The Speed of an Era
In 1995, the Internet connected roughly 16 million people — less than half a percent of the world's population. Five years later, that number had grown tenfold. It was, by all measures, one of the fastest technology adoptions in history.
Until now.
Generative AI reached a comparable threshold of global adoption in less than two years. OpenAI's ChatGPT hit 100 million users in just eight weeks. According to the St. Louis Federal Reserve, by mid-2024, nearly 40 percent of U.S. adults reported using generative AI, matching the adoption level the Internet took almost a decade to achieve.1 Harvard's analysis puts the technology's growth "at a pace far exceeding that of PCs or the early Internet."2
Generative AI reached the adoption threshold the Internet took nearly a decade to achieve — in under two years. Source: St. Louis Federal Reserve; Visual Capitalist.
The Diffusion Curve Has Tilted
The classic Law of Diffusion of Innovation shows how technologies move from innovators to early adopters to the mass market over time. Historically, each stage stretched over years — sometimes decades.
AI compresses that timeline. The "early adopter" phase that once defined a niche group of technologists now includes educators, analysts, marketers, and small-business owners — millions of people experimenting with daily use cases. We've crossed the threshold where curiosity becomes necessity.4
|
Diffusion no longer measures how long it takes a technology to spread — it measures how quickly markets reorganize around it. |
In diffusion terms, we're somewhere between the late early adopters and the early majority — a rare stage when the technology itself has become widely accessible, and the real difference now lies in how people choose to use it. Organizations that adapt their behaviors, processes, and ways of thinking around AI will move ahead, while those waiting for the "next version" of the technology risk falling behind.5
AI Isn't Replacing Search — It's Re-Wiring Discovery
For business leaders, the question isn't whether AI will shape user behavior — it already has. The question is how quickly our digital infrastructure will adapt to engage and measure it.
Just as Google Search reshaped marketing in the 2000s, AI-driven discovery is quietly redrawing traffic patterns today. Platforms like ChatGPT, Perplexity, and Google's AI Overviews are introducing a new kind of referral: machine-mediated discovery.6
The tell-tale sign? Referral links from AI tools are already appearing in analytics reports. When a visitor clicks a link surfaced through ChatGPT, the URL often includes utm_source=chat.openai or a variation of utm_medium=ai_overview. These parameters may seem small, but they mark the emergence of a new acquisition channel — AI Search.7
Measuring the Invisible
To understand this shift, organizations need to expand how they tag, track, and interpret user journeys. A few immediate steps can help:
utm_source=ChatGPT, utm_medium=AI_Assistant, etc.) so they appear distinctly in analytics.Businesses that establish AI attribution frameworks now will have an empirical view of the shift as it happens. We've spent the past six months helping RIAs build exactly that foundation.
The Verdict: Not Hype, Just Adoption Moving Faster
AI isn't an incremental innovation — it's the next distribution layer of the Internet. It changes how information is found, how trust is built, and how decisions are made.
The Internet's first era rewarded those who understood search.
The next era will reward those who understand discovery through AI.
What was once invisible — search traffic — became measurable and then manageable. The same will happen again, only faster.
The firms that act early, by combining curiosity with rigor, will not just follow the adoption curve. They will define it.
Notes