There’s a variety of powerful “use cases” for Large Language Model Artificial Intelligence systems.

But it’s all-too-easy to use their power recklessly, to produce work with is more convoluted, empty, and even damaging, than it may appear at first glance.

And the discernment needed to know when LLM AI use is reasonable will be harder to come by, as these systems get better (which they will … overwhelmingly so).

Best that we all do some hard, firsthand thinking on the topic.

Nice that you’ve already done so.