Are even the computers smart enough for the gray zones?

General Joseph Votel, head of US Special Operations Command, is worried about “gray zones.” As he told the House Armed Services Subcommittee on Emerging Threats and Capabilities back in March, he is vexed by today’s “ambiguity on the nature of the conflict, the parties involved, and the validity of the legal and political claims at stake.” All these complexities, he said, “require us to invest time and effort in ensuring we prepare ourselves with the proper capabilities, capacities, and authorities to safeguard US interests.” That’s a lot to ponder. But the HASC’s “Emerging Threats and Capabilities” panel sounds a lot like our own “Emerging Defense Challenges” Initiative here at the Scowcroft Center, so this got me thinking—just what new structures are needed to think through the problems of modern war?

In just the past week, we’ve heard plenty of ideas. Chairman of the Joint Chiefs General Joe Dunford said that he needs a whole new staff to think about trans-regional threats. In the Wall Street Journal, former top SEAL Admiral Eric Olson wrote that the military needs old-school OSS men—“Ph.D.s who could win a bar fight.” Army Chief of Staff General Mark Milley is considering “Training and Advisory” brigades to try, one more time, to teach Iraqis how to fight, and how to care enough to fight. Dean of the Fletcher School Admiral James Stavridis and our own Harlan Ullman argued in Defense News for a revolution in military education. Under Secretary for Policy Christine Wormuth wants to crowdsource strategic analysis. By the request for information, she wants a team of “50-100 individuals [including]  at least 10 experts from each of the following disciplines: social sciences, science and technology, economics, politics, and functional expertise.” Oh, and lots of regional knowledge and “demographic diversity” too.

But as my colleague August Cole observed, those parameters are rather prescriptive, which is why it’s not actually a crowd. It’s more of a super-panel. The problem is that its deliberations, as with much discussion around Washington, DC, could devolve into BOGGAT: just another—this time bigger—bunch of guys and gals around a table, bloviating about what’s important. Our colleague Alex Ward has observed that, by the standards of the town, any panelist can sound intelligent—while totally evading any question—with a simple tripartite spiel. Tell the audience that we need a regional, whole-of-government, strategic approach to the problem. It’s at once impossible to argue against, and wholly unhelpful in any context.

The crowd-panel is at least a step past the Jasons, taking in a far wider range of expertise than just the comfortable physical sciences. But maybe, some folks are thinking, more people doing more talking and staff work is not the answer. Senator John McCain specifically wants to reef back on the “sprawling mini-Pentagons” of the merely regional joint headquarters that produce big briefings, but don’t so obviously win the small wars. Dunford, for his part, does admit that his Joint Staff will shrink overall next year. So what’s the constructive alternative?

One fresh hell these days, as Evil explained in Time Bandits, is “computers.” Indeed, the US Air Force wants a supercomputer just for navigating its bureaucracy, a scheme I recently compared to Project Cybersyn, “the brainchild of Salvador Allende’s Marxist government for centrally managing the entire Chilean economy with a pair of IBM 360 mainframes.” In the Journal of Strategic Studies, Kareem Ayoub of Oxford and Kenneth Payne of King’s College have just written about “Strategy in the Age of Artificial Intelligence.” Learning machines, they believe, will “in the very near future, have a profound impact on the conduct of strategy and will be disruptive of existing power balances.” If one military power can better leverage all that computing power, then Deputy Defense Secretary Bob Work will have his Third Offset strategy. How this will improve the formulation of strategy is less clear.

That’s because the problem frequently is not how to apply more raw power to previously defined questions, but what questions to ask in the first place. Building a faster-thinking machine may help, for example, warships survive masses of inbound missiles, but they won’t help strategists figure out whether the outbound missiles will ultimately blowback as “North Atlantic scrod with Quaker State.” In that event, as my friend Dan Ward might argue, maybe the Third Offset will be in danger of devolving into the First Order. This bigger-computers strategy has a cinematic quality about it—if you can’t win the war with a battle station the size of a small moon, just build one into a whole planet. I’ll have to watch the movie to find out how that goes, but I think that I already know.

James Hasik is a senior fellow at the Scowcroft Center for Strategy and Security.

Related Experts: James Hasik, Harlan Ullman, and August Cole