The launch of ChatGPT in November 2022 triggered a gold rush in the world of AI, sparking dreams of a future where artificial intelligence would revolutionize everything. Tech giants jumped on the bandwagon, pouring billions into chips, data centers, and everything AI. 

Nvidia, the chipmaker at the forefront of the AI revolution, saw its market cap soar past $3 trillion. In June this year, the company briefly surpassed Apple and Microsoft to become the world's most valuable company. 

The pattern has been similar across other AI mega-cap leaders Microsoft, Apple, Tesla, Amazon, Meta, and Alphabet, which, along with Nvidia, climbed 75.71% during 2023, while the broader S&P 500 Index managed “just” 24.23%.

But then the tide turned. Over the past two months, Nvidia shed $900 billion in market value, sparking questions about whether artificial intelligence investor confidence is collapsing. 

Drive down any of the Bay Area's main arteries today, and you'll see nearly every billboard promoting a product "driven by AI." Five years ago, it was "blockchain." Ten years ago, it was "big data." Twenty-five years ago, practically anything followed by ".com." Not all of these technologies lived up to their promise. 

So, are we in an AI bubble then? 

Well, no one knows for sure, but we would argue against it. While the true impact of AI is only beginning to appear in corporate financial statements, several factors give us confidence that the trajectory will be positive and AI will deliver promising capabilities in the long run.

The Hype Cycle Is Natural

Everett Rogers’s diffusion of innovations theory explains the technology adoption life cycle fairly well. Any new technology moves from being first adapted by innovators and early adopters to mainstream users before reaching a point where even those far behind the curve catch up. 

Most firms today also know the perils of crossing the chasm, the gap representing the difficult transition from visionary, early-adopting customers willing to put up with an incomplete product to mainstream customers who demand a more complete product, thanks to Geoffrey Moore.

unnamed (7)-1


However, these theories do not fully capture the complexities of revolutionary technologies like AI, where another dynamic—the hype—comes into play, affecting everything from budgeting and forecasting to startup investments. Coined in 1995 by research firm Gartner, its annual hype cycle aims to assess whether a technology is on the path to productive use or still mired in the smoke-and-mirrors phase.

There are five phases of a hype cycle: the innovation trigger, the peak of inflated expectations, the trough of disillusionment, the slope of enlightenment, and the plateau of productivity (see figure).

unnamed (6)-2Source: ZDNET’s Coverage of the Gartner's 2024 Hype Cycle Forecast

The innovation trigger phase is about building excitement when a new technology shows serious promise and engineers, marketers, and investors see the potential—even though most of that potential is unfulfilled and remains currently unattainable. ChatGPT in 2022 marked this moment for AI. 

Then comes the second phase, the peak of inflated expectations, where media coverage is breathless and overwhelming, entrepreneurs pitch new startups, and marketers refer to the technology in everything they pitch. 

Next comes the trough of disillusionment, where the initial excitement fades, and technology that once generated exuberant hype falters. After relentless promotion with minimal uptake, expectations come crashing down as the reality sets in that the technology isn’t living up to its promises. VR has been in this phase repeatedly. It may again enter the innovation trigger phase should Apple manage a cheaper, lighter Vision Pro. 

Finally, some technologies crawl out of the trough of disillusionment and begin climbing the slope of enlightenment and the plateau of productivity. These two phases refer to when technology begins finding its footing, its specific value propositions are proven, and it enters some level of productive use without the associated hype.

According to Gartner, Generative AI in 2024 has passed the Peak of Inflated Expectations, although hype about it continues. Going forward, more value is expected to be derived from projects based on AI techniques, either stand-alone or combined with standardized processes to aid implementation. 

While the days when an AI start-up could secure billion-dollar valuations with just a PowerPoint presentation of its 'vision' may be over, this winding down of AI hype marks not the end but the beginning of a more serious and focused era of AI innovation.

This Is Not Dotcom 2.0

Economic bubbles form when the price of an asset (real estate, stocks, or something else) rises far beyond its actual value. When investors see momentum, they start pouring money into it, driving prices up. This creates a bandwagon effect, with more people jumping in, convinced that prices will keep climbing. But eventually, they realize they’ve paid way too much for something that isn’t worth it. Panic selling kicks in, and the herd mentality that drove prices up now sends them crashing down. 

unnamed (5)-2Lifecycle of an economic bubble (Source: Listen Money Matters)

On the surface, the current AI situation may look like the dot-com bubble, but there is an important difference. 

During the dot-com bubble, companies’ stocks soared on nothing but hype; even firms with no real business models reached billion-dollar valuations. Crazy valuations and unprofitable business models were justified by the new business models made possible by the Internet. 

For example, the grocery delivery service Webvan, which went public in late 1999, incurred over $27 in costs per order while charging customers no more than $10. Similar stories unfolded with Pets.com, eToys.com, Flooz.com, and Excite@Home. The common thread was clear: dot-com companies were soaring on unrealistic expectations and unproven business models—growing fast, breaking things, but failing to generate profits.

With AI, the story isn’t quite the same.

The stocks that fell recently belong to tech giants like Microsoft, Apple, Tesla, Amazon, Meta, Alphabet, and Nvidia—companies with well-established business models. Unlike the dot-com era, AI is generating revenue for them. For instance, over half of Fortune 500 companies today use Microsoft's Azure AI models, significantly boosting Microsoft's cloud business revenue.

Other companies are benefiting, too. Arm, a UK-based chip designer whose technology is used in smartphones and PCs, saw AI demand fuel its revenue. Data management company Palantir also witnessed similar growth driven by AI. 

Like the Internet (which people were skeptical about), AI has immense potential, but realizing its full benefits will take time and patience. According to a McKinsey survey, most executives expect it will take three to five years to capture the full value from their AI investments. 

As AI moves from providing point solutions to automating complete workflows, it will likely unlock tremendous value, potentially adding $15.7 trillion to the global economy by 2030. 

Moving From Proven Capabilities To Real Returns

While the possibility of creating Artificial General Intelligence (AGI)—a machine capable of performing any human intellectual task—may seem far-fetched, current AI models already excel at sifting through massive data sets, identifying patterns, automating workflows, and generating human-quality text.

Deloitte and IBM explored the value of output created by Generative AI among more than 2,800 business leaders. Here are some areas where organizations are witnessing a return on their AI  investments:

  • Text (83%): Gen AI assists with automating tasks like report writing, document summarization, and marketing copy generation.
  • Code (62%): Gen AI helps developers write code more efficiently and with fewer errors.
  • Audio (56%): Gen AI call centers with realistic audio assist customers and employees.
  • Image (55%): Gen AI can simulate a product's appearance in a customer’s home or reconstruct an accident scene to assess insurance claims and liability.
  • Video generation (36%) and 3D model generation (26%) can create marketing materials, virtual renderings, and product mockups.

Another independent IBM study also shows that AI delivered a higher average return on investment (ROI) in 2023 than in 2022. Organizations have achieved several benefits through their generative AI initiatives:

unnamed (4)-2

Source: Deloitte, Moving from potential to performance

McKinsey research estimates that Generative AI alone could add between $2.6 trillion and $4.4 trillion to the global economy annually while increasing the impact of all AI by 15 to 40%. More than 100 generative AI use cases across seven business domains have been identified in technology, media, and telecommunications alone. These new-gen AI use cases could unleash between $80 to $130 billion in the media industry annually. 

In his insights on AI, David Cahn of Sequoia Capital introduces what he terms the “$600B Question”—the gap between building AI infrastructure and actually generating revenue. The problem? While massive AI capabilities exist, companies aren’t ready to implement them effectively, which is why many AI investments are falling flat. The real opportunity isn’t creating more AI capacity; it’s helping companies break through the barriers to effective implementation. When they do, the results will follow.

Views from Business Leaders

Business leaders are divided on the AI outlook. Some report that they already see a return on their AI investments and plan to become increasingly bullish, yet others remain skeptical. 

JPMorgan CEO Jamie Dimon says AI is not just hype and plans to add thousands of jobs focused on AI in the next few years. His company already employs 2,000 people focused on AI.

Earlier this year, Ursula von der Leyen, President of the European Commission, emphasized AI's potential to significantly boost productivity and revolutionize sectors like healthcare. She stated, "First movers will be rewarded, and the global race is already on without any question."

Chris Howard, Gartner's chief of research, addressed the Hype Cycle for Artificial Intelligence, emphasizing that AI remains a dominant topic. He clarified that the "trough of disillusionment" phase that AI is expected to enter is not a dark or dangerous period but rather a crucial stage for determining what works, what doesn't, and where the real work and dependencies lie.

Jeff McMillan, head of firmwide artificial intelligence at Morgan Stanley, says that their financial advisors continue to see the real benefits GenAI delivers to their practices and that the investment bank is just starting to unlock the true power of AI.

At the other extreme is Gary Marcus, a professor emeritus at New York University and a long-time AI researcher, who predicts that the AI bubble could collapse within days or weeks if not months.

DeepMind CEO Demis Hassabis also belongs to this camp. In an interview with the Financial Times, Hassabis remarked, “In a way, AI’s not hyped enough, but in some senses, it’s too hyped…We’re talking about all sorts of things that are just not real.”

AI Investments Will Continue

Capital spending by technology mega-caps is exploding, primarily because management teams believe underinvesting in AI is a greater risk than overinvesting. Consequently, Microsoft is expected to spend roughly $73 billion in the calendar year 2024, followed by Amazon (nearly $70 billion), Alphabet Inc. ($50 billion) and Meta Platforms (just shy of $40 billion). 

While not all of this spending is on AI, this aggregate outlay of about $230 billion in one year is a huge change from roughly $100 billion spent in 2020. 

Going forward, investors will closely scrutinize Return on AI Investment. The value of AI will be unlocked in two key areas: companies already benefiting from AI and those aiming to integrate it into their future business strategies.

Chip companies such as Nvidia and tech companies like Microsoft, Amazon, and Alphabet are near-term beneficiaries. Still, it is unlikely that these would recover their $230 billion worth of capital spending in 2024 (and likely even higher spending in 2025) soon. 

However, significant value remains to be unlocked in the second category of companies—those planning for AI but not yet profiting from it. According to a new EY survey, among senior leaders at organizations that invest in AI, about half (51%) admit that three years ago, they spent less than 5% of their total budgets on AI investments. Today, 88% of those same leaders spend 5% or more, with the number set to grow even higher.

Going forward, investors will also focus on value creation in healthcare, finance, retail, manufacturing, energy, publishing, and other sectors as firms integrate AI into their future strategies.

What This Means For Digital Media Publishing

Several publishers have been exploring AI to streamline their operations. Last fall, BDG, BuzzFeed, and Trusted Media Brands tested AI copilots and private chatbots to make their sales organizations more efficient and productive. Time and The Wall Street Journal use AI to improve efficiencies and speed up the rate of doing business. 

Earlier this month, we made the case for why digital media publishers should seize the moment to invest in AI. Foundational AI models have become more powerful, accurate, and accessible, while the cost of deploying them has dropped significantly amidst an expanding array of use cases.

The latest models, such as OpenAI’s Sora and Meta’s SAM2, expand AI capabilities beyond text, allowing publishers to leverage the power of more engaging formats like video. For example, Aeon’s Generative AI-based text-to-video solution today enables publishers to:

We are witnessing a transition from the age of AI hype to the age of practical and tactical AI, where innovative publishers, focused on real-world impact rather than buzz, can build a future where they enjoy a lasting competitive edge.