Is it too early to have an “iPhone 16 disappoints” rumor? Poppycock! It’s never too early!
After a lot of hype about Apple Intelligence driving iPhone 16 sales, Ming-Chi Kuo says hold your horses.
…the expectation that consumers will buy the new iPhone 16 for the Beta version of Apple Intelligence in [second half of 2024] may be too optimistic.
Are we talking about “Windows Phone will overtake Android by 2015” kind of optimistic or they’re just a little too high kind of optimistic?
This makes sense, of course, because Apple Intelligence will be rolled out rather slowly over the course of the next nine months. And, also, AI is as over-hyped as a studio-manufactured boy band.
The Macalope typed that metaphor as a joke, but it’s fairly apt. While AI might churn out a few toe-tapping hits, it is a product largely pushed into the spotlight by venture capitalists looking to ride a hype cycle and secure the investments in Nvidia they made based on other over-hyped technologies like crypto, the blockchain, and NFTs.
Anyway, as a result, it looks like Apple is not ordering as many of these devices that no one has seen yet. It’s July so we’re not saying it’s time to panic about iPhone 16 sales numbers, we’re just saying you should clear your schedule for later in the year.
This isn’t exactly a product cut rumor, but July is pretty early to be throwing cold water on upcoming iPhone sales.
If it’s too early to call an iPhone 16 fizzle, is it too early to call an AI one, then? The iPhone 16 hype fail isn’t the only sign.
Like Dare Obasanjo, the Macalope has been more bullish on using AI based on defined data sets rather than “the open web” which is full of everything from how to correctly use the the English language to the latest conspiracy theories about how solar panels are pushing us further away from the sun. Also, conspiracy theories on how they are pulling us closer to the sun. Both theories actually come from the same 4chan poster.
Sadly, a leading example of trying to use AI on defined data sets, legal AI firm Harvey which has been slathered in VC cash like honey-soy glaze on a ham, looks set to disappoint. Obasanjo opines:
…in actually trying to use LLMs for similar tasks, I’m not sold the technology is ready for high stakes use cases.
If AI doesn’t work that great on large, open data sets and it doesn’t work that great on smaller, well-defined data set, what does it work great on? If we are to rely on AI for answers–you know, to do its job–then at some point it’s going to need to work pretty flawlessly. If the Macalope asks it what goes great on a pizza and it says “pineapple and ham,” that’s fine. That’s an opinion. A wrong opinion, but an opinion actually held by real people. If it says “rocks and glue,” which one has, that’s not okay. Imagine asking it a legal question. Or a medical one.
Currently, the best use of AI seems to be as a means of generating ideas, some of which may be good, but some of which could be disastrously wrong. If you intend to use it in a professional capacity, you still need to be an expert in the field. That could be okay, except a lot of the hype cycle has been built on the idea that AI will lead to massive layoffs, which Wall Street loves. That’s a problem.
After all, what good is a technology if we can’t use it to squeeze workers?
The Macalope has been really down on AI, as you may have noticed. It certainly seems like a promising technology, but one that’s being foisted on people before it’s ready to do the job responsibly just so some can get a payday.
Some bubbles should be popped.
