As we wait for public cloud to devour all enterprise workloads, the heady hype around driverless cars may prove instructive. Writing on the slower-than-expected driverless car revolution, The Wall Street Journal’s Christopher Mims has said, “we’ll have to adjust to the reality that autonomous driving could be headed for narrower–but still transformative–applications.”

“Narrower but still transformative” is a great way to describe what we can expect from public cloud computing, too. Not that public cloud isn’t having a massive impact on enterprise IT, and may actually come to dominate enterprise workloads. But, getting there will be gradual. Unlike in autonomous cars, the problem in cloud computing is less about tech and more about culture.

Autonomous autos…aren’t

Hubris being what it is in Silicon Valley, it’s not surprising that the approach to self-driving cars is an analog to open source’s “given enough eyeballs, all bugs are shallow” thinking. For Google, Uber, Tesla, and others, this thinking becomes: “Given enough data, all roads are self-drivable.”

Except, of course, that they aren’t. Not anytime soon, anyway.

SEE: Cloud migration decision tool (Tech Pro Research)

As Mims writes, “It turns out that the human ability to build mental models isn’t something that current AI can just learn, no matter how much data it’s fed. And even once we have the technology, we’ll still have to deal with all those unpredictable humans in cars, on bikes and scooters, and on foot.”

People complicate data. Remove those people from the streets, however, and we still haven’t solved the problem, as Mims noted:

Over a lifetime of driving, humans become expert at countless subtasks, from noticing distracted pedestrians to questioning the judgment of construction workers waving them through a work site. While much has been made of the total number of miles that various self-driving systems have racked up, conquering these little annoyances actually requires an enormous amount of intellectual labor by many teams of engineers.

Smart as an algorithm may prove to be on paper, real intelligence (even among we less intelligent humans) parses data much better than any computer can. The self-driving car future, in sum, remains a future endeavor, one that will take a long time to reach, no matter how much computing horsepower we throw at it. It’s a tough technical problem to conquer.

People and the public cloud

Meanwhile, public cloud computing won’t take over the enterprise tomorrow for a quite different reason: People. Early on, the suggested roadblocks to public cloud computing were mostly technical (latency, security, etc.). One by one these have faded and public cloud has turned out to be superior to on-premises infrastructure in most instances.

And yet, as Red Hat’s Gordon Haff has highlighted: “All public or all private strategies are rare, especially among large organizations.” Instead, he goes on, enterprises are “essentially all hybrid in some form.” Yes, there are companies that say they’re going all public cloud, just as there have been some pretending to wait it out. But ultimately, everyone ends up running hybrid infrastructure.

Why? Because people.

SEE: Special report: The cloud v. data center decision (free PDF) (TechRepublic)

As I’ve written, in any given enterprise there are, in fact, teams or business units that have gone all-in on public cloud. Within that same enterprise, however, there are other groups that lack the DNA or inclination to move to public cloud. Or, as one head of infrastructure, architecture, and design at an American financial services firm told Forrester (as quoted by Haff): “I think we’ll always be in a hybrid mode. We’ll always need internally hosted apps; I don’t think we’ll ever get everything into public cloud. We might move those apps to a managed service somewhere, but they can’t move to public.”

Perhaps there are legal reasons for that…today. I doubt those same legal reasons will persist. But the human reasons for resisting change, for wanting to keep some applications close, will absolutely persist.

This is why Red Hat has long been smart to stick to a hybrid approach, and why Amazon Web Services (AWS), Google, and Microsoft have all been equally smart to embrace it, if grudgingly (in some cases). Haff cites Forrester research showing that 88% of those surveyed plan to increase public cloud spending, and a nearly identical percentage (87%) plan to increase their private cloud spending. These attempts to replicate public cloud benefits within the firewall are generally Quixotic in my view, but they’re completely understandable.

Self-driving cars is a technology problem. Public cloud-driving infrastructure is a people problem.

It’s why the approach taken by Cloud Foundry–training enterprises to shift their thinking to a public cloud mentality–is so potent. But it’s also why we’ll have hybrid infrastructure for a long, long time.