Is ChatGPT Dead? – Fat Tail Daily | Australian Markets

Silicon Valley’s AI giants are cracking beneath strain. And they’re about to provide away the keys to the dominion.

Within months, probably the most highly effective AI model makers from OpenAI, Meta, and newcomer Thinking Machines will release open-source AI.

Free for anybody to obtain and construct upon.

This isn’t charity. It’s survival.

They’ve watched China’s DeepSeek drive an total industry to open source with its breakthrough second.

Now, one other Chinese lab, Moonshot AI, exhibits that DeepSeek wasn’t a fluke.

Its latest AI beats OpenAI’s model at coding whereas costing a fraction of the price (85% cheaper on average).

If DeepSeek is the West’s Sputnik second, then Moonshot is Pearl Harbour. The plain occasion that forces speedy whole mobilisation.

I’m satisfied that the West wants its own DeepSeek second. Fast.

The Open Source Wave

OpenAI simply delayed its open-source model again.

Sam Altman says they need more security testing. The actual cause? They’re terrified.

When Moonshot launched Kimi K2 final Friday, it modified the whole lot. The model beats GPT-4.1 at coding duties, but it prices a fraction of its Western counterparts.

That’s not a pricing error. It’s the longer term.

Mira Murati sees this clearly. The former OpenAI CTO simply raised US$2 billion for her new Thinking Machines Lab.

Her promise? A frontier model with ‘significant open source components’ within months.

She’s not alone. In March, Google launched its latest open-source attempt with Gemma 3. But as they’ll even admit — China’s are higher.

The math is straightforward. Why pay 1000’s month-to-month for online AI entry to your business when you’ll be able to run the identical intelligence domestically, or for cents quite than {dollars}?

Hardware prices are plummeting. Nvidia’s H200 chips make operating massive fashions 70% cheaper than final 12 months.

By Christmas, any startup can run GPT-4 stage AI on a $10,000 server.

Soon, machine intelligence will likely be as interchangeable as electrical energy — you gained’t care which place generates it.

The Great AI Commodification

But what occurs when intelligence turns into as low-cost as electrical energy?

We’re about to search out out.

Right now, AI firms trade compute for {dollars}. OpenAI prices US$20 month-to-month for ChatGPT. Anthropic desires US$25 for Claude. Microsoft bundles AI into Office subscriptions.

This business model has months, not years, left to stay.

When each developer can obtain high-level intelligence for fractions of a cent, the pricing flooring drops to zero. Just like electronic mail. Just like web browsers. Just like video calls.

The commodification sample is predictable. First, costs collapse. Then, margins evaporate. Finally, worth migrates up the stack.

Think about cloud computing. Amazon didn’t make billions promoting uncooked servers with Amazon Web Services (AWS). They made it packaging compute with databases, analytics, and developer instruments.

AI will observe the identical path. Raw intelligence turns into close to-free. But intelligence utilized to particular issues? That’s value trillions.

Meta understands this. That’s why they’re paying US$100 million salaries to poach expertise. Not to construct higher fashions. To construct higher merchandise.

As I wrote about a few weeks back, OpenAI is busy securing new income streams and tax breaks to cut back its reliance on Microsoft.

China will get it. Kimi K2 isn’t simply low-cost. It’s designed particularly for coding duties. They’re not competing on uncooked intelligence. They’re competing on utility.

Where Smart Money Goes

The AI model makers are desperately looking for worth past their AI.

Companies charging for uncooked AI entry will watch helplessly as clients change to low-cost options. Just like newspapers watched readers abandon print without cost online content material.

Except quicker.

OpenAI’s US$157 billion valuation assumes they keep pricing energy. But how do you charge for one thing that’s practically free in all places?

There’s a historic sample right here — those that construct infrastructure for tech developments. And people who work on the ‘use case’ or utility layer.

The actual alternative for traders has at all times been additional up the chain. Yes, yes, ‘picks and shovels’ investments have made people tens of millions.

But empires have been made by the customers, not the creators. You know the title Rockefeller as a result of he used railroads — not as a result of he constructed them

Just like with the web growth of the early 2000s. Netscape gave away web browsers. Jeff Bezos and Amazon used them to construct a US$2 trillion empire.

So, that is what I imply once I speak about right this moment’s AI revolution.

It’s not the company that comes out with it first — it’s the company that may monetise the infrastructure.

Today’s equal? Companies utilizing commodified AI to create new disruptors.

Take Perplexity (personal, however rumoured to go public subsequent 12 months). They don’t construct basis fashions. They use open-source AI to reimagine search.

Revenue hit US$100 million this 12 months, up from US$20 million in 2024. Users pay for an expertise Google can’t match. Not as a result of Perplexity has higher AI. Because they utilized it higher.

With a new spherical of funding yesterday, they’re now value US$18 billion. Double their early 2025 worth.

Or have a look at Harvey.ai, the legal AI startup. They wonderful-tune open fashions on proprietary legal information. Their edge isn’t the technology. It’s understanding how attorneys really work.

They simply raised at a US$5 billion valuation. Annual recurring income (ARR) grew 4x final 12 months. Every main law firm desires in.

The sample is evident. As AI commodifies, three issues create worth: area experience, proprietary information, and buyer relationships.

The AI itself? Just one other utility, like electrical energy, water or bandwidth.

Investors are already shifting. Enterprise software program firms utilizing AI have been the first targets during this most up-to-date bull run.

Fund managers within the latest survey (FMS) are putting their bets on tech and AI.

The market is aware of what’s coming. A flood of low-cost, open, and highly effective AI that makes right this moment’s closed fashions uneconomical.

Companies charging for uncooked AI entry will undergo the identical destiny as long-distance cellphone firms. Remember when calls value {dollars} per minute? Now they’re free. The worth moved to smartphones and apps.

But firms utilizing low-cost AI to resolve costly issues? They’ll mint money for many years.

Healthcare diagnostics. Financial fraud detection. Industrial optimisation. Legal analysis. These markets are value trillions.

And they’re about to be disrupted by firms wielding low-cost, commodified intelligence.

The West’s latest DeepSeek second isn’t a menace. It’s the best investment alternative for the reason that web’s disruption.

Just make sure you’re betting on the applying layer.

Because when intelligence prices nothing, the whole lot constructed on high of it turns into the true worth.

Regards,

Charlie Ormond,
Editor, Alpha Tech Trader and Altucher’s Early-Stage Crypto Investor