When Apple does finally throw its hat into the generative AI ring, Deepwater Asset Management’s Gene Munster doesn’t expect Apple’s product to be much better than the competition, but he does it expect it to be good enough to boost the stock.
Ian Krietzberg
for TheStreet:
Unlike its “Magnificent Seven” peers, Apple remained conspicuously quiet about any efforts it’s been making in the generative artificial intelligence space throughout 2023… And when Apple does finally throw its hat into the generative AI ring, Munster doesn’t expect Apple’s product to be much better than the competition.
Munster said that Apple’s position in terms of privacy will likely force the company to purchase third-party training data just like its competition… [W]ithout tapping into its own proprietary data, Apple will “be in the game,” but it won’t be the best.
But beyond the technical performance of the models is their usability. And Munster expects Apple to be on par with Google when it comes to the sheer number of people using their generative AI products. Enabling a generative AI Siri, Munster said, will give Apple a “huge advantage.”
“The average consumer doesn’t care that this is copying out to GPT-3.5 when GPT-4’s out. They care if can they hit a button and get what they want,” Munster said, likening his Apple AI expectations to the iPhone, which has “never been the best phone. It’s always been the phone that works best. I think that’s exactly what they’re going to do.”
MacDailyNews Take: We actually can’t remember the last time we used the “phone” (voice calling) portion of our iPhones.
Munster expects the stock to start getting credit from its coming push into AI sometime this year, likely in the wake of a potential June unveil.
MacDailyNews Take: Generative AI. M3 Ultra. A Siri that doesn’t suck. This year’s WWDC is already sounding like a jam-packed blockbuster!
Please help support MacDailyNews. Click or tap here to support our independent tech blog. Thank you!
Support MacDailyNews at no extra cost to you by using this link to shop at Amazon.
Definitely agree with MDN that a Siri that doesn’t suck would be a big welcome – and a long time coming. But how will use of generative AI make Siri less annoying? Because it’ll come up with a more conversational/human way of saying “Sorry, I can’t help with that”?
Generative AI is useful where large amounts of data on a subject can be used to build a model that the user can query conversationally. Where would that data come from? The only large data stores that exist on the iPhone are photos, music, and iCloud Drive. Maybe also email. Contacts and Notes data stores are probably pretty small. But Siri is most lacking in providing answers to general questions. She usually punts and gives you the results of a web search. Which is frustrating when you’re driving. And generative AI on the device is not useful at all in this general use case.
Apple would have been better off training customers to make very specific requests of Siri that would work 100% of the time, even without an internet connection. Take something as simple as “Siri, read the contents of note X”, impossible. “Siri, how much does X Apple device cost?” kicks you to a web search, WTF?
We’ll see what Apple reveals, but I get the sense it’ll be whatever they’ve slapped together with popsicle sticks and shoestring since ChatGPT was announced. I’ve noticed NO improvement in Siri in at least 5 years. If they were doing all kinds of AI research over the years they would have incorporated it into their marquee “virtual assistant” by now.
My feeling is these analysts are looking to down the stock so they can make a buy and kill it. If there is better phones on the planet, good luck finding it. Again a ploy so his clients can make more money
the telling thing is when apple starts to train its AI model it will most likely be using NVIDIA hardware. No one is using Apple M series (mobile chips) for any heavy lifting. Imagine how things would be different if apple used its pole position with TSMC to design workstation level chips and not mobile chips masquerading as desktop chips.
Could not agree more, #include. Well said.
Recent rumors of high-end Mac Pro computers being phased out can only lead to one conclusion — their M series chips, no matter how carefully cherry picked touted in Apple speed tests — simply cannot compete with workstation level chips from Intel, AMD and NIVIDIA integration. Thus, they will never compete for DOMINANCE in 3D modeling, video production, graphics, gaming.
The sad part is decades ago Apple commanded the lead, but under iPhone/iPad Cash Cow Cook surrendered the lead immediately after taking over.
Yes, yet ANOTHER reason piling up we need a creative, forward thinking CEO who lead Apple into the future, AI included…
Add to that a section of the Apple services dedicated to Vision Pro (news, arcade, music, exercise, tv+, etc)
hope apple come out with Siri Pro!