The last time hype outpaced substance this aggressively, it was 1999. Today, the term “generative” has become the center of gravity in AI marketing, venture capital, product development, and enterprise strategy. Startups are launching generative slide decks, corporates are racing to add “gen AI” labels to every legacy product, and investors are pouring billions into anything that outputs text, code, or images. The pattern is all too familiar: massive capital, inflated valuations, vague value propositions—and minimal revenue.
We are now in the middle of the “Generative Everything” bubble. And like the dot-com frenzy before it, this wave is setting up both a collapse and a reset.
The Promise of Generative AI
Generative AI refers to machine learning models that can produce original content—text, images, audio, video, or even 3D objects. It’s a real technological breakthrough, opening up possibilities in automation, creativity, communication, and software generation.
But the problem isn’t the technology—it’s the overextension of its promise. Products that barely scrape the surface of generation now claim to be revolutionary. A chatbot that regurgitates a FAQ is labeled “conversational AI.” A document summarizer becomes “gen AI for legal.” A slideshow generator is pitched as “next-gen storytelling.”
Much like in the dot-com era, companies are using the buzzword as a placeholder for an actual business model.
VC Capital Is Pouring In—But Where’s the Value?
Venture capital has again adopted a quantity-over-quality mindset. Rather than carefully evaluating differentiation, many firms are backing multiple startups in the same category—just in case one breaks out. This is how dozens of startups can be simultaneously building generative search engines, AI meeting tools, or design assistants.
The same occurred during the dot-com boom: massive funding of functionally identical startups, each with slightly different branding and little to no path to profitability. The assumption was that scale and hype would compensate for weak fundamentals. Today’s “generative” era is no different.
“Everything is AI” Thinking
In boardrooms and product teams, there’s a growing assumption that every existing tool must now become AI-enhanced. Whether or not customers asked for it, whether or not it improves outcomes, and whether or not it adds real value—products are being retrofitted with generative features just to stay “competitive.”
This is reminiscent of early-2000s businesses slapping “.com” onto their names to attract attention. In many cases, AI features are being bolted onto stable, functioning workflows—introducing complexity and risk while delivering little benefit.
Too Many Tools, Not Enough Problems
A core sign of a tech bubble is when builders chase possibilities before identifying problems. Right now, there’s a flood of generative AI products on the market, and many are solving non-existent issues.
Do users really need AI to generate generic LinkedIn comments? Is there a business case for six different AI resume builders that offer the same templates? Do enterprises benefit from AI-generated meeting summaries that still need manual editing?
As with the early days of the internet, the excitement is overwhelming the need for actual utility. Just because a tool can generate something doesn’t mean that it should—or that anyone will pay for it.
Unsustainable User Economics
Another parallel with the dot-com bubble lies in unsustainable business economics. Many generative AI platforms offer services that are compute-intensive but charge users little or nothing. Startups are burning through capital to provide GPU-heavy services without a clear path to profitability.
The logic is familiar: grow the user base first, monetize later. But monetization is proving elusive. Consumers are already overwhelmed with similar tools, while enterprises are hesitant to pay for outputs they don’t fully trust.
In the dot-com era, it was “eyeballs over dollars.” Now it’s “prompts over profit.”
Platform Dependency and Fragile Moats
Many generative AI startups are built almost entirely on top of large foundation models from a few dominant providers. Their core differentiation often lies only in UI design, workflow integration, or slight model tweaking.
This creates fragile business models. If foundational model providers change pricing, limit API access, or launch competing features, these startups lose their edge instantly.
The same issue plagued early internet companies that relied on unstable partnerships or proprietary platforms. When those environments shifted, entire businesses vanished.
Signs of Saturation Are Already Here
In less than two years, the generative AI landscape has gone from a handful of players to an overcrowded marketplace. Every niche—education, marketing, productivity, coding—now has dozens of generative apps vying for the same audience.
The noise is deafening. Users are struggling to differentiate between products. Feature lists are becoming indistinguishable. Pricing is a race to the bottom.
This is the phase where consolidation, collapse, or correction typically follows.
What Comes After the Pop
Like the dot-com bust, this bubble will burst. Many generative startups will shut down, funding will dry up for “me too” apps, and large enterprises will start scaling back experimental deployments.
But also like the dot-com crash, the survivors will define the next decade. In the early 2000s, most internet startups failed—but the ones that survived became the digital backbone of modern life. The same will happen in AI.
Real value will emerge from companies that move beyond flashy demos and focus on domain-specific applications, robust infrastructure, and measurable outcomes. The winners won’t be the ones yelling “generative” the loudest—they’ll be the ones who can generate results.
The “Generative Everything” moment is real—but it mirrors every classic tech bubble in structure, velocity, and sentiment. Companies are overpromising, VCs are overfunding, and users are underwhelmed.
This isn’t a warning against generative AI. It’s a reality check. The current wave is a noisy prelude to something more stable, more durable, and more meaningful. The crash will come—but what remains will be worth building on.
Leave a Reply