The Daily: What kind of impact Apple Intelligence will have, what this means for OpenAI, and get ready for the AI PC

Audio by Yoram Wurmser, and Jacob Bourne | Jun 24, 2024

Episode Transcript:

Marcus Johnson (00:00):

This episode is made possible by Roundel. If you partner with roundel, you can reach over 165 million guests who look to target for joy and inspiration. It's a lot of people together with you. They'll design curated media solutions that are a seamless extension of the target experience and everything. All of it is batched by unparalleled first party data and of course measurements. If you would like to find out more about what I'm talking about, you can head to roundel.com.

Jacob Bourne (00:29):

It's moving very cautiously on AI. It doesn't even like to use the term AI, so it calls it Apple intelligence, and so I think it's offloading some of the riskier generative AI capabilities to OpenAI, and then it's doing some of the more simple tasks using smaller AI models that are on device.

Marcus Johnson (00:51):

Hey gang, it's Monday, June 24. Yory, Jacob and listeners, welcome to Behind the Numbers DAIly, a Newmarket podcast made possible by Roundel. I'm Marcus. Today I'm joined by two folks. We start with our technology analyst. He's based in California. It's Jacob Bourne.

Jacob Bourne (01:06):

Hi Marcus.

Marcus Johnson (01:07):

Hey fellow. We're also joined by one of our principal analysts who covers everything, advertising, media and tech. This, of course, Yory Wurmser.

Yory Wurmser (01:14):

Hey Marcus, how are you?

Marcus Johnson (01:16):

Hey, chap. Very good, very good. Based in New Jersey, he is, today we're talking all about Apple intelligence. We start with today's facts. How do Americans pay for things? So this is according to a visual caps article that used Federal Reserve data to track the share of payment methods. So gentss, what do you think the number one, two, and three payment methods are today? You can choose from debit, credit and cash.

Jacob Bourne (01:46):

I'd say credit,

Marcus Johnson (01:47):

Credit number one, Yory.

Yory Wurmser (01:48):

I'd go with credit number one and debit number two.

Marcus Johnson (01:53):

Exactly right. It wasn't the case in 2016. 2016, what was that eight years ago? It seems like very recently, 2016, but it was eight years ago. Cash was king with 31%

Jacob Bourne (02:04):

Marcus Johnson (02:05):

The share of payment methods, 31% debit cards, 27% and credit cards used 18% of the time, but by 2022, I guess it's not, yeah, it's not eight years because this data is from 20 20, 20 16 to 2022. So six years later, credit cards had taken the crown going from 18 to 31%. Debit cards were still second upper touch with 29%, and cash had fallen from 31 to 18%, landing it in third place. A few points ahead of electronic fund transfers EFTs at 13%.

Jacob Bourne (02:35):

Now a lot has changed,

Marcus Johnson (02:37):

A lot has changed, and this visual capitalist article one, Marcus L pointed out some very probable drivers of cash losing its importance. One, hygiene concerns, fear over covid transmission using cash push folks through payments. Two, social distancing, the shift towards online shopping and delivery services dYoryng lockdowns. And number three, banking adjustments reduced access to ATMs and banking services dYoryng lockdowns made it harder to get cash. Still a lot. That's 18% in terms of a share of payment methods is still decently high for cash, so it's hanging on. But going down today's real topic, apple intelligence and a partnership with OpenAI.

Marcus Johnson (03:19):

So in today's episode, first in the lead we'll cover Apple Intelligence. Then for another news, we'll discuss the potential of the A IPC gentss. We're talking Apple Intelligence and so Zoe Kleinman and Liv McMahon of the BBC, right? The Apple is to boost its Siri voice assistant and operating systems with open AI's chat GPT as part of their new personalized AI system. It's called Apple Intelligence, and it'll help users navigate Apple devices more easily. Apple Intelligence uses Gen AI to help users summarize messages. They can instantly write replies to emAIls and text in your own writing star edit photos or create reaction based gen emojis as they called and speak to a somewhat more capable Siri voice assistant as well. The goal for Apple is to have a ubiquitous AI system that knows all about you from your messages, calendars, maps, et cetera, and then it can surface the right information and even take action on your behalf if you so allow it.

Marcus Johnson (04:13):

An economist article describes it as a cross between your personal assistant who can handle a series of tasks on your behalf and your 10-year-old kid who can tell you how to make use of your iPhone's latest features. For some tough questions though, Siri will ask users if they're okay, if she asks open AI's GPT for some help. The feature is free for Apple customers and their data won't be used to trAIn Chat. GPT Apple Intelligence arrives later this year. No financial agreements were disclosed regarding the OpenAI and Apple Partnership. So that's the backdrop. Let me start with Yory, what was your initial reaction to the introduction of Apple Intelligence?

Yory Wurmser (04:46):

I'm pretty excited about it, to be honest. I think it's a big deal. A couple of things for Apple itself, for the iPhone itself, it's an integrated strategy around AI, so all parts of the iPhone are kind of woven together by this AI strategy. It's personalized AI strategy, so I think it's a really good move for Apple, the Apple intelligence overall. The most interesting part is Syria. I'm really cYoryous to see how open AI benefits from this. I know what their strategy is or I think I know what their strategy is. That is they want to drive traffic and upsells to chat GPT, but there are a lot of unknowns on that side.

Marcus Johnson (05:23):

Well, let's talk about that for a second then. Yeah. What does this mean for OpenAI? We'll come back to Apple in a second, but they're doing this partnership with OpenAI. What does this mean for that company?

Jacob Bourne (05:32):

I think this means a ton of exposure for OpenAI and really right now in the commercial generative AI race, it's really about getting more market share. Doesn't matter how much money you're making off of it, yet, it will soon, but they really want to get more people using chat, GBT. That's the name of the game, and I think by having this run right on to the Apple's ecosystem, they're definitely going to get some more traction

Marcus Johnson (05:56):

And a significant amount of scale. Yeah, OpenAI says it has around a hundred million weekly users. Apple says it has 2.2 billion devices in use around the world, so a tremendous amount of scale in a hurry. Yory, what else does this mean for OpenAI?

Yory Wurmser (06:11):

Well, I'm cYoryous what it means in terms of their need of compute power. I know a lot of this stuff is happening locally on the phone, but not all of these answers are going to be answered on the phone, and I think the majority of more complicated ones are going to go bounce to the cloud, and OpenAI is going to have to pay for that or at least use their credits with Microsoft. So I think there's going to be a huge challenge for them to deal with that big step up in usage.

Jacob Bourne (06:35):

Yeah, I think Gary makes a great point because OpenAI has had challenges keeping chat chie online at points, and I think it's because of the amount of traffic going into it and the amount of compute resources.

Marcus Johnson (06:47):

Yeah, this also seemed to me like it was a bit of a moment for OpenAI as it establishing itself as a company in this space. I mean obviously it has tried to do that for the last 18 months, but Apple, the third richest company in the world conceding that it can't do AI better, at least for the moment, felt like a bit of a milestone, and this agreement isn't exclusive, and so they could partner with other people. I think Apple, maybe there's a world where they continue to develop their own AI and at some point they say, we don't need you anymore. We've got a system which is somewhat comparable to what Open AI has, if that's even possible. But I thought that that was interesting. And then gentss, what do you make of this? I mean, it seems as though the antitrust folks, I mean they were circling already wasn't a New York Times article a few weeks ago. The US clears way for antitrust inquiries of Nvidia, Microsoft and OpenAI. How much do you think this ratchets up the pressure on them from an antitrust perspective?

Yory Wurmser (07:37):

I think it definitely ratchet it up in two ways. It gives Apple a lot of power. One is just in terms of hosting these different chat bots, whether it's Gemini from Google or Open AI's Chat GPT, it gives them a lot of power in terms of whether they stick with them or provide other alternatives. But I think secondly, once you create this ecosystem where you can go strAIght into apps and get information, it gives Apple even more power in terms of over the whole app ecosystem and how much exposure these apps have. So I think that's probably the more fundamental way that it rAIses antitrust issues.

Marcus Johnson (08:13):

Jacob, that was one of my concerns. There was one ask, I can't remember what it was now, a headline or sub-headline, maybe a sentence in it, but it was basically saying, what does this mean for apps? Because it's talking about surfacing a lot of this info. One of the examples it says in terms of how Apple intelligence can help you is if you say, open that podcast my wife sent me a few days ago, and Siri goes into your emAIl or your messages from the person it knows is the user's wife, it finds the recent link to the podcast and then opens the file in the podcast app. So that's like three or four apps you might have to open to find this thing, and now it's just going to basically fire up the one you need

Jacob Bourne (08:46):

And it kind of changes people's relationship with their or could change people's relationship with their phone in terms of how they navigate their phone instead of swiping in looking for apps. They're accessing what they need through a Jared I model, and so we could see some apps benefit from that and then some maybe potentially be the disadvantage.

Marcus Johnson (09:07):

That's true. Yeah, there might be some winners here as well. Yeah, some other examples. It was saying it could find out what time your mom's flight lands, find a certAIn photo and send it to a certAIn friend. I mean, these things do seem helpful, but yeah, what does that do to the usage? It did seem, one of my takeaways was it felt underwhelming, but that appears to be by design. It does the Apple Intelligence announcement.

Jacob Bourne (09:26):

Yes, and I think here Apple is really concerned about its reputation. It's moving very cautiously on AI. He doesn't even like to use the term AI, so it calls it Apple intelligence. And so I think it's offloading some of the riskier generative AI capabilities to OpenAI and then it's doing some of the more simple tasks using smaller AI models that are on device on the iPhone on your Mac. But I still think even with these more simple tasks, there are still things that could go wrong, and I think that is something for Apple to be cautious about.

Marcus Johnson (09:58):

Yep. Yeah, they're not trying to help you do something you've never done before. It's focusing, its AI on being personalized, private, and practical is what it says, and that seems like it practical is probably just what the doctor ordered. AgAIn, it doesn't feel revolutionary, but that's not what people are looking for. That's not what most consumers can handle. And Gen AI has been criticized for being a solution in search for a problem. Apple is trying to use it to reduce friction for its iPhone users experience every day. And there was actually a good quote, Nathaniel Whitmore, CEO of AI education company Super Intelligence saying one of the realities of gen AI right now is that the impact isn't about wiping out entire industries off the map, but about individuals finding small but significant time savings through new AI powered processes. Apple Intelligence is betting that the way most people are going to get into AI is by doing the same things they're doing now just a little faster.

Jacob Bourne (10:54):

Yes, and I think what Apple is doing here is really trying to sell its iPhone and its Max. It's not trying to sell AI, and I think that's going to be a difference between IT and the other tech giants.

Marcus Johnson (11:05):

Ai at the end of the day, they're a device company by and large, they do some services for, I think it's only about 20% of their business revenue wise, and almost everything Apple does is about selling more phones, sales of phones, which make up iPhones, which make up half of Apple's revenues are slowing down. Apple's worldwide shipments of iPhones just saw its lowest Q1 in three years, and then also a year ago, iPhones accounted for 40% of all smartphones activated in the us. That figure is now 33% according to consumer intelligence research partners. So they're trying to do things to stem the tide of iPhone sales plateauing and starting to tick down a bit, hope King of Axios pointing out that the newest AI features are avAIlable only in the company's most powerful phones, meaning customers will have to upgrade to get access to Apple Intelligence. They're putting it in their new devices. You want access to Apple Intelligence, you're going to have to buy those new devices, and a lot of people are due for an upgrade. A lot of the upgrade cycle was about four years ago when 5G was kind of hitting devices. And so three, four years is typically when people say, this phone isn't working anymore or it feels slow, or I can't scratch it anymore than I have, and so they're looking for a new one.

Yory Wurmser (12:12):

Yeah, I think it is really going to drive a super cycle. I mean, depending on how good it is when it comes out, but if it is legitimately a game changer, unless you have an iPhone 15 Pro or newer at that point, probably iPhone 16, you won't be able to use it. And I think that'll drive a lot of upsells if in fact it is revolutionary. Yeah,

Marcus Johnson (12:33):

For sure. It seems like I mentioned some of the examples that says it can do already, some of the examples it can do. I mean, you mentioned Siri being a huge part of this Yory, Jason Avin suggesting a few ideas. One was few ideas of what it could do. Siri could notice an emAIl from a colleague notice. Notice an emAIl from your colleague asking if you're avAIlable for a meeting next Tuesday. See your calendar is pretty full suggest a different time, and a step further would be it would send you a notification and offer to reply to the emAIl in your tone and voice, so noticing things before you even notice them and suggesting things so that there is less friction. It says it's trying to be this kind of personal concierge whilst making gen AI more consumer friendly. The other part of this gens, I wonder what you make of this is Apple kind of got made fun of for relying on open AI for its AI, but others haven't fared much better, right? Google and Microsoft both face backlash over the mistakes for their AI products already.

Jacob Bourne (13:24):

Not only that, but if you think about it, Apple's approach isn't wildly different from Microsoft's. I mean, both companies are working to improve their internal AI capabilities while in the interim relying on OpenAI, the difference is that Microsoft got to it first, but it's really, it's a very similar strategy.

Yory Wurmser (13:43):

Let's not forget that that's what Apple did with maps and navigation and search and Google. They make 10 plus billion in traffic acquisition costs from Google, so there is some upside there for Apple. Yeah,

Marcus Johnson (13:57):

And like I was saying, they can always develop their own AI and then cut open AI out. It's not exclusive relationship Apple. Yeah, they could use Gemini, Alphabet's, Gemini potentially. And I've always sAId this content is king. I mean the content here is chat GT's answers, is it going to surface the best answers? But the wild card there is that the pipes that carry the content, in this case, the iPhone are the ace. You need the devices to carry the answers in the first place. So Apple's in such a strong position in that regard. The other element here is partnering with OpenAI gives Apple a bit of a liability shield. David Wagner, portfolio manager at Apple shareholder, Aptis Capital Advisors was saying Apple is passing off the liabilities of AI to a partner like OpenAI because it knows which battles to fight. Jacob, to your point, very strong brand, particularly when it comes to primacy. And so is that fAIr to say this gives them a bit of a liability shield? They say we're relying on OpenAI.

Jacob Bourne (14:49):

It is a bit of a show, but still the Apple intelligence itself, the part that's stopped using chat g PT is still accessing some pretty personal content on people's stones, even if it's not doing so through the cloud. True. If Apple Intelligence gets something wrong when using people's calendar or contacts or anything and it does so frequently, and then that could be a trust issue for consumers.

Marcus Johnson (15:13):

I mean, this might have led me into my last question, which is why might this be a bad direction for Apple? Any other reasons?

Jacob Bourne (15:18):

I mean, I think apart from that, it's largely probably a good direction for Apple just because it's a big tech company and not getting into generative AI at this point. It's not an option. It pretty much had to do something, and so it's doing something, it's offloading some of the risks to open ei, which I think is probably a safe move. If Apple Intelligence works as advertised, then I think it should be in the clear. I think that's probably the biggest question.

Yory Wurmser (15:43):

Yeah, I mean, I think one other downside is that just like they do get traffic acquisitions caused from Google, but they don't get the advertising, they don't get upsells of, they're giving those upsells to open AI or Gemini or whatever for professional services, which is something they're, I think, willing to give up, but that is something they're losing.

Marcus Johnson (16:02):

So real quick, just to quantify it for folks, we talked a lot about a lot. How much of an impact Jacob would this have on the AI race out of 10

Jacob Bourne (16:10):

Ities? I would say it's high. I would say it's an eight out of 10. And the reason being so far, we're seeing a lot of enthusiasm coming from the enterprise over generative AI and consumers are maybe unfamiliar with generative AI, maybe don't know how to use it, maybe not trust it. Apple's move here is directly AImed at the Albert's consumer, and it's taking such a different approach with this very simple approach to stick to technology. It's not a lot of fanfare around this. Meanwhile, the other tech giants, some of them have grand plan announcing grand plans about how they plan to build an artificial general intelligence which consumers maybe don't want or need. So I think that Apple's kind of being a pioneer here in terms of potentially bringing generative AI to the masses, and I think that's going to make a huge difference in the news.

Marcus Johnson (16:54):

Yeah, Yory,

Yory Wurmser (16:55):

I was going to say out of 10, but that would be boring. So I'm going to say seven out of 10.

Marcus Johnson (16:59):

7.9. Yeah,

Yory Wurmser (17:02):

I think it is important. I think a lot depends on the execution. So I'd say it's a little too early to say whether it's an eight, but I think potentially it's definitely very important. And one other way, it's very important too. I think devices are going to wearable devices like smart glasses and will utilize something like Siri or Gemini or Meta AI or whatever. So this helps Apple with that as well. Yeah.

Marcus Johnson (17:27):

Alright, gentss, that's where we'll leave the lead story for today time now for the fourth quarter of the show today. In other news, just one story, get ready for the AI C.

Speaker 4 (17:39):

Pour back one

Marcus Johnson (17:42):

In free. Axios is saying that we should get ready for the A IPC. They note that the biggest names in computing are all poYoryng resources into building gen AI into their devices, hoping to inject the venerable PC market with a boost of AI excitement. The promise of the AI PC is that because of an AI processor called a neural processor or a neural engine, you'll be able to do more of the work on the device as opposed to in the cloud. Ms. Reed points out that that could strengthen data privacy and reduce climate stressing data center power demands. Microsoft has already sAId it will equip Windows PCs with AI features including a dedicated copilot AI keyboard key. But Jacob, I'll start with you. What interests you most about the advent of the AI pc?

Jacob Bourne (18:26):

Yeah, so I think first, in order to really get these AI PCs up and running to their full potential, we're going to need to see some more powerful and highly efficient AI chips and what I think are currently avAIlable today. Of course, the chip companies are working on it, but I think the really exciting thing is that once we get there, we're going to see a huge spillover effect into other devices. And ultimately we're going to see things like generating AI powered smart appliances that don't require the internet access to function. And I think at the end of it all, we're going to see the robotics industry really use these highly advanced AI chips to really power independent, advanced robots that can act out in the real world.

Marcus Johnson (19:12):

Uri.

Yory Wurmser (19:12):

Yeah, I mean, I agree with Jacob, bro. I also think it changes the interface with desktops and a potentially has revolutionary effect as something like Windows did 25, 30 years ago where how you interact, how you choose things could just change to more voice mode or using fuller descriptions rather than clicking and typing things in. I don't think it's going to be that extreme right away, but I think it provides an entryway for that type of interface.

Marcus Johnson (19:42):

One of the stories here was that Microsoft was going to launch something called recall, a new AI feature, but it's since pulled that feature. It was supposed to be part of its new copilot plus, but because of secYoryty concerns, it's decided not to release that. So recall is going to let you find, quickly, find anything you've ever seen or done on your laptop. It can also rewind your work a session back to a specific moment so you could change your mind about a particular direction you took quickly. Jacob, what's going on there?

Jacob Bourne (20:07):

Well, I think there's just such speed and momentum in terms of some of these feature rollouts around generative AI and the secYoryty community is really concerned about potential unintended consequences of these features, especially when you have the AI accessing everything on people's computers, even if it is on device. There's just a significant amount of concern about secYoryty breaches that we haven't seen to the scale that could be at risk of much bigger breaches than the past. How

Marcus Johnson (20:40):

Is this different from Apple Intelligence though? That is saying we're going to access everything that's in your phone and we're going to try to present the most relevant information when you need it. This is kind of a similar feature ish.

Yory Wurmser (20:50):

I mean, I think it is pretty similar. Apple generally has, I think, a better reputation than the Windows ecosystem for secYoryty, but it's definitely something that's relevant in the Apple ecosystem as well.

Marcus Johnson (21:03):

Yeah, yeah, I PCs, they're already here. According to canals, nearly 20% of PCs shipped this year will have AI capabilities that's rising to 60% by 2027, and the PC market could use a shot in the arm worldwide PC shipments have been falling since 2021 and by a lot shipments were down 30%, 30% from 2021 to 2023. So just those two years down, 30% according to IDC. That's what we've got time for this episode. Thank you so much to my guests as always. Thank you to Yory.

Yory Wurmser (21:35):

Yeah, glad to be here.

Marcus Johnson (21:36):

Thank you, Jacob. Thank you, Marcus. Thank you. Yory. Thank you to Victoria who edits the show. Stewart, who runs the team, Sophie who does our social media. Thanks to everyone for listening in. We hope to see you tomorrow for a special video podcast of the Behind the Numbers DAIly, an e-Marketer podcast made possible by Round Dew. You can go to YouTube and search for e-marketer if you would like to watch that episode. Or you can listen to it all the places you normally listen to us.