by Dana Gardner
Just when cloud computing seems inevitable as the dominant force in IT, it’s time to move on because we’re not quite at the end-state of digital transformation. Far from it.
Now's the time to prepare for the post-cloud world.
It’s not that cloud computing is going away. It’s that we need to be ready for making the best of IT productivity once cloud in its many forms become so pervasive as to be mundane, the place where all great IT innovations must go.
That’s right, as soon as we’ve wrestled all the step-change benefits from cloud — and we’re close — it’s time to relegate cloud’s impact to simply modest, incremental improvements tempered by the providers’ commercial interests.
I’ve seen it before, as I’ve been a born-again cheer leader for each IT innovation heyday over the last 20 years — from OLE, web browsers, app servers, Linux, Java, n-tier, XML, LAMP, Web Services, REST, dynamic languages, to, yes, the lofty cloud models.
And as each of these mighty productivity improvements stood on the shoulders of the ones before it, its own barbarian at yet another gate, disrupting as well as improving, it’s tempting to think of the progression as somehow maturing, plateauing, coming to a close.
Cloud may be big, it may change everything, but it's not the end of change. And change is not slowing down.
In the fast-approaching post-cloud world, the new differentiator is in how well an organization procures and consumes many digital business assets simultaneously, dynamically, and in deep synergy — across a diverse and complex supply chain and easily reconfigured installed base.
Technology’s new poster child, artificial intelligence (AI), and its engine, machine learning, will be a major reason why cloud goes mundane, practically invisible, fairly soon.
To make the most of a spectrum of apps and data hosting options, many variables need to be considered, evaluated, measured, reassessed and implemented. Repeat. All the time.
To me, managing the complexity of such hybrid computing real-time procurement is a killer app of AI. These are not necessarily the skills resulting from completing a Microsoft certification process.
"I want the computers to actually optimize themselves and compose themselves so that I can be on the beach, and I will know that my infrastructure is changing, scaling and flexing to be able to meet the demands of whatever we throw at it,” a CTO of a highly technical company recently told me.
Oddly, like the cobbler whose kids have no shoes, IT infrastructure vendors seem to be dragging their feet in employing AI to manage IT complexity. Who yet is digesting and analyzing all their logs, learning the buying options of all the clouds providers, leveraging automation and orchestration capabilities, and applying an algorithmic approach to real-time, dynamic, microservices-level hybrid deployments?
If I missed you, let me know.
I nonetheless expect this very data-driven and productivity-optimized approach to hybrid IT procurement to outshine cloud in the next few years. Perhaps Wal-Mart would be interested in providing such a service? But I digress.
Suffice to say, mere humans, even with their swivel chairs well-greased, are not going to be able to pick and choose from their walls of whiteboard and sticky pads which cloud service to apply to which business requirements and then buy it as the best price as needed until it's not.
Manual IT practices and post-cloud do not mix. Instead IT leaders will become the authors and keepers of their IT-supporting-business-defining algorithms, which will both satisfy their businesses stakeholders’ requirements best — and achieve attainment of such satisfaction at the lowest cost, lowest risk, leanest organization, and most agile means possible. Think of it as AI-enabled minimal viable automated IT.
And there are other factors that point to the post-cloud era dawning.
Those seeking to innovate will always do an end-run around the common denominator in their markets. Major cloud usage will flip from being a first-mover advantage to being the status quo. When everyone uses the same IT formula, how to gain advantage? By being best at using more options for IT, including your own closely guarded datacenters for certain apps and data to gain the skills and intellectual priority to out-perform the pack.
Furthermore, when a single company overly captures a market, it never lasts. Should we expect Amazon to be the runaway leader in cloud for very long? If Microsoft and Amazon go to the mats for a long time, will they pummel each other into lower profits, lower expectations.
Already the tier-2 cloud ecosystem competition is adjusting. We're seeing how the latest advances in data center infrastructure are bringing the performance up and costs down for private cloud, high performance computing, software defined data center, hyperconverged infrastructure and the Internet of Things.
Data centers are rapidly becoming smaller, faster, tougher and cheaper per output than ever. They can be better managed from afar, moved easily. That trend will only continue, probably until price parity works in favor of the on-premises world because they can procure creatively, paying with use, still getting depreciation benefits.
As in any market, eventually, and I think quite soon, price parity will develop between private cloud and public cloud. Each will have their advantages, and so the bell curve on general use will go to where the risk is lowest, convenience highest, and ease of data portability inherent.
In other words, in the post-cloud world, movement among and between, and productive exploitation of, all hosting options for all workloads is the best core competency for IT operators.
I expect those operators to increasingly seek atomization of their IT infrastructure – via micro services, containers, server-less computing and more to come – while seeking to make them multi-cloud friendly.
Cloud may change a lot of things but economics, human nature, and the will of governments worldwide to regulate against commercial monopolies are not among them.
Welcome to the post-cloud world, same as it ever was, again.
Dana Gardner, founder and principal analyst at Interarbor Solutions and host and producer of the 11-years-running BriefingsDirect podcast and blog, has been covering and analyzing IT since 1988.