Apparently OpenAI Is Deep in the Red Too. So Who Will Pay for AI in the End?

I was reading tech news when I saw a story about leaked financial information from OpenAI.

Revenue at around 2 trillion yen.

And costs at around 5 trillion yen.

In other words, a massive deficit.

Of course, leaving aside whether the exact numbers are true, at least one thing seems clear: the AI industry is running while burning an unbelievable amount of money.

At this point, the scale is already beyond what ordinary people can easily feel.

I can imagine a deficit of tens of millions of yen, or an investment of several hundred million yen.

But when someone says several trillion yen, my only reaction is something like:

“Is that the budget of some country?”

That is about as far as my intuition goes.

AI Eats Money

This has been obvious for a while.

AI looks like magic.

You ask a question, and it answers.

It writes code.

It creates images.

It even does research.

But behind the scenes, massive data centers are consuming electricity, and GPUs are roaring away.

In other words, AI is fundamentally a money-eating machine.

And not a small one.

Investors are carrying that cost for now.

But at some point, someone has to pay the bill.

There is no free lunch.

The “Fable 5 Is Expensive” Problem May Be Only the Beginning

Recently, when Anthropic released Fable 5, many users were surprised by the pricing.

Honestly, my reaction was:

“Uh… okay.”

I understand that it is powerful.

But high performance and high prices tend to arrive as a set.

OpenAI’s Codex also looks relatively cheap right now.

But there is no guarantee that this will last forever.

If anything, it feels more likely that we are still in the market-share phase, and price revisions may come later.

Netflix was cheap at first.

Cloud services also looked cheap at first.

Subscriptions usually work that way.

Before you know it, the price goes up.

I do not see why AI would be the exception.

Then Chinese AI Enters the Picture

From here on, this is pure speculation.

But I do not think it is impossible.

What happens if American AI companies move toward a high-price strategy?

Chinese players may arrive.

“It has 90% of the performance.”

“It costs one-third as much.”

“You may have some security concerns.”

What would happen if products like that appeared?

In fact, Chinese products have taken global markets many times with this same pattern.

Home appliances.

Smartphones.

Solar panels.

EVs.

Drones.

At first, people doubted the quality.

But low prices and practical usefulness expanded the market.

AI could very well follow the same path.

I Would Probably Use It Too

Of course, there are security concerns.

It would be difficult to put enterprise systems or confidential information into such tools.

But what about personal use?

If there were an AI that costs $100 per month and another that costs $20 per month.

And if the performance gap were only around 10%.

Honestly, I think many people would move to the cheaper option.

I would probably use it too.

I am not planning to ask an AI about the Tiananmen Square incident. I just want it to fix bugs.

Will We See the AI Version of the “Chinese Product Era”?

Chinese products are cheap across many categories.

And they have expanded markets through that low cost.

Of course, the quality has also improved dramatically compared with the past.

If the same thing happens in AI, future competition may shift from a performance race to a price race.

OpenAI and Anthropic are both trying to build smarter models.

But users do not necessarily want the world’s best-performing model.

They want something smart enough and cheap enough.

When you think about it, this is the same road that PCs, smartphones, and home appliances have all traveled.

There is no guarantee that AI will remain special forever.

A few years from now, maybe we will no longer be talking about “OpenAI or Chinese AI.”

Instead, it may become a much more familiar kind of market:

“Premium AI or budget AI?”