This story from a scientist, stuck with a 5.8GB Excel file that refuses to open, isn't just a technical hiccup. It's a clear sign that the tools we rely on for data work are no longer fit for the data we actually have. Our opinion is plain: the problem isn't the file size. It's the tool.
This scientist did everything right. They tried the Data tab, attempted to filter unnecessary values, and used the Get Data feature to reduce the load. And Excel still errored out after thirty minutes of "establishing a connection." That's not a user error. That's a tool hitting a wall it was never designed to climb. Traditional spreadsheets were built for rows and columns that fit on a single screen. They were not built for modern datasets that measure in gigabytes. The scientist's workflow stalled, not because of poor planning, but because the software assumes data will always fit inside its memory limits. When it doesn't, the whole process grinds to a halt.
What this means in practical terms is that millions of professionals, scientists, analysts, engineers, are spending their time fighting file limits instead of doing the actual work. They're breaking large files into smaller chunks, losing context. They're waiting for connections to establish, only to get error messages. They're told to "clean" or "reduce" data before they can even look at it. That workaround is accepted as normal, but it shouldn't be. An AI-native spreadsheet doesn't treat a 5.8GB file as an exception. It treats it as a dataset to be queried, not loaded. It opens the file in place, processes only what you ask for, and lets you filter, sample, or aggregate without ever forcing the entire file into memory. The scientist's workflow should have been: open the file, ask for the relevant columns, and start analyzing in seconds.
The solution here isn't a better version of the same approach, more RAM, different import settings, or a smaller file. The solution is a tool that was built for data at this scale from the ground up. If you are routinely working with files that stall Excel, the takeaway is not to work harder at workarounds. It is to ask why you are still using a tool that treats your data as a burden rather than a resource. The scientist in this story deserves a tool that meets them where their data actually lives. So do you.