The Boise Brake: Why Your Neighbors Want to Hit Pause on AI

Imagine you are building a car that can travel at the speed of light.

You’ve got the engine running, but you haven't figured out how the brakes work yet.

That is exactly how the folks at Pause AI Boise feel about our current tech race.

This local group is part of a global movement pushing for a temporary halt on the development of "frontier" models.

The Quest for AGI

The group is specifically worried about AGI (Artificial General Intelligence).

Think of AGI as a digital "Einstein" that can do anything a human can, but a million times faster.

It’s like a Swiss Army knife that is smarter than the person holding it.

If we build this without a "kill switch," the Boise group argues we might lose control of our own invention.

A big concern for the movement is something called Alignment.

Alignment is the process of making sure an AI’s goals match human values.

Think of it like training a high-energy puppy.

If you tell the puppy to "get the newspaper" and it breaks through the front window to do it, the puppy is not "aligned" with your rules.

With AI, an unaligned system could solve a problem in a way that accidentally hurts humans.

Pause AI Boise is advocating for Compute Caps.

In plain English, "compute" is just the raw computer power (the number of chips and electricity) used to train a massive AI.

A compute cap is like a speed governor on a truck.

It prevents the machine from going faster than the infrastructure—or the law—can handle.

By limiting how much power companies can use, the group hopes to slow down the "brain growth" of these machines.

You might wonder why a group in Idaho is tackling a global Silicon Valley issue.

The members believe that tech safety shouldn't just be decided by billionaires in California.

They are pushing for transparency, which means "showing your work" so everyone knows what the AI is being taught.

It’s like making sure a chef lists every ingredient in a new dish before serving it to the whole town.

Critics argue that pausing will just let other, less-friendly countries win the tech race.

However, Pause AI Boise suggests that winning a race toward a cliff isn't really winning at all.

They want us to treat AI development like a new medicine—testing it in a lab for years before letting it out into the wild.

We have reached a point where "moving fast and breaking things" might include breaking the very things we need to survive.

If we finally build a mind greater than our own, will it view us as its creator or just a legacy system that needs an upgrade?