one wants to pick it up and run with it after its complete - The fact that the domain is complicated and you are the only one that knows about it is NOTHING to do with MS Access - thats a policy decision to create such a crazy administrative system by management.
I assure you there are a lot more people able to pick up a second hand MS Access application than there are able to pick up a second hand web application.
OK but the CLOUD
The CLOUD has a lot of potential if you want to design one system between multiple buildings that aren't networked but why if you have spent a decade or more putting networking into a building do you suddenly abandon that bought and paid for hardware and pay Amazon or Microsoft to rent a duplicate in their clouds. Seriously these people really don't think too deeply sometimes.
the argument is usually there is an app out there that if you rent will mean that you don't need IT support and that's potentially a good saving.
A lot of these apps are so bespoke they often cost 5k or more per year and often designed by very few people - these app companies go bust all the time often because they have so few customers that their business model really struggles. This will leave you back at square one.
The best solution is nearly always a hybrid - yes use bought in apps - but really be careful about it - choose your battles carefully - do some in house development for no other reason than it gives you an indication of what is difficult and easy to do. Work with outside contractors for real value added services but don't be afraid to try and develop things yourself especially if absolutely nothing exists. Always be flexible and ready to adapt and don't go all out for the cloud or all out for internal.
If you have assets you SHOULD Sweat them