Inklings #006 📧

Come for some talk about Apple's succession plans, stay for a dive into the about-to-be-20-year-old classic, The Prestige. I back last week on the Escape Hatch podcast covering yet another Christopher Nolan title. (And now perhaps you know better understand my angle on the piece about John Ternus.)

You can find it here: YouTube | Apple Podcasts | Spotify

Thoughts On...

📦 Amazon's FedEx and UPS Killer – This was always inevitable. And obvious. 'Amazon Supply Chain Services' may be an awful, generic name, but it will soon just be called 'ASCS', just like 'Amazon Web Services' is 'AWS'. Speaking of, that was clearly the model here. Build up the capabilities to do it for yourself, then open it up to the world. One big difference is the direct competition Amazon faces here in the aforementioned FedEx and UPS – both of which, are, of course, also partners. Or were. We'll see how both react now with their stocks dropping around 10% on the news. And, of course, there's the US Postal Service, which Amazon just struck a new deal with after nearly ending their 250+ year run simply by suggesting they may no longer deal with them. While Amazon may not have a smartphone (yet?), they do have the physical logistical mechanism by which their competitors increasingly must play ball with them... [Reuters]

📲 Can Intel Land the iPhone? – Are we about to see Apple with Intel Inside yet again? Not so fast. While it is potentially big news that Apple would work with Intel to make chips for their "main devices", it sounds like such discussions are super early. And concurrent to ones with Samsung, who would seem far better suited to handle such a task in the short term given their current position as a foundry (though obviously Intel's American roots help narratively here, if nothing else). This is all an offshoot of earlier reports that Apple was talking to Intel about working with them to fabricate chips for some lower-volume devices (read: not iPhones). But now that NVIDIA has started to dominate the TSMC relationship, to the point where even almighty Apple is being pushed back in the line, well, Apple needs to have other options. That said, it still feels like we're at least a few years from Intel reliably making chips on the type of cutting-edge nodes that TSMC can do – if Intel can get there at all. Then again, TSMC is also apparently cutting back on some costs for cutting edge chips, so maybe there's a window... It would be fun to have Intel back with Apple, even just in a way. [Bloomberg 🔒]

🥽 Is Apple Killing the Vision Pro? – No. But because of an earlier report saying as much, John Gruber had to lay out the common sense argument as to why a funeral is not happening – at least not any time soon. Which is not to say the device is a success. Obviously, it's not. And while I think Apple made a mistake is launching it before it was fully baked (including with developer support), it's slowly but surely getting more baked as an ecosystem and product over time. A bunch of the recent content releases have been good! The problem is that the device itself remains far too cumbersome and laborious to use. And, of course, is too expensive, perhaps by an order of magnitude. But if they can bring new elements of "Spatial Computing" up from the ground with a new Smartglasses product, while bringing others from the top down from Vision Pro, perhaps there's a sweet spot in the middle. The question is timing on that – is that a few years away, or a decade. It will be one of the key first questions for John Ternus. But unless something goes wrong and Apple needs to start cutting costs, it's more like a product on ice for the moment. Entertaining relatively few of us! [Daring Fireball]

🛎️ The End of Ask.com, RIP Jeeves – I was never much of an Ask Jeeves user – I went from Lycos and Yahoo to Metacrawler before switching, like everyone else, to Google – but I definitely have a fondness for both the idea of an "answer engine" and the P.G. Wodehouse-inspired Jeeves character. Both seem more top of mind than ever in a way in our current Age of AI. LLMs now power "answer engines" in a way that web search never fully could, of course. And there's an open question as to whether or not to anthropomorphize the chatbots. Certainly, Siri and Alexa kicked this off in one direction, Microsoft has since tried to give newer AI a face, and voice may be the thing that pushes it all over the edge. Regardless, how long until an AI service buys the ask.com domain from IAC (now, incidentally called 'People' after their publication) for a lot of money? Though probably less than the $1B+ IAC paid for the entire company back in 2005. [NYT]

Asides...

• Speaking of web search, queries hit an all-time high at Google last quarter. Why? AI, of course. Well played, Google. [Verge]
• OpenAI pushes out a separate ChatGPT app for those forced to use Microsoft's Intune endpoint management (at work and/or school). I can't tell if this helps or hurts my 'BYOAI' case... [9to5Mac]
• Just as Anthropic rolls out their $1.5B JV for corporate AI deployment, OpenAI formally announces their $4B company for this. [WSJ 🔒]
• Is OpenAI fast-tracking their phone project for the first half of 2027? Ming-Chi Kuo sure seems to think so in an update (are we still sure this isn't about their other AI devices?). [@mingchikuo]
• With 'May the Fourth' yesterday, Nielsen released the "winners" of streaming in the Star Wars universe. It's mostly about the movies – except for the newer trilogy, which doesn't seem to have any lasting power with any demographic. Time for some new ones, Filoni! [THR]

Below, members of The Inner Ring will find thoughts on:

• OpenAI's Journals v. Elon's Texts
• OpenAI's $29B Man
• A Growing Gap Between Frontier vs. "Open" AI?