As today’s digital disruptions bounce and smash their way through conventional technologies and conventional wisdom alike, predicting their path is a multifaceted challenge. So many areas of technology advance on Moore’s Law-like exponential curves that divining the future is fraught with danger.
Such is the problem with artificial intelligence (AI) and its related concepts, including cognitive computing, machine learning, and deep learning.
Today, we have interesting but hardly revolutionary technologies like Apple Siri and Amazon Alexa. And tomorrow? If you listen to the doomsayers, The Terminator’s Skynet or its killer robot kin will take over the world from us puny humans.
While we cannot entirely disagree with the dire warnings of certain pundits, the reality is that there is quite a lot of territory to cover between Siri and Skynet. Even well before some kind of singularity where AI begins to think for itself, we face enormous disruption across the board — from the consumer marketplace all the way to the enterprise.
In fact, the writing is on the wall. AI will completely upend how we build, use, and even conceive of software. The AI we have today merely scratches the surface — and long before we invent Skynet, AI will change everything.
We all love to hate auto-complete — the AI built into our phones that tries to guess what word we’re typing or what word or phrase we might type next.
It’s true that auto-complete could work much better than it does today. And in fact, it will. Much better.
We’re already seeing auto-complete-like technologies in enterprise applications. For example, some modern business process designers suggest the next best action, as the user assembles a process. We’ve also seen a visual data integration tool that suggests the next step in a complex integration.
The low-code/no-code space is picking up on this trend, as well, building AI wizards that discern the intent of the application creator, thus fleshing out as much of the completed application as it can automatically.
As this AI matures, we can expect it to autocomplete more and more of the task the user is attempting, pulling in relevant contextual information as necessary to fill in any gaps that human beings have heretofore had to complete.
For example, it won’t be long until data entry is a thing of the past. Why type in a person’s information into a form when it is sufficient to tell the software what information you’d like it to enter and let the software find it wherever it may be and enter it for you?
Among the voluminous quantities of Steve Jobs lore of questionable veracity is the story of the iPod design meeting. The team was coming up with many complex interface designs when Steve walked in, drew a rectangle with a circle in it, and behold, the iPod was born.
The insight: If a single button will do, why have anything more complicated? The same goes for AI. As AI advances, how many complex tasks can we boil down to the single click of a button — or no click at all?
Complete a form, even a complex data entry task in an ERP application? Click a button and done.
Close the books for a multinational bank? Click a button.
Comply with a regulation? How about complying with all regulations? Click and done.
Recruit, hire, provision, and orient an employee? All it takes is a single click.
It’s important to note that while clicking a button may be literal, it might also be a metaphor for any ultra-simple, single user action. It might mean saying, “Siri, buy me that house” (mortgage included). Or perhaps waving your hand in an augmented reality interface to, say, enroll in college.
Consumer examples may be easier to envision, but enterprise scenarios are potentially more powerful because existing enterprise tasks tend to be more complex and time-consuming.
Essentially, any task or process that boils down to busy work or paper-pushing or broadly speaking, bureaucracy should be as simple as pushing the proverbial button — and soon, AI will make this vision a reality.
Furthermore, for many tasks, even pushing a button is too much trouble. Take the example of complying with regulations. Our AI should simply know that we must be compliant, and proactively take whatever action it must to ensure that we are so – fully automated, no button pushing necessary.
The Moral Dilemma (Hint: It Ain’t Skynet)
The future path of AI leads us to a grand utopia free of busy work and paper-pushing. Only one problem: busy work and paper-pushing give people jobs. Such careers may be mind-numbing to be sure, but they pay the bills.
The moral question around AI of our day, therefore, isn’t whether we might create killer robots, but rather whether extending the inexorable progress in enterprise automation with AI is a good thing or not.
The answer? Such moral arguments, in the end, are irrelevant. Just as water flows downhill, business inevitably finds a path to maximum profitability — for better or worse.
True, some businesses may decide to forego AI in order to keep cube farms full of data entry personnel busy. But remember, we’ve seen this story before, many times.
What happened to the companies that said the same thing about their typing pools, eschewing PCs to keep throngs of typists busy pecking away at their Selectrics? Not only were such efforts futile, but led to competitive disadvantage and eventual demise for the companies that persisted.
Today, the increasing pace of technology innovation is driving global digital disruption. The timeframe connecting competitive weakness to abject failure grows ever shorter.
Failure to adopt AI as the technology matures may save some jobs in the short term, but how much more will be lost when the entire enterprise goes belly up?
While every wave of automation inevitably puts some people out of work, the bigger picture is how automation shifts the emphasis on the types of work companies require from their employees. AI-driven automation is no different.
After all, there will remain plenty of tasks within organizations large and small that AI is not well-suited to tackling. We’re all used to interacting with software, and yet as social beings, there will always be a role for human-to-human interaction, both in social and business contexts.
Busywork, however, is on the chopping block. If you’re an executive or knowledge worker or some other category of personnel whose value to your organization extends beyond busy work, your life can only be improved by eliminating it.
Imagine if you will a “deal with my email” button. Push it and voila! Your inbox is empty — or perhaps, almost empty. I say almost because there should always be emails (or other examples of human interactions) that you really need to deal with personally. The rest of them, however, are busy work.
To conclude, then, some advice: seek to build skills that AI will not be able to supplant, at least within your career timeframe. If you’re hiring someone, always look for such skills, regardless of the role.
Whether we like it or not, AI is going to streamline and simplify our lives, and every piece of software we encounter will be AI-driven. Resistance is futile. Instead, do what you can to maximize the upside, both for you personally as well as for the organizations where we work.
Copyright © Intellyx LLC. Intellyx publishes the Agile Digital Transformation Roadmap poster, advises companies on their digital transformation initiatives, and helps vendors communicate their agility stories. As of the time of writing, none of the organizations mentioned in this article are Intellyx customers. Image credit: Blake Emrys.