The future of war arrived. We aren’t ready.

Steve Blank, the Silicon Valley entrepreneur and educator, recently warned startup founders that if their company was built on assumptions from more than two years ago, their business plan may already be dead. The same is true of military plans.

Replace “startup” with “combatant command,” and the warning becomes existential.

Every OPLAN, CONPLAN, and campaign design sitting in a classified vault somewhere — from European Command to Indo-Pacific Command, from NATO’s Article 5 contingencies to the South China Sea access-denial scenarios — was built on assumptions that are no longer true. The target sets have changed. The kill chains have changed. The very physics of the battlefield have changed. And the institutions that own those plans are, by and large, heads down — still recruiting for the team, still building the product, still defending the roadmap. The question is no longer whether the force must change. It is whether the institution can move fast enough to matter.

I am not writing this from the outside looking in. I spent 37 years in the United States Navy — as a submarine commander, a nuclear engineer, the first non-Engineering Duty Officer to serve as Chief Engineer of the Navy, and finally as the 26th Chief of Naval Research, overseeing $4 billion in science and technology investment for the Navy and Marine Corps. I championed autonomy, AI, and advanced technologies from inside the institution — and I know exactly how hard that institution can resist being told that its assumptions are wrong. I watched more promising ideas stall in bureaucratic amber than I care to count.

The science was not the problem. The technology was not the problem. The urgency — or the lack of it — was the problem. The planning processes, the acquisition timelines, the risk aversion baked into a system designed for a slower world: those were the problems. And they are the same problems now rendering every OPLAN on every classified server obsolete.

The hedge strategists were right

For years a small community of defense visionaries argued that the future of warfare would be faster, cheaper, more distributed, and driven by autonomous systems and artificial intelligence rather than by manned platforms and mass. They were largely ignored — called too commercial, too theoretical, too Silicon Valley. Their concepts were interesting, the establishment said, but not yet ready for the battlefield. The institutional response was to fund some new ideas on the margins while protecting the legacy programs of record.

We are out of time. The future has arrived.

We’ve got a dead moose on the table — the kind of problem that only gets worse the longer we pretend it’s not there.

Ukraine compressed the kill chain from days to minutes. Cheap FPV drones destroyed million-dollar platforms. Electronic warfare became an hourly contest. Recent attacks on exposed aircraft and logistics nodes have shown the same lesson elsewhere: fixed assets, predictable basing, and legacy protection assumptions are liabilities. In the Pacific, that logic applies to both sides of the Taiwan Strait. Surface ships, ports, airfields, tankers, and logistics hubs are all now targetable at scale.

Meanwhile, in the U.S., companies once treated as peripheral — Anduril, Palantir, Shield AI, and dozens of others — are no longer future capabilities. They are present tense operational realities. AI-enabled logistics, autonomous ISR, and software-defined weapons are reshaping what is possible on a battlefield faster than any OPLAN was written to account for.

But most plans were not written with these capabilities as the centerpiece.

That is the dead moose on the table: Most operational plans reflect a world that existed five to 10 years ago. They are built around force employment models designed for the platforms we have had for decades, not the capabilities that almost anyone can buy off the shelf. They assume timelines, logistics chains, and command structures that autonomous systems are already beginning to dissolve.

No one is saying it out loud because the institutional incentives cut entirely the other way. You do not get promoted for declaring the OPLAN obsolete. You do not win a program of record by arguing that the program of record is no longer the correct way to field new capabilities. The services have spent decades and hundreds of billions of dollars building the force that these plans employ. Rewriting the plans from first principles means questioning much of what sits upstream.

But that is precisely what needs to happen.

The vindication no one wanted to celebrate

This moment is uncomfortable for the visionaries, too. Being right about the future when the present catches up does not feel like a celebration — not when the validation comes in the form of burning armor in eastern Ukraine, destroyed tankers on Central Command ramps, or contested airspace over the Taiwan Strait. The autonomy advocates, the AI-in-the-loop theorists, the commercial defense investors who built these ecosystems at the margins of the institutional military — they did not want to be vindicated by a war. They wanted to be vindicated by a transformation that happened before one was necessary.

The future has just smacked the present in the face, which makes the urgency sharper, not softer.

The right question now — the one every combatant commander, every J5, every service chief should be forced to answer — is the one Blank poses to startup founders: If you were designing the operational plan today, using today’s tools in today’s operating environment, what would you build?

The answer is not in any OPLAN currently on file. And the institutions that cannot confront that fact — that cling to the sunk costs of their legacy planning constructs, their manned-platform employment models, their 72-hour targeting cycles — will face the same fate as the founders Blank is warning. They will arrive at the next conflict with an obsolete plan, a team sized for the wrong problem, and a product the battlefield will refuse to buy.

What meaningful change looks like

This is where the argument has to get specific — because “we need to change” is not a strategy. The question to ask is: what exactly needs to change, and who does it?

Start with the plans themselves, rewriting them from first principles. A new OPLAN is not transformation if the same staff uses the same assumptions and bolts on drones, AI, and autonomous systems as supporting tools. Meaningful change means campaign designs built from the ground up around the actual capabilities now available — where AI-enabled ISR, autonomous logistics, and human-machine teaming are the primary design constraints, not exotic edge cases. That requires different inputs to the planning process, not just different outputs.

Second, move money at visible scale. The services do not need another innovation fund at the margins. They need to shift real dollars from legacy programs whose operational logic is eroding toward autonomous systems, software-defined capabilities, and commercial defense companies that are already moving at battlefield speed. This is not an argument against platforms. It is an argument that the ratio is wrong and everyone in the Pentagon knows it.

Third, rewrite doctrine. The Air Force’s Agile Combat Employment concept assumed dispersal plus air superiority would substitute for hardened shelters. Ukraine and CENTCOM proved that wrong at devastating cost. Army doctrine still treats excavation as something done with backhoes and shovels. Neither service has a doctrine for rapidly deploying the kind of protection and survivability infrastructure that a drone-saturated battlefield now demands. Rewriting doctrine is unglamorous work, but it is the foundation everything else rests on.

Fourth is hardest, but necessary: put different people in the room. Some existing staffs can pivot. Many cannot. Too many senior planning processes are run by officers trained and promoted in a pre-autonomy world. They were selected and promoted for mastery of planning frameworks that are now obsolete. The people who warned that the battlefield was becoming cheaper, faster, more distributed, and more automated need authority, not advisory roles. The visionaries who were sidelined should not be invited back to brief the plan. They should help write it.

New OPLANs written by old thinking are not transformation. They are renovation. The battlefield is not asking for a remodel.

This is why I now work with companies building the autonomous systems, AI-enabled decision tools, and software-defined capabilities the next fight will require. That is not a departure from my Navy career. It is a continuation of it. The mission hasn’t changed. What changed is my assessment of where the leverage is.

Steve Blank asks founders to imagine starting over with today’s tools and today’s market. I ask the same question of the institution I served: if you were designing the force and writing the plans from scratch today, knowing what Ukraine has taught us, knowing what drone warfare has done to fixed assets, knowing what autonomy is doing to the speed of decision-making, what would you build?

Answer that question honestly and the path is clear: new plans, new doctrine, new money flows, and new people with real authority.

The visionaries warned this moment would come. They were right. The only remaining choice is whether the institution catches up or waits for the next conflict to make the case.