I Never Expected OpenAI to Last This Long

When ChatGPT launched in late 2022, I gave OpenAI about six months.

That's not an insult. It was simply my professional prediction.

I've been around technology long enough to know that the internet is littered with the graves of products that looked revolutionary for a few weeks before reality arrived with a baseball bat.

Remember Clubhouse?

Remember Google+?

Remember Quibi?

Exactly.

So when people started posting screenshots of ChatGPT writing poems, solving math problems, composing emails, and occasionally hallucinating itself into another dimension, I thought:

"Interesting. Fun. But surely this will fade."

I was wrong.

Spectacularly wrong.

The Week The Internet Lost Its Mind

ChatGPT launched on November 30, 2022.

Within five days, it had crossed one million users. Within two months, it reached 100 million users, becoming the fastest-growing consumer application in history at the time.

The internet reacted exactly as the internet always reacts.

Half the population declared that AI would save humanity.

The other half declared that AI would destroy humanity.

And a third half—yes, internet math—started asking it to write rap songs about their cats.

Everywhere I looked, people were posting screenshots.

Students.

Lawyers.

Marketers.

Software developers.

People who had never touched a line of code suddenly became prompt engineers overnight.

LinkedIn transformed into an AI carnival.

Everyone became a futurist.

Everyone became an expert.

Everyone became insufferable.

It was glorious.

OpenAI's Greatest Achievement Wasn't AI

This may sound strange, but OpenAI's biggest achievement wasn't building a language model.

It was making ordinary people care about artificial intelligence.

For decades, AI existed mostly inside research papers, conferences, and science-fiction movies.

Then OpenAI did something brilliant.

They turned it into a chat window.

That's it.

No complicated interface.

No technical jargon.

No 200-page manual.

Just a blinking cursor and a question:

"How can I help you today?"

That simple decision may end up being one of the most important product design choices in technology history.

The GPT-4 Moment

Then came GPT-4.

This was the moment many skeptics, including me, stopped laughing.

According to OpenAI's technical report, GPT-4 demonstrated performance approaching or exceeding human-level results on numerous professional and academic benchmarks, including scoring around the top 10% on a simulated bar exam.

Researchers examining an early version of GPT-4 published a paper titled Sparks of Artificial General Intelligence.

The title alone caused approximately seven million arguments online.

The researchers argued that GPT-4 displayed capabilities spanning coding, law, psychology, medicine, mathematics, and reasoning at levels that felt fundamentally different from previous systems.

The reactions were predictable.

Researchers got excited.

Journalists got dramatic.

Twitter got weird.

Business consultants doubled their rates.

The Strange Thing About AI Predictions

I've noticed something funny.

The people who confidently predict AI's future are usually wrong.

In 2023, many people said AI would replace programmers.

Programmers are still here.

Others said AI would replace writers.

Writers are still here.

Some predicted that AI would replace Google overnight.

Google is still here.

And yet...

The world has undeniably changed.

That's the strange part.

The most dramatic predictions failed.

The technology still transformed everything anyway.

OpenAI Survived What Usually Kills Companies

Most startups don't die because of competition.

They die because they become irrelevant.

OpenAI somehow avoided that fate.

Think about everything that happened.

Massive competition.

Public controversies.

Leadership drama.

Lawsuits.

Regulatory scrutiny.

Hallucinations.

Security concerns.

Infrastructure challenges.

Existential debates about humanity.

And somehow the company kept growing.

At one point, the conversation around OpenAI became bigger than the products themselves.

Every new release felt like a global event.

Every CEO statement became breaking news.

Every model update generated thousands of YouTube videos within hours.

I've never seen anything quite like it.

The Numbers Are Getting Ridiculous

Here's the part that genuinely surprises me.

The growth never stopped.

OpenAI reported ChatGPT reaching hundreds of millions of weekly users, and by 2026 reports indicated the platform had reached roughly 900 million weekly active users while approaching or surpassing one billion monthly users.

Pause for a second.

One billion.

When a product reaches one billion users, it stops being a product.

It becomes infrastructure.

Electricity is infrastructure.

The internet is infrastructure.

Search engines are infrastructure.

We're witnessing AI move into that category.

And that realization still feels slightly absurd.

My Favorite OpenAI Story

My favorite OpenAI story isn't about GPT-4.

It isn't about billion-dollar valuations.

It isn't about funding rounds.

It's about ordinary people.

A teacher using ChatGPT to prepare lessons.

A founder writing a business plan.

A student learning a difficult concept.

A developer fixing a bug at 2 a.m.

A writer staring at a blank page.

Technology becomes important when people stop talking about the technology and start talking about what it helps them accomplish.

That is what happened to ChatGPT.

The Future Nobody Understands

The funniest thing about AI is that nobody understands where this ends.

Not me.

Not journalists.

Not investors.

Not politicians.

Probably not even OpenAI.

Every few months, reality produces something that wasn't on anyone's bingo card.

Remember when people thought AI was mostly about text?

Then image generation exploded.

Then coding.

Then voice.

Then reasoning.

Then agents.

Then research systems.

Then whatever happened last week.

Trying to predict the future of AI feels like trying to predict the weather on Mars.

So Was I Wrong?

Yes.

Completely.

I never expected OpenAI to last this long.

I thought ChatGPT would be a fascinating technology demonstration.

Instead, it became one of the most influential consumer products ever created.

I expected a trend.

I got a platform.

I expected a curiosity.

I got a cultural phenomenon.

I expected six months.

Here we are years later, still talking about OpenAI.

Still arguing.

Still experimenting.

Still being surprised.

And if there's one lesson I've learned from watching this story unfold, it's this:

Never underestimate a company that manages to make incredibly advanced technology feel simple.

The technology world is full of brilliant inventions.

Very few become habits.

ChatGPT became a habit.

And habits are much harder to disrupt than technology.

Which means OpenAI may be around a lot longer than many of us originally expected.

Including me.