Members of my C-suite are doing this today on a raspberry pi, so the promise is not empty. Key difference is they aren't using local AI to do it.

For the M5 Ultra, I suspect it would be valuable for someone who wants to achieve all the above and more, but with local AI due to data privacy concerns, and also not regulated data that comes with lots of other requirements solved by more traditional approaches.

Three possibilities:

1. The type of person who deals with lots of intellectual property using expensive Mac-only desktop applications that aren't meant for servers, whose mind has formed positive associations with the term "Apple Intelligence", whose values overlap with Apple's lawyer's values, who actually stands to profit from having a Mac that's more powerful than anyone else's Mac, whose long-term goals are not impacted by planned obsolecense on a piece of computer hardware costing over $25k ($50k after 1TB SSD add-on).

2. Trust fund beneficiary who wants to show off, LARP as #1, prime target for Apple's marketing.

3. 2026 kit for billionare-class iPad babies. All brain rot content is 100% local AI-generated. Never have to speak to your children again. A true "we have dead internet theory at home" machine.

Weird swipe, I'm not trying to sell it to you, this is just where I see it going, and why these things are going to sell (and why 512 gig M3 Ultra Mac Studios have skyrocketed in demand/price). It's not been an empty promise, in that it's been providing lots of very concrete value to us already.

What exactly is empty about any of what he wrote?