The AI Breakthrough You Can’t Afford to Ignore

“Since the Industrial Revolution first made the perception of change through technology clear, the rate has continued to increase, until now the wind of change has risen from a zephyr to a hurricane. It is change, continuing change, inevitable change, that is the dominant factor in society today. No sensible decision can be made any longer without taking into account not only the world as it is, but the world as it will be—and naturally this means that there must be an accurate perception of the world as it will be.”

—Isaac Asimov, 1978

Do you ever feel like you’re being steamrolled by change? Everywhere we turn, we see mind-blowing advances that make our childhood years feel like a bygone era. 

We’ve all heard our parents and grandparents weigh in on change. It always starts with, “I remember when…” 

…phones had a cord and were attached to the wall (gasp!). 

…we didn’t have a computer in the house. 

…cameras had film.

Every generation experiences transformation, but we are on the brink of something unprecedented. Let’s take a break from news about the Middle East and oil to focus on a story you may have missed—one that deserves your full attention. 

A Preview of Mythos

Anthropic, an artificial intelligence company, just announced results from its latest AI model, Mythos. From the company:

Mythos Preview has already found thousands of high-severity vulnerabilities, including some in every major operating system and web browser. Given the rate of AI progress, it will not be long before such capabilities proliferate, potentially beyond actors who are committed to deploying them safely. The fallout—for economies, public safety, and national security—could be severe.

What Anthropic is saying here is that its new AI model can find serious, previously unknown security flaws in every major operating system and web browser and develop working exploits for them. That puts virtually every entity that touches the internet at risk.

Even if you are a proud Luddite, refusing to touch a computer or cell phone, you have exposure. If you use a bank or electricity, or are even marginally integrated into modern society, there is no escaping software.

Claude Mythos Preview demonstrates a major leap in cyber skills—the vulnerabilities it has spotted have in some cases survived decades of human review and millions of automated security tests, and the exploits it develops are increasingly sophisticated.

Mythos has not yet been released to the general public, but Anthropic has sounded the alarm, releasing a preview of the model to a group of companies with a lot to lose if the world does not find a way to solve its cybersecurity issues. It has enlisted the help of firms including Apple, Google, CrowdStrike, JP Morgan, and Microsoft to collaborate on “cyberproofing” critical software systems. 

All of this might sound alarming, but it’s exactly what other nations are racing to accomplish. The day an adversary develops its own version of Mythos is the dawn of a new era of conflict.

Again, from Anthropic (emphasis mine):

The work of defending the world’s cyber infrastructure might take years; frontier AI capabilities are likely to advance substantially over just the next few months. For cyber defenders to come out ahead, we need to act now. 

I encourage you to watch this video from Anthropic on the steps being taken to shore up cybersecurity in the face of rapidly evolving AI models.

As my friend Stephen McBride says, today’s AI model is the worst it will ever be. There are reasons to be optimistic—perhaps we were the first to get to this level of AI, and we have time to prepare.

It sounds like the power of Mythos has rattled the folks at Anthropic, but not every company—nay every nation—would be as forthright about the challenges coming in the months ahead.

You can read Anthropic’s press release on Mythos here.

As you may recall, several months ago we sent you a survey about AI and received over 7,000 responses. Those responses led us to develop what we think will be the essential course on artificial intelligence— covering how we got here, where the technology stands today, where it is heading, and how you can use AI to your advantage while preparing yourself and your family for the changes underway. Look for an announcement in late May or early June as we wrap up this multi-month research project. 

Until then, I encourage you to sign up for our Strategic Investment Conference. From AI to the Middle East to the US economy, we will cover the pressing topics of today, with ideas on how to manage your way through this rapidly changing world of ours.

Thanks for reading, as always, and for your support of what we do here at Mauldin Economics. I’d love to hear your thoughts on Mythos and what comes next. Reply to this email or post a comment on our website.

I hope to see you, virtually, in May at the SIC!

Ed D’Agostino

Partner & COO