The Jagged Frontier
- JB Higgins
- May 8
- 4 min read
By JB Higgins - FrontierOps Advisory
There’s a phrase I heard repeatedly at this week that immediately resonated with me:
“The Jagged Frontier.”
The term comes from Ethan Mollick — one of the clearest thinkers currently writing about artificial intelligence and the future of work through his Substack, One Useful Thing.
As someone who has spent decades inside large operational systems — Fortune 25 companies, enterprise transformations, large-scale Jira implementations, executive governance structures — I think “The Jagged Frontier” may be the single most accurate phrase for the state of AI in 2026.
And ironically, long before I heard Mollick say it on stage, the concept itself had already inspired the visual identity of FrontierOps.
That’s why our logo looks the way it does.
Not smooth hills. Not polished curves. Not some utopian glowing future.
Jagged mountains.
Because that’s what this moment actually feels like.
AI Is Not a Smooth Curve
One of the biggest mistakes executives are making right now is assuming AI capability progresses linearly.
It does not.
AI behaves like uneven terrain.
One moment it produces astonishing strategic insight, writes functional code, summarizes complex systems, or generates useful architecture recommendations.
The next moment it confidently invents nonsense.
That inconsistency is the frontier.
Not because AI is “bad.” Not because it’s “fake.” But because we are interacting with a fundamentally new kind of cognitive tool that is still discovering the edges of its own capability.
Mollick’s framing is important because it avoids both extremes:
the doom cult
the hype cult
Instead, he describes reality.
And reality is jagged.
The Demo Problem
At Team ’26, the central tension underneath nearly every AI conversation was obvious:
The demos are impressive.The operational reality is harder.
That gap matters.
A lot.
Because many organizations are still treating AI like a feature rollout instead of a structural transformation.
Executives see a polished agent demo and assume the enterprise is ready.
But most enterprises are not structurally coherent enough to support large-scale agentic workflows yet.
That’s the real story.
The hard part is not generating text.
The hard part is integrating intelligence into systems full of:
fragmented ownership
unclear governance
inconsistent workflows
bad signaling
contradictory incentives
low-trust operational structures
performative reporting
unstable data foundations
AI doesn’t magically eliminate organizational entropy.
In many cases, it exposes it.
Faster.
(As my grandfather used to say, power tools won't make you a better carpenter, but they'll sure cut your fingers off faster.)
“HR Is the New R&D”
One of the more important ideas Mollick has been discussing recently is this:
“HR is the new R&D.”
That line sounds strange at first.
Until you realize what agentic AI actually changes.
Historically, technology scaled primarily through systems.
AI scales through people plus systems.
The differentiator is no longer just tooling capability.
It is organizational adaptability.
Which teams can learn fastest?
Which leaders can redesign workflows?
Which organizations can tolerate ambiguity?
Which companies can retrain themselves operationally?
That becomes the new competitive advantage.
Not who bought the software first.
But who redesigned themselves around it first.
The Human–AI Workforce Is Already Here
One thing became very clear at Team ’26:
The “future AI workplace” is no longer theoretical.
It already exists.
Quietly.
Unevenly.
Messily.
Some teams are already using agents as operational force multipliers:
summarizing meetings
generating requirements
accelerating development
analyzing workflow bottlenecks
drafting communications
synthesizing knowledge
identifying operational drift
Meanwhile, other teams are still debating whether AI should even be allowed.
That spread is enormous.
And it creates a strange new organizational reality where two departments inside the same company may effectively be operating in different centuries.
The Real Bottleneck Is Structural
This is where my own work increasingly intersects with what Mollick is describing.
Most organizations assume their AI problem is technological.
Usually it’s structural.
An AI agent dropped into a chaotic operational system behaves exactly like a human dropped into one:
confused.
Because AI depends heavily on:
clarity
signaling
ownership
workflow consistency
governance integrity
accessible knowledge structures
The cleaner the operational architecture, the more useful the AI becomes.
Which means the organizations that win in the AI era will not necessarily be the ones with the most advanced models.
They’ll be the ones with the most coherent systems.
That distinction matters enormously.
Why the Frontier Matters
The frontier has always been dangerous.
But it has also always been where disproportionate opportunity exists.
That’s true in exploration. That’s true in business. And it’s true now with AI.
We are entering an era where intelligence becomes abundant.
Which means the scarce resource becomes something else:
judgment
coordination
trust
operational clarity
systems thinking
moral architecture
organizational coherence
The companies that understand this early will build enormous advantages.
The companies that mistake AI for magic will create very expensive confusion.
The Mountains in Our Logo
When I built FrontierOps, I chose jagged mountain peaks intentionally.
At the time, I saw them as symbols of operational reality:
hard terrain
limited visibility
high consequence
navigation through uncertainty
Now they mean something even more specific.
They represent the actual shape of the AI era.
Not smooth.
Not clean.
Not predictable.
Jagged.
And the organizations that succeed will not be the ones waiting for the terrain to flatten.
They’ll be the ones who learn how to navigate it.
If your organization is moving from AI curious to AI operational, the question is no longer whether AI matters.
The question is whether your systems are structurally ready for it.
That’s the frontier we help organizations navigate.
Comments