A viral video about a fantasy phone called PhoneBloks inspired my Scientific American column this month. It’s a concept that almost everyone loves immediately: a cell phone whose components you can snap in and out as newer, better ones came along. Far less e-waste—and money-waste; you’d own only one phone and just keep it up to date with new parts.

That idea will never see the light of day, for the reasons I mention in the column. But PhoneBloks isn’t the first idea that seemed great but never flew; the corridors of high-tech history are littered with similar appealing ideas that flopped. For example:

The plug-and-play gadget: A spinoff of the PalmPilot called The Handspring Visor held much of the same attraction as the PhoneBloks. It was a handheld organizer with a big cartridge slot on the back. Into it you could snap a range of accessories: more storage, a GPS receiver, a camera, a cellular transmitter, a remote control, a Bluetooth module, an MP3 module or a voice recorder.

The Visors were on the market for about four years (1999 to 2003) and have fans to this day. But clearly, the concept of interchangeable gadget parts alone isn’t enough to start a revolution.

The simple kitchen computer: Over and over again manufacturers dream of making an Internet-connected computer for the kitchen, bringing the convenience of quick Web lookups and the power of modern processing to technophobes and families. There was 3Com’s Audrey, a $500 machine (released in 2000) intended for “Internet snacking.” There was the Netpliance i-Opener (1999, $100). There was the Virgin Webplayer (1999, $100 for three years of Internet service).

They were all flops; none lasted more than a few months on the market. They were made to be inexpensive, and therefore they were slow and unsatisfying to use—and consumers were leery of buying a single-purpose computer when a regular laptop could do much more.

The simpler word processor: Everyone loves to complain that software—especially Microsoft’s—is overwhelmingly complex and loaded down with unnecessary features. But every time Microsoft tries to satisfy the demand with a simpler, less bloated option, nobody buys it.

There was Microsoft Write, for example, a slimmed-down, much less expensive version of Microsoft Word for the Mac. It tanked.

There was Microsoft Bob, a cartoonified front end for Windows 95 and Windows NT. It became a global joke.

The basic tablet: More recently, there was the Surface RT tablet, a stripped-down, less expensive tablet that couldn’t run standard Windows programs (as its big brother, the Surface Pro, can); instead, it runs a new class of apps developed just for it. The RT has been a gigantic sales flop.

Apparently, we don’t always want what we say we want. We don’t like complexity—but we also like to surround ourselves with power we’ll never use.