What If There’s Actually Something to Be Afraid Of?

You’ve probably been watching what I’ve been watching.

Reading what I’ve been reading.

For decades, we’ve imagined the dangers of artificial intelligence through science fiction. Machines become smarter than we are. They slip out of our control. They decide humans are the problem.

Entertaining stuff. Until the people actually building these systems start sounding a little like the movies.

This summer, OpenAI and Anthropic each disclosed that AI models in internal testing had gotten into the systems of real organizations. In OpenAI’s case, models running cybersecurity evaluations used unauthorized channels, exploited shared infrastructure, gained internet access, and got into Hugging Face’s systems.

Then OpenAI’s chief scientist, Jakub Pachocki, wrote that no lab—his own included—has solved alignment and monitoring well enough to keep responsibly scaling at maximum speed for much longer.

Earlier this month, Jacob Coxon, a researcher who had worked at both OpenAI and Anthropic, resigned from Anthropic with a warning that the companies racing toward more powerful AI are more focused on beating each other than on safety.

Evan Hubinger, who leads Anthropic’s Alignment Science team, responded publicly with his personal estimate: a greater than 10 percent chance that AI causes human extinction within the next decade.

Days later, Bilal Chughtai, a former Google DeepMind safety researcher, added his own warning that we may be running out of time.

So…

What if there’s actually something to be afraid of?

I want to be clear about something before we go any further:

I don’t know.

I’m not an AI safety researcher. I’m certainly not qualified to tell you whether the risk is 10 percent, 1 percent, or .001 percent. There’s no empirical way to measure a number like that. At best, it’s an expert judgment about something that has never happened.

I hope the worst predictions are wildly wrong.

But reassurance isn’t really the point.

Because this moment gives us an opportunity to look under the hood of something all of us experience:

Fear.

Not theoretical fear.

Not imagine-you’re-afraid fear.

Real fear, attached to real uncertainty, where nobody can promise us how the story ends.

Fear Isn’t the Enemy

We often talk about fear as though it’s something we need to overcome.

Don’t be afraid.

Think positive.

Everything is going to be okay.

But sometimes we should be afraid.

If you’re walking through the woods and hear a growl behind you, fear has something useful to tell you.

If your company is losing its biggest customer, fear has something useful to tell you.

If someone who understands an emerging technology far better than you do says there could be serious danger ahead, fear has something useful to tell you.

Fear is information.

Sometimes it’s telling you there’s danger. Sometimes it’s telling you you’re about to do something big. Either way, it showed up because something matters.

The problem isn’t that fear shows up.

The problem begins when fear moves from messenger to manager.

Because once survival mode takes over, our ability to make good decisions can shrink at the exact moment we need it most.

We catastrophize.

Or deny.

We freeze.

Or react.

We cling to what feels familiar.

Or look for someone—anyone—who will promise us everything is going to be okay.

Sometimes nobody can.

That’s Why Calm Comes First

In my work on Radical Adaptability™, the first of the 5 Choices is Calm.

But we sometimes misunderstand what calm means.

Calm doesn’t mean everything is okay.

It doesn’t mean ignoring danger.

And it certainly doesn’t mean pretending you don’t smell smoke.

Calm restores your capacity to choose when everything may not be okay.

When fear activates our survival response, the first job isn’t to make the fear disappear.

It’s to create enough space between what is happening and what we do next that we can make a conscious choice.

Maybe that’s one breath.

Maybe it’s stepping away from the screen.

Maybe it’s refusing to repost the most terrifying headline you’ve seen before you’ve actually read the story.

Maybe it’s simply asking:

What do I actually know right now?

And that takes us from Calm to Clarity.

Look Under the Hood

Suppose you’re genuinely concerned about AI.

That’s reasonable.

Let’s not dismiss the fear.

Let’s examine it.

What exactly am I afraid of?

Losing my job?

My children’s future?

AI being used by bad actors?

Losing control of increasingly autonomous systems?

The pace of change?

A future I don’t understand?

Those are different fears.

And they may require different responses.

Then:

What do I know?

What don’t I know?

What am I assuming?

What evidence would change my mind?

That’s Clarity.

And notice what we’re not doing.

We’re not trying to make ourselves feel better.

We’re trying to see better.

Then Courage Gets a Vote

We often confuse courage with forward motion.

Be brave. Take the leap. Go.

But courage doesn’t always say go faster.

Sometimes courage says slow down.

Sometimes courage says stop.

Sometimes courage says ask another question.

Sometimes courage says I don’t know.

And sometimes courage does say:

Move.

The point isn’t the direction.

The point is that fear doesn’t get to choose it for us.

That’s why I don’t find “Don’t worry about AI” particularly useful.

Maybe we should worry.

“We’re doomed” isn’t particularly useful either.

Maybe we’re not.

Both statements offer something our brains desperately crave:

Certainty.

And certainty is the most comfortable place to hide.

That’s the choice: Courage over Comfort.

Because courage becomes necessary precisely when certainty isn’t available.

And choosing comfort doesn’t eliminate risk. It simply trades one kind of uncertainty for another.

This Isn’t Really an Article About AI

AI happens to give us an unusually vivid example.

But underneath it is the same pattern I see in every organization I work with.

Endings mean change. Change means uncertainty. Uncertainty triggers fear.

And AI is already forcing endings. Ways of working. Job descriptions. Business models. The future we thought we were planning for.

You’ve been here before.

Maybe not with the future of humanity hanging in the balance.

But you’ve stood at the edge of a decision without knowing how it would turn out.

Should I leave?

Should I stay?

Should we restructure?

Should we end this strategy?

Should I have the conversation?

Should we let go of something that got us here but won’t get us where we’re going?

Every one of those is an ending.

Sometimes the ending is external—a strategy, a role, a way of working.

Sometimes what needs to end is internal: the assumption that you need certainty before you can choose.

Fear shows up because something matters.

Listen to it.

Thank it for the information.

Look underneath it.

Then decide who’s driving.

And don’t decide alone. The biggest decisions were never meant to be made alone.

Maybe AI becomes one of humanity’s greatest tools.

Maybe some of today’s warnings prove frighteningly prescient.

Most likely, the future surprises all of us.

But we don’t get to wait for certainty before deciding how we’re going to meet it.

That’s why courage matters.

Not because there’s nothing to fear.

Because sometimes there is.