Industry observers note that A cash cow. Dave Rowland/Getty Images/Getty Images AsiaPac hide caption.
Google became one of the most valuable companies in the world by revolutionizing how humanity finds information. Now, a new generation of AI rivals is challenging its dominance and pushing Google to spend a fortune to secure its future. Can Google embrace this disruptive technology without breaking the business that made it rich?
Former Google executive Marissa Mayer remembers when the organization was just a scrappy startup, working out of a small office above a bike shop in Palo Alto, Calif. She joined the company back in 1999, fresh out of nearby Stanford with a master's in computer science.
"βIt's pretty wild. When I interviewed at Google, there were just seven people there," Mayer tells Planet Money. By the time she got hired, a month or two later, she became its 20th employee.
Sources indicate that then Google vice president of search Marissa Mayer in 2009. ROBYN BECK/AFP via Getty Images/AFP hide caption.
Recent findings suggest that that year, Google brought in just over $200,000 in annual revenue. By 2022, the eve of the AI boom, the tech giant β by then reorganized under a parent company called Alphabet β was generating more than $280 billion. That's more than a millionfold increase.
Recent findings suggest that google made its fortune selling ads next to its search results. It monopolized internet searches so thoroughly that its name became a verb. And its search business became like an unstoppable money machine.
According to recent reports, in 2012, Mayer left and became the CEO of Yahoo, trying to revive a competitor that Google had blown out of the water. But neither Yahoo nor any other rival managed to dethrone King Google. The company's dominion just kept expanding. Its share of the U.S. search-engine market grew from 80% in 2009 to nearly 90% in 2020.
For a time, it seemed like the only forces that could cut this juggernaut down to size were politicians or antitrust lawyers.
Recent findings suggest that but the craziest part? The biggest threat to Google was maybe coming from inside the house! From what it was developing in its own labs β and the industrial race it would unleash.
In a recent development, google plowed the profits from its search business into research and development of AI. And, in 2017, a team of eight Google employees developed the transformer, a breakthrough technology that made today's AI chatbots possible.
In a recent development, mayer, who is now CEO of the AI startup Dazzle and remains a Google shareholder, sees Google's pioneering AI work as a reason for optimism about the organization's future. "This is not a technology that's happening to Google," she says. "They helped invent this fresh technology.".
Industry observers note that yet, today, Google is spending enormous amounts to build its AI capabilities and fend off rivals. And it's overhauling its search business, using AI to answer questions directly instead of just providing links.
In a recent development, wall Street is getting nervous. In late October, Alphabet released its most recent earnings report. The company's stock price plummeted more than 12% the next day β the biggest single-day percentage decline since the company went public in 2004.
If this whole AI transition goes sideways for the organization, it's a bit like a mother giving birth to a baby and then that baby slapping her in the face and stealing money from her wallet.
Which brings us to⦠a classic business theory we've been thinking about recently: the innovator's dilemma.
The late Harvard Business School professor Clayton Christensen developed the theory in his 1997 book, The Innovator's Dilemma: When New Technologies Cause Great Firms to Fail.
Christensen theorized that the very thing that makes a business wildly successful can lead to its downfall in the face of a "disruptive" technology. He helped put disruption in Silicon Valley's vocabulary, long before it became a clichΓ©.
Recent findings suggest that the core innovator's dilemma is basically this: embrace a disruptive technology, and risk the profits you're making today. But if you fail to embrace it and let others run with it, you may end up risking your tech giant's health tomorrow.
Recent findings suggest that today in the Planet Money newsletter we ask: has Google been caught in the innovator's dilemma? And can it pivot to a upcoming technological era without sacrificing its enormous profits?
Recent findings suggest that google didn't just invent the transformer. Over the decades, its researchers made computers translate languages, recognize speech and images, drive cars autonomously, and even beat one of the world's best players at the game Go.
The tech giant even developed an internal chatbot called Meena that employees played with years before AI chatbots became a sensation.
Industry observers note that then, on Nov. 30, 2022, OpenAI released ChatGPT and offered a completely latest way to find information and a glimpse of a future where people didn't need to go to Google to find answers anymore.
Google scrambled to respond, reportedly declaring a "code red" as ChatGPT captivated the masses.
So, first question, why wasn't Google the first out of the gate with something like ChatGPT?
It has been revealed that christensen's theory offers one possible explanation: When you're making tons of money from your existing business in search, why rush to release something that could undermine it? Why disrupt yourself?
His theory wasn't just that companies seek to protect their cash cow. It was that a manufacturer can get too focused on the best way to nurture and milk that cash cow. Like, dang, I love this cow. Then when a disruptive technology comes along, they might be like, wait, this won't help me milk my beautiful cow. No thanks.
In other words, a manufacturer's focus on its core business can lead it to not see a disruptive technology's market potential. It can create a kind of myopia.
It has been revealed that last year, Google's chief scientist, Jeff Dean, offered an explanation for why Google hadn't released something like ChatGPT soonerβ and it suggested the company had fallen victim to the kind of myopia Christensen warned about. Dean told podcaster Dwarkesh Patel that they looked at their existing chatbots through the lens of their search business. You know, their cow.
Sources indicate that "From a search perspective, you want to get the right answer 100% of the time," Dean remarked. "These models were not near that bar." They were prone to hallucinating and also saying offensive things.
According to recent reports, those were problems ChatGPT had as well. It made tons of mistakes and stated racist, sexist, and other incendiary things that made headlines. Yet it took off like a rocket ship when they released it.
Recent findings suggest that and it turned out that people didn't just use ChatGPT as a way to find information or get answers. Dean acknowledged that Google had underestimated how useful chatbots could be for things people wouldn't necessarily turn to a search engine for, like drafting a note or summarizing text.
In a recent development, christensen warned about this very trap in The Innovator's Dilemma. Companies that keep disruptive technologies "bottled up in their labs," he wrote, risk losing out to rivals willing to run with them, however imperfect they may be.
Mayer, however, questions whether releasing a chatbot first would have gone as well for Google as it did for OpenAI. "The standards for an established company like Google in terms of what customers expect, in terms of quality, how good the search is, how great the product is fundamentally different than for an upstart or a newcomer," Mayer says.
According to recent reports, we don't know what would have happened if Google had released a chatbot first. Would a Google chatbot have taken off in the same way, giving Google a first-mover advantage? Or would its flaws have created major headaches for a tech giant with an established reputation to protect?
Whatever explains Google's initial hesitation, it's now going all in. It's fought back with its own chatbot, now called Gemini, and by putting AI-generated answers directly onto its search page, even at the risk of undermining the advertising business that made it rich.
The question now is whether all this will leave Google as dominant β and as profitable β in the AI age as it was in the age of old-school web search.
Benedict Evans is a widely read technology analyst who was previously a partner at the Silicon Valley venture capital firm Andreessen Horowitz.
It has been revealed that he states that every 10 or 15 years or so, a revolutionary "platform shift" shakes up the tech industry, giving newcomers a chance to challenge dominant companies. It happened with the rise of personal computers, then the internet, then smartphones. Now, he says, it's happening with AI.
"The whole table gets reset and the things that meant that they had won and you couldn't take them on suddenly don't matter anymore," Evans says.
Industry observers note that benedict Evans speaks on stage in 2018 in Oxfordshire, England. John Phillips/Getty Images/Getty Images Europe hide caption.
For Google, the consumer pivot to AI could mean it loses some of the competitive advantages that helped it monopolize search.
It has been revealed that the process toward monopolization began with Google building a better search engine than rivals. Then, Evans indicates, its success became self-reinforcing: more users gave Google more data to improve its results, helping it attract still more users. Economists call this a "network effect," a kind of positive feedback loop that helped Google cement its dominance.
It has been revealed that of course, Google also actively worked to catch and keep customers as well. It, for instance, paid to make its search engine the default in browsers and on devices β deals that a federal judge ruled in 2024 had illegally helped maintain its search monopoly.
In a recent development, google also cemented its dominance by weaving itself into our lives. It runs the operating system on many of our phones. We use free Google services to send emails, write documents, find directions, and watch a stranger explain how to fix our washing machine. Many of us keep going back to Google search out of sheer habit or because, when we're navigating the Google digital complex, we keep getting nudged to go there.
It has been revealed that but a platform shift could loosen Google's grip on us. Why go to Google dotcom to search for information if you're using an AI system made by a rival that can answer your question well enough? And, maybe, those rival AI systems will borrow from Google's playbook and work to intertwine themselves into our lives.
It's still early days in this rapidly transforming market. One massive question, Evans claims, is whether AI models will develop winner-take-all dynamics like those that helped Google dominate search. So far, it seems like the frontier labs keep catching up with each other. One possibility, he says, is that AI could become increasingly capable of improving itself β what those in the biz call "recursive self-improvement" β allowing whichever company gets there first to build a lead that rivals struggle to overcome. But whether the AI race will end with a dominant leader β and whether Google will be the one that emerges victorious β is still a huge question.
Sources indicate that mayer doesn't expect a single winner. Yes, the mammoth cost of the computing power and energy needed to develop and run frontier models creates a barrier to entry, making it hard for smaller companies to compete. But she believes there's room for several giants. "There will be multiple big players," she says. "I don't think there will be just one.".
Recent findings suggest that google has a really good shot at remaining one of the big players, but it could lose the dominance it has enjoyed in search. As Evans puts it, "Google's share of search has really got only one way to go.".
Industry observers note that but losing market share doesn't necessarily mean making less money, Mayer argues. She expects AI to attract more people looking for answers and draw more advertising dollars into the market. Sure, she notes, Google's share could get smaller, but the pie could get bigger.
Evans points to Microsoft as an example of how a organization can lose dominance after a technological disruption but still keep growing. When high-speed internet took the world by storm, it opened up new ways for rivals to compete with Microsoft in software. Companies like, yes, Google developed products like Docs and Sheets that people could use in a web browser, whether or not they were running Windows. Controlling the dominant operating system no longer gave Microsoft the same advantages in the internet era.
In a recent development, but the internet also gave more people a reason to buy computers β and many bought PCs that ran Windows. Microsoft sold more copies of Windows and Office, and its business kept growing because of this and other reasons.
It has been revealed that "Well, go look at Microsoft," he claims. "Microsoft lost, but at the same time, the market became so much bigger" that "the shareholders got a lot more money.".
That noted, Microsoft shareholders needed some patience. Its stock price didn't surpass its dotcom-era peak until nearly 17 years later.
Beyond just the threat to Google's monopoly in search, AI is also challenging the advertising model that made search so lucrative.
Google's historic stock plunge in late October suggested that investors are concerned about this.
In a recent development, on the organization's earnings call, executives acknowledged that the growing use of "AI Mode," its chatbot-style search feature, was contributing to slower growth of search advertising revenue because it earned less per query than traditional search. Meanwhile, Alphabet stated it expects to spend a whopping $200 billion this year alone on infrastructure, including data centers, largely to support its AI push.
Watching Google's stock price plummet that day, it felt like seeing Google struggle with the innovator's dilemma in real time. Like it was willing to risk cannibalizing its core business to feed its ambitions in the AI age, accepting short-term pain in the hope of long-term gain.
Sundar Pichai, the CEO of Google and its parent company Alphabet, has made clear he believes embracing innovation is the only way forward. He was actually asked about the innovator's dilemma on the All-In podcast last year. "I think the dilemma only exists if you treat it as a dilemma," Pichai stated. "All along in technology you have these massive periods of innovation and you lean into it as hard as you can. It's the only way to do it.".
Pichai Sundararajan, known as Sundar Pichai, CEO of Google Inc. speaks during an event to introduce Google Pixel phone and other Google products on Oct. 4, 2016, in San Francisco. The Google Pixel is intended to challenge the Apple iPhone in the premium smartphone category. (Photo by Ramin Talaie/Getty Images) Ramin Talaie/Getty Images/Getty Images North America hide caption.
Mayer is optimistic about Pichai's aggressive approach and Google's position in this fresh technological era. Google helped pioneer this technology, she says, and using it to answer people's questions draws on what has been its core expertise. She even questions whether there is really any dilemma for Google when it comes to using AI to give consumers better answers.
"They actually aren't doing something upcoming by deploying it," Mayer says. "They're just answering questions differently and arguably better.".
Recent findings suggest that she does, however, acknowledge there is a dilemma when it comes to advertisers, who pay most of the company's bills. Google may need to fundamentally revamp its system of clickable ads it has spent decades showing alongside search results.
Sources indicate that "βSo I think that piece of it is hard," Mayer notes. "How do I take that advertising spend and move it to a new format that keeps my customers, my advertising customers, as happy?".
But advertisers will follow consumers, she argues, and AI could provide the basis for an even more lucrative business.
Industry observers note that "I don't see why it can't be as profitable," she claims. "In fact, I think it probably, in the end, will be more lucrative than it has been in the past.".
In a recent development, who knows what advertising will look like as the AI era advances. Maybe we'll see familiar-looking banner ads or clickable links on chatbot screens. Or maybe we'll be confiding in an AI assistant with a disembodied voice, like Scarlett Johansson in Her, sharing our most intimate thoughts, only to hear a word from our sponsor. Awkward.
Sources indicate that mayer sees other possibilities. Her startup, Dazzle, uses AI to analyze people's photographs. Those photos can help the system learn about you β which brands you like, what products you own, and what you might want to buy next. You can also snap a photo of something you like and ask where to buy it.
"If we can identify the product and send you there," she claims, "there's an opportunity for advertisements or lead generation.".
The more people share with AI systems, the more those systems learn about them, creating valuable personalized data that could be used to microtarget ads or nudge people to buy things.
For Google, the question is whether AI's newly introduced ways of connecting people to products can replace β or exceed β the profits of its existing advertising business.
In a statement to Planet Money, Google mentioned its AI features are already leading to more searches on its platform. "This past quarter, Search queries grew at their fastest year-over-year rate in five years," the company said. It states they're taking a "deliberate approach" to developing ad revenue in this upcoming technological era: "building the best user experience first and then scaling monetization.".
But the AI race is about much more than just building systems that can help people find information and make money from ads along the way. Investors are dreaming of AI models that can do things that once sounded like science fiction, including automating vast amounts of knowledge work. They're betting revolutionary applications will give rise to incredibly profitable businesses.
Evans believes the most exciting market opportunities will be found in entirely fresh uses for AI β some we haven't even dreamed up yet.
Recent findings suggest that what will those applications be? How profitable will they become? Will they justify the mind-blowing investments Google and other companies are making in this technological revolution?
Evans states the uncertainty around fundamental questions like these helps explain stock market volatility. Investors don't know how much money these companies will ultimately earn β or how long they'll have to wait. And even if AI creates enormous wealth, he says, it's far from clear who will capture it.
One risk Evans sees is that competing AI models become so similar in their capabilities that their developers struggle to charge much for them. The most lucrative businesses could be the products and services other companies build using those models.
He points to cellphone service providers as a cautionary tale. They spent billions to build out the network infrastructure for the mobile internet, allowing people to watch videos, order taxis, and much more on their cellphones.
"All the stuff that they were talking about happened," Evans indicates, "but it happened 10 years later.".
Sources indicate that and when that future finally arrived, the service providers weren't the major winners from all that spending. Ouch! Yes, they owned the infrastructure, but it was the companies making smartphones and building apps and services on top of that infrastructure that captured the biggest rewards of the mobile revolution.
Evans suggests that the same risk hangs over the companies building today's AI models. "The point is just because it's amazing and cool and exciting and it's going to change the world doesn't mean that you have value capture or pricing power." In other words, AI could change the world without sending these companies' profits β or stock prices β soaring.
In a recent development, of course, the fate that befell the cellphone service providers is just one possible scenario. Evans says the AI industry could develop like advanced chip manufacturing, where the rising cost and complexity of making chips left fewer and fewer companies able to compete. Or, he says, it could look like cloud computing, where several significant companies make good money supplying the infrastructure that other businesses run on.
Industry observers note that google has a good shot at thriving under more than one of those scenarios. Its empire spans search, AI chatbots and assistants, YouTube, cloud computing, and even driverless cars through Waymo.
Recent findings suggest that google can make money both from supplying AI infrastructure and building products and services with AI, Evans indicates. On one side, Google already sells access to its own AI models and supplies computing power to other companies, including Anthropic. Google is also a major investor in Anthropic, and has a stake that could become even more valuable when that company goes public.
According to recent reports, but if the business of building and running frontier AI models turns out to be less profitable than so many investors hope, Google could still make money using the technology in its own products. It also has the money and in-house expertise to compete for entirely fresh AI businesses.
The question is whether it can make enough money from all this to make its huge investments worthwhile β and make up for whatever it might lose in search.
Google helped invent the technology behind this revolution. Now it's betting a fortune that AI can deliver a whole newly introduced herd of cash cows β without killing the one that made it rich.
NPR does not offer or accept money for coverage or interviews. Anyone who makes such an offer or request is not an NPR employee and does not represent NPR.
