Once upon a time, maybe in the middle of the decade before last, there was an idea that was called “Bring Your Own Device”. iPhones and MacBook Airs showed that computing devices didn’t need to be heavy and ugly. They could be light and stylish, with batteries that lasted more than an hour.
The mostly privileged owners of these sleek and stylish devices demanded that they should be free to display their privilege and panache by using their sleek and stylish devices at work. Because of irritating issues like security, this was often quite difficult, but luckily, most of the privileged didn’t really use much more than email at work, so solutions were cobbled together.
Then The Cloud happened, and people started to realise that the devices weren’t the problem. The problem was the shitty software that large organisations used because large organisations had used shitty software for as long as anyone could remember. But there was no such thing as “Bring Your Own Software” Instead we had “Shadow IT” and Shadow IT was Bad.
Things rubbed along. People with good intentions trying to find better ways to get their jobs done. Organisations trying to maintain the idea of walls. A perpetual whack-a-mole.
And then came the Large Language Models. At first, the focus was on yet more “attack vectors”, but soon people across organisations started to find ways to roll their own software, and with this newfound power and freedom, they rolled and rolled and rolled. People started to show each other their own-rolled software, suggesting that the thing that they had built might be useful for others. But because everything that had been built had been built for a user base of one, it was easier for everyone to roll their own than to use software from others.
And at that point, people realised that bountiful plenty wasn’t as good a thing as they had been led to believe. New walls were built, and most people were quietly relieved to have someone else decide where the boundaries lay. A few, of course, were already looking for ways around them.
mmitII 