Why Generative AI Is a Bigger Threat to Apple Than Google or Amazon

The conversation over the last three years about the shift to digital has been largely about the share of consumers who once did most things in the physical world and who now do more of those same things online.

The more interesting conversation is how this shift to digital has reframed the competitive landscape.

New tech has moved the application of digital beyond smartphones and apps to new ecosystems that blend the physical and digital worlds and can be accessed by a variety of connected devices, including cars and voice-activated speakers and earbuds. Many traditional industry sectors are being disrupted by these digital competitors, some new and some old, who appear to have come out of nowhere to take share — and grow it.

For instance, airline CEOs never gave a minute’s thought to Zoom, Teams and similar apps as competitors for the business traveler’s dollar until client, prospect and staff meetings shifted almost exclusively to online video channels during COVID. Those channels remain a CFO-friendly alternative to frequently hopping on a plane to visit clients or staff in person, reducing both business travel expenses and airline profits.

Car OEMs never thought of Amazon, Apple, or Google as competitors until they saw the opportunity to monetize the cockpit and the user experience while in the car and decided they should create their own car-centric apps and ecosystems to fight back.

Grocers used to think of Whole Foods as their only Amazon grocery competition — and with its small share and Whole Paycheck reputation, they didn’t worry too much. Until Amazon Subscribe & Save put a noticeable dent in center-store sales, and now has 7.1% of U.S. consumers ordering products and having them delivered for free this way, according to PYMNTS data.

Then there’s OpenAI’s ChatGPT.

Its public launch with GPT in November is a disruption like no other we’ve seen in the modern internet age. Unlike the digital competitors who may have been hiding in plain sight, slowly chipping away at traditional players’ market share, ChatGPT came out of nowhere, it seemed, with the potential to change the world.

In a matter of six months, GPT has monopolized nearly every conversation among CEOs, boards and investors about whether it’s a threat, an opportunity to fast-track a new way to do business, or both.

Ever since, news reports suggest that Google and Amazon have pinned their respective GPT-anxiety meters now that OpenAI, with Microsoft’s support, has created a new way to build and scale computing’s next big thing.

I’m not so sure.

For one thing, six months since OpenAI’s debut is much too early to predict winners and losers, especially since we know that both Google and Amazon have invested heavily in AI and LLM for many years.

For another, I don’t think that winners — at least not the Big Tech winners — will be crowned because they are successful in disrupting search, but because they have been successful in disrupting how search is monetized using large language models.

Winners won’t try to force-fit the old cost-per-click model into these new innovative content platforms, but create new models that embed commerce into search and monetize that conversion.

It’s a model that potentially favors Amazon and Google more so than Microsoft, given the existing scale of each of their platforms to consumers and businesses globally — the hundreds of millions of tokenized and embedded payments credentials that can enable a transaction and complete a sale today and their cross-platform access to the billions of consumers with smartphones.

And one that puts Apple at risk.

Cupertino’s stance on privacy and user data, and its closed ecosystem, may have Apple winning the hearts and minds of consumers but losing the Generative AI war, despite having a 58% share of the U.S. smartphone market to Android’s 42%. Unless something changes, Apple may end up playing the role of smartphone, mobile OS and App Store, just like it does today — with an App Store that comes chock-full of everyone else’s GPT app innovations.

It’s already happening.  OpenAI’s GPT app for iPhone users is now available in the App Store.

Everyone’s 15 Minutes of GPT Fame

You know the ChatGPT story well by now.

UBS says it’s the fastest growing consumer application of all time, capturing 100 million users in just two months with 28 million visits a day to its site by the end of January. Pitchbook reports that in Q1 2023, $1.7 billion was invested in 46 Generative AI startups, with another $10.68 billion announced but not yet closed. VCs report that seed rounds for GPT startups can go as high as $15 million and start at $5 million. Those are very big seeds.

Countless press releases tout the various applications of GPT tech to businesses — PR newswire says it has published about 150,000 of them since January. The CE 100 Index, the PYMNTS stock index of the 100 listed companies that are leading the transition to a more digital, connected economy, has seen a boost as companies announce the integration of GPT models into their businesses and the market takes notice. Personally, I have never seen more announcements about company products that start with the prefix “AI” — even for companies that I know that have always used AI as part of their tech.

AI, especially Generative AI, has become the ticket for everyone’s 15 minutes of fame.

The Big Tech GPT Landgrab — Google

Nowhere has the GPT spotlight shone brighter than on Google and Amazon. Microsoft’s $10 billion investment in OpenAI immediately set off news accounts about Bing finally being able to gain share in search — and the advertising dollars it generates.

Whether that’s wishful thinking or reality remains to be seen.

ChatGPT isn’t exactly Microsoft’s first attempt to use AI to make Bing more competitive. It has invested billions in Bing since its launch in 2009. There’s not much to show for any of those dollars spent — at least not so far. Google’s search share in the U.S. remains almost 15 times that of Bing’s (Google at 89% to Bing’s 6.3%).

For Microsoft and Bing to win, they must convince users — tons of them, and pretty soon — that it is better, and then they must show merchants that they can drive more sales their way. And then convince Apple and Samsung to ditch Google for Bing to power their search. That’s probably not happening anytime soon.

That said, Google has cause for concern when it comes to search and and how to monetize it — which is its giant cash machine today, and largely the result of advertising.  Q1 2023 results showed a continued decline in ad revenue from search as the combination of Apple’s privacy blockers and economic conditions throttled results. Google, it was reported, “nervously” rushed out an announcement for Bard, its Generative AI platform, on February 6, 2023 to arrest market concerns that Microsoft might out-search the search giant. Of course, this wasn’t really news — Bard has been in development at Google since 2015.

Understanding Google’s plans for commerce using Generative AI is perhaps even more fundamentally important than how it will use Bard to make search more efficient, authentic and reliable. Google has been more cautious in its public statements about Generative AI to manage expectations while Bard gets smarter and better behaved.

But the Generative AI ticket that Google needs to punch is getting people to start more of their product searches there. Google may have more search traffic overall than Bing — but depending on whose source you believe, anywhere from half to 65 percent of searches for products start not on Google, but on Amazon.

That’s not news or a new fight for Google. Nor are its aspirations to become a payments and commerce force. Over the years, Google’s focus has been on using Chrome to make checkout easier with form fill and capturing and storing tokenized payments credentials there to expedite online checkout. It’s been less successful in monetizing the GPay and Chrome ecosystem in other, more proactive ways.

The story that’s yet to be written is how Bard and its Generative AI models are pointed at moving Google away from an ad-based search platform to a contextual, commerce driven ecosystem — how Google will use Android to distribute Bard and its Generative AI derivatives, and how Cloud will embed those capabilities into other applications and use cases.

The Big Tech GPT Landgrab — Amazon

Then there’s Amazon. The Economist wrote in 2019 that AI has been a core element of Amazon’s strategy since 1999, when Jeff Wilkie joined the firm. The strategy then was to use data and AI to improve Amazon’s operations and logistics to the benefit of the consumer. Of course, no one would recognize what they were using back then as AI, given the rapid development in that area. But AI in some form and voice are the DNA of its Alexa voice operating system. According to a study of nearly 3,000 U.S. consumers conducted by PYMNTS in April, 60 percent of U.S. consumers want Alexa to become even smarter as the idea of a smart virtual assistant to simplify the complexity of the daily grind becomes more appealing.

Also in April, Amazon announced the rollout of a suite of Generative AI tools.  In addition to using Generative AI to make product searches on the Amazon marketplace more efficient, Amazon is embedding Generative AI tools into AWS so that businesses can build and scale their own GPT use cases using its Bedrock platform. Amazon’s expansion into other connected ecosystems like healthcare, grocery, restaurant delivery, streaming services and gaming (remember, it owns Twitch) gives Amazon a number of adjacent ecosystems and content to enable with its Generative AI capabilities.

More recently, Amazon announced that its AI-powered voice assistant, Alexa, has taken a step closer to being Alexa everywhere a consumer wants to take her. In addition to new Echo devices with and without screens suitable for home use, Alexa is now available in earbuds, glasses (not smart glasses, but those with sensors that allow a noise cancelation conversation with Alexa), and a USB plug that takes Alexa inside the car, whether or not the car OS is powered by Alexa.

Amazon Pay is both a payments and identity credential that powers conversion on the Amazon platform, and increasingly off it, as more merchants participate in its Buy with Prime program. Amazon Pay has reinvented store checkout with biometrics and no-checkout checkout in its smaller-format grocery stores.

As a commerce platform accessible to anyone with a smartphone and the Amazon app, it’s easy to declare Amazon the player with a running head start to connect search with commerce, since that’s its business today.  Unlike Microsoft, Google and Apple, it is a commerce platform with embedded credentials and a ton of data that it can use to make its models smarter and results more relevant to the user. Even AWS, which reported slowing growth, is likely to have a new lease on life as companies clamor to leverage its Bedrock platform and other Generative AI tools to embed LLM into new experiences for their end users.

What Amazon doesn’t have is any control over the mobile operating systems and handset OEMs that host its app — and neither does Microsoft. Amazon’s Fire Phone was a disaster, mostly because it wasn’t better than — or even as good as — existing products that users liked and used. The decision to fork Android made it more challenging for developers who needed to create three versions of their apps, instead of just two — and for a device that had few users. The user/app developer virtuous circle never got off the ground.

Unlike Apple’s iOS and Google’s Android, both of which can (pretty much) force their apps on the phone because they own the OSs, Amazon depends on users and user downloads to claim real estate on those home screens and dictate how it is used. That hasn’t been much of a problem, so far — but it could become one if Apple and Android decide to use Generative AI to impose new rules that govern how they appear and how much it may cost to operate there.

The Big Tech GPT Landgrab — Apple

Then there’s Apple. We have three data points on Apple’s current market position and thinking on Generative AI.

First, during Apple’s Q2 2023 earnings call earlier in May, there were the usual vagaries when asked about Generative AI and Apple’s plans. Analysts didn’t probe further when told by Tim Cook that Apple is “thoughtfully” considering the application of Generative AI and won’t comment on product roadmaps.

Second, it’s been reported that Apple has posted 100 job openings for engineers with Generative AI experience. Although we don’t have any definitive word from Apple HQ on Generative AI, it seems that — of the current crop of Big Tech players — Apple may be the one scrambling the most because they are the furthest behind.

Three, Apple seems fine with having Generative AI apps in the app store.

It’s possible that one of the reasons that we may not have heard much from Apple on the topic is because it is a tough one for Apple to navigate.

Apple is no AI novice or lightweight. It’s neural network uses AI — and its own AI chip — to improve the iPhones’ computational efficiencies so that users have a better experience. But for Apple to compete in the Generative AI space, it needs loads of data. Apple has made a point of stating that it only collects enough data about a user to maintain their accounts. So, it doesn’t have much today, other than what people search for in the App store. And that’s not much.

Collecting more data isn’t straightforward. Apple would have to figure out a way to thread the needle, balancing the image of the platform that protects users’ privacy and doesn’t collect and monetize their data with a desire to be a player on the Generative AI stage. And to be able to do that on a global scale, where iOS share is dwarfed by Android.

Also not straightforward: getting consumers to think of — and use — Apple as a commerce platform, which also includes using Apple Pay to pay for purchases. The latter has been a struggle for Apple both on and offline. Going on almost nine years after its launch, PYMNTS data shows that cards at checkout in the store are used almost 31 times more than Apple Pay, which accounts for a tiny 2.1 percent of retail sales. Apple Pay’s growth over that time has averaged about .25 basis points per year.

It’s also not obvious that consumers think of Apple as a commerce platform, even though they use the iPhone to access apps that connect them to commerce. As commerce becomes more distributed across connected endpoints that cross platforms and operating systems, smartphones will become one of the ways — but not the only way — in which consumers interact with apps and ecosystems. Those interactions could even become device-agnostic over time — where Apple, as a closed ecosystem, has built an inherent disadvantage.

It’s tricky.  It may mean that Apple is relegated, perhaps not happily, to playing host to GPT apps in the app store — at least for the foreseeable future. And maybe being the de facto regulator to make sure harmful GPT apps don’t get loose.

What’s Next

Many of you reading this have played around with GPT and are amazed by its power and potential. Me, too. You may even be a little bit scared of how many industries, including your own, will be changed when these large language models are more refined.

I decided to use it to help with this piece. Not in writing it — which would have saved my Sunday and produced something not nearly as brilliant 😊— but in the image that I used to accompany this piece.

I put into DALL-E a few prompts: disruption, technology, change, Big Tech, future. What you see is the best of what was generated after several iterations — but at least it only cost me only a few cents and a few minutes to produce. I know it’s not the greatest image.

Most interesting about the selection of results was that most of the images depicted disruption as rearranging the letters used to spell the word disruption. Disappointing, since rearranging words and spelling seems an elementary interpretation of the concept — and a rather underwhelming way to depict the future of a technology that holds such great potential.