I was shocked by the very one-sided view Prof. Shneiderman presents towards his view that AI is only a "tool" for humans, like a bulldozer or airplane. It's true he makes some good points, and the position he seems to hold is a reasonable one, but presenting it as the only reasonable view seems like an extreme position.
In my opinion, every technological improvement is a tool as Shneiderman suggests, until suddenly it is something much more. Take for example the job of human computer - a common profession in the early-mid 20th century that helped with everything from the war efforts to sending man to the moon. When the slide rule was first invented, this new tool did much of the calculating of a human computer's job, but certainly did not render them obsolete. Neither did the invention of the electric calculator. However, once personal computers became mainstream, they did almost perfectly replace human computers. Now, the job just does not exist anymore - electronic computers are not mere tools for these human computers, but rather eliminated them entirely.
But don't take my word for it - listen to people who actually know what they're talking about! It is well known that Elon Musk and Stephen Hawking believe AI will result in unemployment unlike anything seen before in human history, and consulting firms like McKinsey believe that almost 25% of human work could be fully automated using existing technologies. Or listen to true AI experts - researchers such as Martin Ford and Erik Brynjolfsson believe "this time could be different" - that instead of simply making humans more efficient at their work as previous industrial revolutions have done, advances in AI will eliminate large swaths of the job market altogether.
Neither one of these possible futures is certain, but presenting only one - that AI is nothing more than a "tool" - seems irresponsible.