It's on the falling edge of the hype curve. It's quite expected, and you're right about where it's headed. It can't do everything people want/expect but it can do some things really well. It'll find its niche and people will continue to refine it and find new uses, but it'll never be the threat/boon folks have been expecting.

People are using it for things it's not good at thinking it'll get better. And it has to an extent. It is technically very capable of writing prose or drawing pictures, but it lacks any semblance of artistry and it always will. I've seen trained elephants paint pictures, but they are interesting for the novelty, not for their expression. AI could be the impetus for more people to notice art and what makes good art special.

That’s a good comparison.

Note though that input prompt voices have not revolutionized anything. I personally avoid them as much as possible.

To call it 'ending' though is a stretch.

That's only the title and it is only referring to the hype ending, not development of the technology:

Eighteen months later, generative AI is not transforming business. Many projects using the technology are being cancelled, such as an attempt by McDonald’s to automate drive-through ordering which went viral on TikTok after producing comical failures. Government efforts to make systems to summarise public submissions and calculate welfare entitlements have met the same fate.

If you read the article you'll find the author is not claiming we have universally reached the end of AI or its hype cycle yet:

A Gartner report published in June listed most generative AI technologies as either at the peak of inflated expectations or still going upward.

There's a whole section of the article dedicated to answering why this is the case, too. I recommended you read the article, as you seem to have misinterpreted it based on the title.