I don't think it an either / or. The current AI models, as they are, improve productivity quite a bit. They're just super expensive but the expense is being subsidized so it appears reasonable.
An alternative possibility is that the models become much cheaper and their use becomes more ubiquitous which would be helpful.
I think you're vastly underestimating the technical illiteracy most people have with computers.
99% of people buy a desktop and don't even consider what the operating system is let alone think about changing it to something else. I would imagine they don't even know that a difference exists between operating systems.
I think people don't speak at their phones simply because the experience is so horrible.
I would LOVE to be able to voice control the various aspects of my phone but it's so mind numbingly frustrating that I'm annoyed just thinking about it. If it was better I would sacrifice other features for it.
Everyone mentioning multiple apps such as "rectangle" or "amethyst" doesn't really understand how restrictive they are compared to tools like i3 or sway.
Window snapping, tiling and screen navigation is a HUGE deal. I basically don't use my mouse when programming in Linux and it's really difficult to do the same on Mac.
Maybe your code is ok and understandable to you when you first write it, but when someone else has to work on it a month or two later, a lack of types makes it significantly more difficult.
It sounds like your process is a bit of a drag on you and I think you should improve it.
What? No, I think it'd be great! I'd love to be able to say out loud "kube get pods pipe grep service" and the output to be printed on the terminal. I _don't_ want to say "Hey Google, list the pods in kubernetes and look for customer service".
The transfer between my language and what I can type is great. It starts becoming more complex once you need to add countless flags, but again, a structured approach can fix this.
Totally guessing here but this would be a property that gradually appears as the model size increases.
With a tiny model you would get gibberish and with ever increasing models the response would increasingly approach a coherent answer to a finally correct answer.
The author touches on this in the very article linked. His point is that the low accuracy answers provided by ChatGPT risks reinforcing the doctor's biases and then not thinking outside the box. In other words it risks the doctor blindly trusting the AI too much.
I'm sure I remember Zuckerberg talking about this very thing in one of the Congress interviews he had.
I don't disagree with you, the content producers are always going to outnumber the moderators by a massive margin. It makes reasonable moderation very difficult.