Apple and Meta Discussed AI Partnership Despite Longtime Rivalry

[00:00:00] Welcome to Tech News Briefing. It's Friday, June 28. I'm Zoe Thomas for The Wall Street Journal. One of the world's biggest online retailers is looking for new efficiencies and cost savings by turning to generative artificial intelligence. We'll tell you how Amazon's finance teams are

[00:00:37] using the tech. And sticking with AI, longtime rivals Apple and Meta have held talks about potentially integrating the Facebook parent's generative AI model into Apple Intelligence, its AI system for iPhones and other devices. Our reporter Salvador Rodriguez will explain

[00:00:55] what each company could gain from working together. First though, many companies are looking for ways to effectively move from testing and experimenting with generative AI to using it in prime time. Amazon's finance teams are doing just that. Here to tell us more is WSJ

[00:01:17] CFO Journal reporter Mark Maurer. Mark, can you tell us about some of the specific applications Amazon's finance teams have been using generative AI for? Amazon is using generative AI in a bunch of different ways right now throughout the finance function. And there are some tools

[00:01:34] that they're still experimenting with and others that they've put into practice. But right now, the company's finance teams are applying generative AI to things like detecting fraud tied to transactions, forecasting future finances, reviewing contracts, and for certain tax-related

[00:01:53] work. Can you tell us a little bit more about the tool that its tax team has built? Amazon's tax compliance team recently created a tool that uses generative AI to help validate invoices that are coming in for value-added taxes. And these are taxes that companies pay

[00:02:14] in varying stages of production of a good or service. So when the team receives an invoice, the AI tool automatically checks the validity of that invoice before payment goes through. So instead of manually validating all of these invoices, generative AI

[00:02:37] allows a team to automate a process that long had a strong manual component to it. How does Amazon use generative AI to help with its many contracts? So a company like Amazon has thousands of contracts with suppliers and customers and others.

[00:02:56] And it has an AI tool that takes into account things like standard reference data, policies, history of past contracts. And it uses that to gauge the changes that it needs to make to accommodate those contracts. So the AI extracts key details for sort of a more streamlined review

[00:03:19] that humans will do. And one of the executives I spoke to noted that this is particularly valuable because it allows humans to not have to wade through sometimes incredibly dense contracts. Why has Amazon chosen these areas?

[00:03:34] These are areas that Amazon saw as opportunities to try to be more efficient and maybe reduce the extent to which humans were spending time on repetitive manual tasks. And perhaps those employees could be spending time elsewhere, devoting more of their week to

[00:03:58] critical thinking. And the company at the same time can also help save costs and increase the accuracy of this work. That was reporter Mark Maurer from the WSJ's CFO Journal. Coming up, could a potential deal for generative AI make partners out of long-time rivals?

[00:04:18] We'll learn more after the break. In its hustle to catch up on AI, Apple has talked with long-time rival Meta, the parent company of Facebook and Instagram. According to people familiar with the matter, the discussions have been about integrating Meta's generative AI model into Apple Intelligence,

[00:05:02] the recently announced AI system for iPhones and other devices. Here to tell us more about the talks and why they matter is our reporter Salvador Rodriguez. So Sal, really simply, why would Meta want to integrate its generative AI model into Apple

[00:05:16] Intelligence? Well, at the end of the day, even though the tech industry is having a generative AI revolution right now, Apple is still the king when it comes to hardware. The way that

[00:05:27] humans interact with this technology is typically on an iPhone or on a Mac. And so if you want your technology to be distributed far and wide, being friends with Apple is not going to hurt you. What does this mean for Meta's own AI ambitions?

[00:05:43] Meta has been working on AI for a decade. But when this revolution started, it's because OpenAI launched ChatGPT in late 2022. And since then, Meta has fallen under the radar when it comes to

[00:05:56] the conversation around AI. And so what they've been trying to do the last few years is popularize their own large language model, which is called Lama. So by striking a partnership of this

[00:06:09] magnitude, that would certainly raise the profile of Lama and just all the work that Meta's AI organization has been doing over the last few years. Sal, you reported that Meta and other companies developing generative AI are hoping to take

[00:06:22] advantage of Apple's massive distribution through its iPhones, similar to what Apple offers with its App Store on its iPhones. What makes the process of getting an external AI onto Apple's AI platform different from getting something onto the App Store?

[00:06:38] What we're talking about here, and this has come up a lot since the iPhone was birthed into the world in 2007, it's one thing to have an app, it's another thing to be more tightly integrated

[00:06:50] into the actual operating system. If you've ever thought about anyone who uses the Safari browser or uses Apple Maps to get around, a lot of folks, they just want their phone to work. They don't

[00:07:01] really want to go and shop and find the absolute best option. So if you are the AI model that's on there by default, it's kind of like prime real estate when it comes to these handheld devices.

[00:07:13] You know, we should note that discussions with Meta and other AI companies haven't been finalized and they could fall through. But Sal, why would a deal between Apple and Meta be particularly noteworthy? The significance of these discussions that have occurred is noteworthy because Apple and Meta

[00:07:31] have been at each other's throats for the last decade. There's very rarely any rivalries in Silicon Valley that get as heated as that between Apple and Meta. You know, in 2022, Meta told Wall Street that they were going to lose $10 billion in revenue due to some privacy

[00:07:49] changes that Apple had made for its iOS mobile operating system. And then, you know, in April, Meta was telling advertisers to use this other way of buying ads so that Apple wouldn't get a

[00:08:01] cut of that revenue. These are companies that are always going at it. Just this year, Apple released the Vision Pro headset computing device that puts them in direct competition with Meta and their

[00:08:12] VR devices. These are two companies that are not friends. So at the very least, they took the time to discuss if a partnership would make sense. You know, when Apple Intelligence was unveiled earlier this month at the company's worldwide developer conference, OpenAI's Chat GPT was

[00:08:30] announced as the company's first partner. And I just want to take a minute to say that News Corp, the owner of The Wall Street Journal, has a content licensing partnership with OpenAI. But what other AI companies has Apple been talking to? Based on our reporting,

[00:08:45] Apple has also been talking to AI startups, Anthropic and Perplexity. And we also know that Apple has talked with Google. So certainly the iPhone maker is talking to a lot of the who's who's in the tech industry's AI race. What has Apple said about offering numerous AI options

[00:09:04] to users? Apple's software leader, Craig Federighi, has said that it makes sense for the company to offer users numerous AI options because they could prefer different models for different tasks. And so as these companies build their own models using their own different data sets,

[00:09:23] they could find areas where one model works better than another for a different kind of thing. That was our reporter, Salvador Rodriguez. And that's it for Tech News Briefing. Today's show was produced by Julie Chang. I'm your host, Zoe Thomas. We had additional support this week from

[00:09:39] Melanie Roy. Jessica Fenton and Michael LaValle wrote our theme music. Our supervising producer is Catherine Millsop. Our development producer is Aisha Al-Muslim. Scott Salloway and Chris Zinsley are our deputy editors. And Philana Patterson is The Wall Street Journal's head of

[00:09:54] news audio. We'll be back this afternoon with TNB Tech Minute. Thanks for listening.