Building the future into what you do today
In my CA World ’17 keynote I talked about how enterprise IT must recast itself to sense, adapt and respond to a rapidly changing world.
Technological change has a way of sneaking up on us.
If I had asked you twenty years ago if you’d own an electric car someday, you would probably have said no. Ten years ago, you would have cautiously said maybe. Today, you would either say your next car could be electric, or you already own one.
That’s disruption. It’s when something that you never really believed would happen actually unfolds in front of your eyes.
I believe that by 2030, what we now know as “IT” will be gone, or it will be virtually unrecognizable. Today, your software IS your brand, and your survival depends on learning how to build and deliver it well. But enterprise IT is struggling to recast itself to match the speed of technological change.
In the midst of this, how can you possibly make decisions around what will happen in the next year, let alone five or ten years into the future?
There are three things I think about.
First, I look at today’s environment, and then look backwards. The rate of change of any vector can only be determined if you have at least two points. If you have multiple points, you can estimate acceleration. The past is often the future’s best predictor. Take containers. Over the past two years, we’ve seen a very steep adoption curve. That helps me predict where containers are headed two years from now.
Second, like any responsible scientist, I must account for uncertainty in my model. When I project the curve of change out into the future, it’s just an informed guess about what I believe will happen. When Jobs and Wozniak were still in their garage, the future was not knowable to the IT guy running a System 360. Yet the work in that garage would change that IT guy’s world forever.
Third, once I get a read on where technology is likely to go in the future, I start building for that future today. This is where a lot of companies get into trouble. They either take a leap of faith and tackle something that nobody even needs yet, or they build something that becomes almost immediately obsolete, or is simply an attempt to catch up to the competition. You need to start with what you believe about the future, and then iterate with your customers to shape and re-shape what you deliver.
The Modern Software Factory is the single most important enabler of your business. It’s a way to sense, adapt, and respond to a rapidly changing world. Let’s take a look at where the Modern Software Factory principles are taking us in the future:
Data and analytics will revolutionize agile as we know it. In the future, agile without data analytics will be like fuel injection without oxygen. And data scientists won’t just throw insights over the transom to scrum teams – they will become integrated into a highly granular and fast-paced process for creating value. Sophisticated insight-generating engines tied into real-time business metrics will drive this new model.
Automation is all about increasing the throughput, efficiency, and quality of your factory. This happens today through things like continuous testing and integration, and business process automation. Automation will accelerate your development but only if you standardize and integrate workflows smoothly across the DevOps process and toolchain. The future isn’t about brute-force manual automation; it’s about intelligent automation that learns, adapts, and constantly self-optimizes the entire system.
There’s a lot of hype out there about AI, but AI is not one specific thing. You don’t “do AI” – it’s essentially a set of algorithms expressed as code operating on data. AI and machine learning represent a fundamentally different approach to software development. In the old model, we coded machines to achieve hard-coded outcomes. With machine learning, we give them examples of outcomes so they can reach conclusions on their own. I believe that machine intelligence will finally deliver on the promise of big data, and will increasingly find its way into everything.
I like to think of security as simply “trust”. It is the foundation of your brand and your relationship with your customers. As AI gets more sophisticated, it will become an intelligent hacking actor and part of our security threat landscape. The challenge will be to use AI effectively to guard against its malicious use.
The basic things that are at risk in the enterprise today will still be at risk tomorrow. And as with automation, we must build truly intelligent security systems to guard against tomorrow’s threats.
The real shift toward intelligent learning systems is just beginning. AI and automation can help us if we use them in the right ways. Let’s be clear: it’s not a choice. We are never going to have enough skilled people to do the work that needs to be done. According to research from the US Bureau of Labor Statistics, by 2020, there will be a million more jobs available in computing than applicants who can fill them. AI and automation are going to be central to helping us do more with the skills that we have.
The Age of Cognition is coming, and we need to be ready for it because it’s going to change how we work and what we work on. The efficiencies we gained by going digital over the last two decades will be harvested through the use of AI and machine learning.
We need to understand and accept that the way we will organize work in the future will be shared with our technology. That means elevating technology from “tool” to “partner” and keeping our eyes open through this evolution.