Charles Esther
“We’re on AI time,” is the phrase I now use to explain how fast artificial intelligence and related technologies are advancing.
Even though OpenAI launched ChatGPT only two and a half years ago, I’ve sensed for months that technology is no longer following Moore’s Law, which said the number of transistors on a chip doubles every two years.
Instead, we’re living by what I call AI Model Law, where the power of generative AI models doubles every three months.
Whether or not you believe Large Language Models (LLMs) are improving at that speed, you can’t deny how quickly AI is being adopted around the world.
AI’s Unprecedented Pace and Adoption
A recent, extensive report from Mary Meeker, a general partner at BOND Investments, clearly shows just how transformative AI really is and how unlike any previous technology shift.
Meeker and her team wrote, “The pace and scope of change related to the artificial intelligence technology evolution is indeed unprecedented, as supported by the data.”
One striking fact from the report is that Google took nine years to reach 365 billion annual searches, but ChatGPT reached the same number in just two years.
Meeker’s findings echo what I’ve been trying to explain for a while: we’ve never seen change like this before.
Comparing Past Tech Shifts to AI Time
I’ve witnessed major tech shifts—the rise of personal computing, the move from analog to digital publishing, and the growth of the internet.
Most of those changes, though rapid at the time, happened gradually over many years.
For example, digital publishing tools emerged in the mid-1970s, but it took until the late 1980s for many to switch, and personal computers only became common a decade later.
By contrast, AI time moves so fast it leaves little room for reflection.
The Internet Era Versus AI Adoption
When the public internet arrived in 1993, it took years for broadband and most people to connect.
Workers gradually adapted over a decade before the internet became an essential part of daily life.
I remember the confusion in 1994 when The Today Show hosts asked, “What is the Internet?”
AI tools like ChatGPT, Copilot, and Claude AI have faced no such confusion.
According to Meeker’s report, ChatGPT users soared from zero in October 2020 to 400 million in late 2024, and 800 million in 2025.
An impressive 20 million of those users pay for subscriptions.
It took decades for people to pay for internet content, but AI users are already opening their wallets.
The Road to the AI Era
Perhaps the rise of the internet and mobile computing paved the way for the AI era.
Artificial intelligence hasn’t suddenly appeared out of nowhere, though it might feel that way.
Almost a decade ago, IBM’s Deep Blue beat chess grandmaster Garry Kasparov.
By 2005, an autonomous car completed the DARPA challenge.
Ten years later, DeepMind’s AlphaGo defeated the world’s best Go player.
While groundbreaking, these milestones arrived at a pace people could digest.
Things began accelerating in 2016, and concerns about AI started to surface.
Though terms like “LLM” and “generative AI” weren’t widely used, major companies like IBM, Amazon, Facebook, Microsoft, and Google’s DeepMind created the Partnership on AI.
This nonprofit aimed to explore AI’s opportunities and challenges for society.
That group still exists, but AI time leaves little room to fully heed its advice.
Predictions and Concerns About AI’s Future
A 2016 Stanford study on AI in 2030 found no reason to fear AI as an imminent threat.
Meeker’s presentation, however, paints a faster-moving future that raises new concerns.
For example, by 2030, AI is expected to create full-length movies and games.
Projects like Gemini’s Veo 3 show we’re on track.
AI is also predicted to operate human-like robots, a leap I’ve never seen in my 25 years covering robotics.
It may even build and run autonomous businesses.
In ten years, AI might simulate human-like minds.
Since ChatGPT’s knowledge comes from what we already know, these predictions might be too cautious.
Even AI can’t predict what’s truly unknown.
The Role of Hardware in AI Development
Some argue the driving force behind AI is not model advancement but hardware improvements.
Nvidia CEO Jensen Huang’s “Huang’s Law” suggests GPU performance doubles every two years.
Without powerful processors, AI progress could stall.
Still, generative AI models are improving faster than silicon chips.
Embracing AI Time and the Future
If you accept the concept of AI Time and the rapid doubling of AI capabilities, then ChatGPT’s vision of the future becomes easier to believe.
You might not feel ready, but the AI revolution is coming — fast and unstoppable.
What do you think about living in AI Time?
Would you be ready for a future shaped by AI creating art, running businesses, and thinking like humans?
Only time will tell.
