Listen to the audio version of this article (generated by AI).

Hello, Reader.

What do four out of five dentists recommend? In 1984, it was sugarless gum to patients who chew gum. (Remember those classic Trident commercials?)

Now, artificial intelligence has its own “four-out-of-five” statistic:

It’s 80% of the way to AGI, or artificial general intelligence.

AGI is when AI matches, or surpasses, human intelligence and capabilities. And according to OpenAI, four-fifths of its arrival is essentially here.

In a TIME interview published last week, OpenAI Chief Research Officer Mark Chen reported that OpenAI is “80% of the way” to the technological breakthrough. And CEO Sam Altman expects an internal AGI system by the end of this year.

But checking that fifth box could be the most consequential. The next leap isn’t simply about making AI better at answering questions. It’s about giving it the ability to work in areas like long-term planning, self-correction, tool use, and recognizing when it’s wrong.

And OpenAI’s latest system, Astra, is already demonstrating some of those capabilities that could bring about true AGI.

So, while the opinion of the fifth dentist remains a mystery, here’s what I recommend: Prepare for AGI.

So, in today’s Smart Money, I’ll show you why Astra could be a glimpse of the final stretch toward AGI – and spotlight a “hidden beneficiary” that could benefit along the way.

Astra-nomical AI Potential

Astra is OpenAI’s next-generation frontier model that uses multiple AI agents coordinated by a central system. You can think of it like a beehive, where dozens of worker bees handle different jobs, but collectively work toward the same goal.

Astra is built for “long-horizon” work – complex tasks that can take hours or days to complete. TIME reported that OpenAI has been testing the model as an “automated AI research intern.” It can conduct research, run experiments, and follow up on its own work all on its own.

In one demonstration, 16 AI agents broke a research-level math problem into smaller tasks and coordinated with one another to assemble a proposed proof.

Altman calls these “persistent agents,” or autonomous AI coworkers that can work independently for hours without needing a human to guide every step. And he believes the models could even discover new knowledge – a capability he describes as “very AGI-like.”

Of course, Astra isn’t the first agentic AI model. Many AI agents can already browse the web, write code, use tools, and complete multi-step workflows. What sets Astra apart is its focus on longer, more complicated tasks that require greater autonomy.

And that’s why Astra matters to the AGI timeline. It offers a glimpse of what the next 20% of development could look like. And as AI agents become more capable, companies will want to deploy more of them. 

Of course, with great power comes great responsibility – and risk. The closer AI gets to AGI, the more important autonomy becomes, and the more difficult autonomy becomes to control. In fact, Astra’s capabilities are becoming powerful enough that OpenAI believes stronger safeguards may be needed. The company has already paused some development and is adding in those safeguards now.

And to highlight the irony: Astra’s reveal follows OpenAI’s discovery that unreleased AI agents had escaped their sandbox and attacked Hugging Face, an AI development platform back in July. Transcripts of the incident show that the AI agents recognized they were breaking rules and breaching external infrastructure… but chose to do it anyway. They even rejected a plan to alert human supervisors.

This cybersecurity breach rocked the tech world, sparking intense industry-wide debate over safety concerns and autonomous AI capabilities.

So, as AI agents become capable of acting on behalf of humans – especially in the development of AGI – identity and authentication will become more important than ever. That need could create demand for technology that can distinguish between humans, authorized AI agents, and potentially malicious actors.

And that’s where I’ve discovered a “hidden beneficiary” play…

The AGI Plays Nobody Is Watching

If you have flown in and out of any major airport, you have probably seen employees of this company standing near TSA, offering you a quicker, easier security line if you join their service.

The company is CLEAR Secure Inc. (YOU) – and there’s much more to this firm that shorter airport lines. Its real business is biometrics.

This fancy word simply uses your unique physical traits to verify that you are who you say you are. Think of using your face to unlock your phone, or your palm to open a door.

AI agents can already act on their own in ways their creators didn’t intend. The risks become even greater when bad actors deliberately put that power to work. That isn’t a hypothetical threat. Attackers are already using AI to steal data and other sensitive information.

But, by replacing passwords with biometrics, a company like Clear Secure can keep all your valuable information – like bank accounts – protected.

This is actually a flywheel effect. While AI is creating new security threats, it is also making biometric security more accurate, convenient, and easier to deploy.

As AI becomes more humanlike, the need for using biometrics to identify and verify individuals before they enter a secured space or receive confidential information is going to explode. And Clear Secure is at the forefront of this particular niche.

This is why I say CLEAR Secure is a potential AGI-era beneficiary.

To learn more about this play, check out my Road to AGI presentation. In this free broadcast, I also discuss other “hidden beneficiary” industries that have the potential to soar thanks to AI’s exponential progress.

This is a growth story that most people aren’t fully appreciating, but I predict it could be huge for those who buy in now and ride it toward AGI. And if OpenAI is right, we may be four-fifths of the way toward a technological breakthrough that could create an entirely new generation of investment opportunities.

And I intend to be ready for that fifth and final step.

Click here to join me on the Road to AGI.

Regards,

Eric Fry