Mon, Nov 24, 2025|Sean Michael Cummings

Alphabet Just Proved the AI Boom Is Nowhere Near the Finish Line

Sam Altman shook the AI world three months ago...

During an August interview with CNBC, the OpenAI CEO commented on one of the industry's biggest ideas: artificial general intelligence ("AGI").

These are computer models that can think better than people. OpenAI's stated mission is to ethically build AGI models. So Altman's opinion means a lot to folks who are watching this industry.

But what he said threw the whole idea into question...

I think it's not a super useful term...

It doesn't really matter and it's just this continuing exponential of model capability that we'll rely on for more and more things.

Altman unsettled the AI bulls. And his timing didn't help.

After two flagship AI models showed only slight improvements in August, analysts have begun to worry that large language models ("LLMs") have finally hit their ceiling.

However, Alphabet (GOOGL) just proved that we're nowhere near the limit of what AI models can do... or how much they can scale.

August saw the release of not one, but two flagship AI models...

A week before Altman's AGI comments, Anthropic released its new Claude Opus 4.1 model. Then, OpenAI released ChatGPT-5, an updated version of its chatbot.

But both models showed only minor improvements over their previous versions.

Take a look at how the August models performed...

Both new models barely beat their predecessors. This incremental growth fueled worries about a so-called "scaling wall" – a point at which AI can't get any smarter, no matter how much computing power it uses.

But last week, Alphabet put these fears to rest with the release of its Gemini 3 Pro chatbot.

The AI Scaling Wall Just Crumbled

Instead of the graphic processing units ("GPUs") that have powered the AI boom so far, Gemini 3 uses specifically AI-optimized chips called tensor processing units ("TPUs").

These chips are less versatile than GPUs. They can only do the math that supports AI training. But that limitation makes them highly energy efficient – and extremely fast at AI computations.

This power and speed led to massive improvements over Alphabet's last model. Take a look at the leap Gemini took compared with the slight improvements from the models this August...

Unlike GPT-5 and Claude 4.1, Gemini 3.0 made huge leaps on a wide variety of thinking tasks.

As Gemini Co-Lead Oriol Vinyals remarked on X, "The delta between 2.5 and 3.0 is as big as we've ever seen. No walls in sight!"

Gemini 3.0 proves that AI has room to scale as long as chips and algorithms become more sophisticated.

In other words, the scaling wall that appeared in August was an illusion. We're just scratching the surface of what's possible with AI.

So ignore the bears saying that today's AI models are as good as they get. Alphabet just confirmed that these companies are still advancing – and their stocks still have massive gains ahead.

If you're not on board the AI trade today, you should be. This boom has plenty of room to run.

Good investing,

Sean Michael Cummings

Editor's note: AI stocks have already rewritten the record books. Yet one 15-year market veteran says the biggest risk isn't a crash... It's that you might miss this next leg of the AI megatrend. He believes a $7.4 trillion "Melt Up Tsunami" is on the way, and it could send a new wave of AI winners soaring even higher. In a new briefing, he reveals the No. 1 AI stock to own before it hits.

Alphabet Just Proved the AI Boom Is Nowhere Near the Finish Line

The AI Scaling Wall Just Crumbled

Further Reading