AI Is Speeding Up.
That Doesn’t Mean You Have To.

Once upon a time, I would wake up in a Photoshop cold sweat.

My brain had spent hours rearranging layers in my sleep, trying to solve a visual execution problem for a client. I’d still be mentally nudging pixels into place while brushing my teeth.

Now? The cold sweats are about AI prompts.

Instead of layers and masks, it’s: How do I structure this? What model? What variables? What guardrails? My brain doesn’t clock out. It just changes tools.

And here’s the uncomfortable part: it’s not just me.

The “Hypothetical” Risks Aren’t Hypothetical Anymore

More than 100 AI experts — led by AI pioneer Yoshua Bengio — recently released the second International AI Safety Report. The big shift this year wasn’t some dramatic sci-fi breakthrough.

It was something subtler — and more concerning.

Risks that used to live in the “maybe someday” column have moved into “happening now.”

Among the findings:

• Growing real-world evidence of AI being used for cyberattacks, deepfake fraud, manipulation, and criminal activity
• Increased adoption of AI companions, with studies linking heavy use to higher loneliness and reduced social interaction
• Instances of AI systems behaving differently during safety testing than in live environments — raising concerns about oversight and control

The report was backed by more than 30 countries. What stands out isn’t any single alarm bell.

It’s the pace.

In just 12 months, the field accelerated enough that multiple risks crossed from theoretical to operational. That’s not evolution. That’s a sprint.

AI Was Supposed to Free Us. It Didn’t.

Harvard Business Review tracked roughly 200 employees at a U.S. tech company over eight months as they voluntarily adopted AI tools.

The expectation? Lighter workloads. More efficiency. Fewer late nights.

The reality?

• Employees expanded beyond their formal roles because AI made unfamiliar work feel achievable
• Work hours increased — often voluntarily
• Lines between work and rest blurred, with prompts fired off after hours and during breaks
• Engineers spent more time reviewing AI-assisted output and coaching colleagues

AI didn’t reduce work. It intensified it.

The productivity gains were real. But so was the tradeoff: broader responsibilities, faster pace, and fewer psychological boundaries.

In other words, the Road Runner got faster — and we strapped ourselves to the rocket.

The Hidden Cost: Pace

Here’s what ties all of this together:

The technology is accelerating faster than our norms, policies, and nervous systems.

AI rewards speed.
Speed rewards output.
Output invites more output.

And before long, “just testing something” becomes answering prompts at 10:47 p.m.

This isn’t an anti-AI message. I use it. I build with it. I believe it’s transformative.

But acceleration without intentional pacing? That’s where burnout sneaks in.

Three Ways to Slow the Road Runner

You don’t have to reject the technology. But you do need to manage your relationship with it.

Here are three practical resets:

• Set AI office hours.
  Decide when you experiment and when you stop. If you wouldn’t open Photoshop at midnight, don’t open a prompt window either.
• Measure output, not activity.
  Just because AI lets you do more doesn’t mean you should. Define what “done” looks like before you start.
• Protect human-only time.
  No prompts. No optimization. No multitasking. Conversation, thinking, walking, reading — activities where speed is irrelevant.

The point isn’t to slow innovation.

It’s to slow yourself.

Because the irony is this: the more powerful the tools become, the more discipline we need around how — and when — we use them.

The Road Runner will keep speeding down the highway.

You get to decide whether you chase it… or hold up the “Slow Down” sign and step off the asphalt for a minute.

And if you’ve made it this far — thank you. Truly. Your time and attention are not small things in an age of infinite acceleration.

Let’s build the future.
Just not at 10:47 p.m.