There’s been a great deal of interest in the potential cost savings and efficiencies of scope and scale promised by autonomous IT processes. Abhishek Bhattacharya, Global Head of Strategy and Planning at cognitive automation solutions provider Digitate adds some perspective for companies wanting to get started with autonomous AI technologies.
There’s been a great deal of interest in the potential cost savings and efficiencies of scope and scale promised by autonomous IT processes. However, well-publicized examples like the autonomous driving car demonstrate that there are still (and perhaps there will always be) some bugs to be ironed out. No one wants to be the one responsible for onboarding an under-baked technology that interrupts or interferes with business processes.
Given that context, how should companies get started with autonomous AI technologies?
Here are a few considerations to add to your perspective.
Understanding Autonomous AI and Its Benefits
Organizations of all sizes today are drowning in vast amounts of data. Data complexity is growing exponentially while hiring or reskilling workers to manage it remains practically a flat line. IT departments are grappling with many issues, all at the same time, along with outages and hacks. So naturally smaller issues get pushed down the ‘things to do’ list and often get lost in all the noise.
These issues are often symptoms of a larger problem not much different than a tumor growing inside an organization’s network: potentially manageable if detected early, but often escapes detection by tired and overworked eyes, thus causing systemic and irreversible damage. In te enterprise context, these damages could result in in data or business loss such as an eCommerce outage, a payment gateway failure, stock out in a retail store and so on.
Even if you could afford to hire all the experts that you possibly find, there aren’t enough skilled professionals to examine every signal coming from your IT system. To add some context here, IDC predicts that the collective sum of the world’s data will grow to 175 zettabytes in 2025 – that’s a compounded annual growth rate of 61%. There’s just no possible way that humans can grapple with all of this alone. Organizations of almost any size will have to implement some kind of automated system to grapple with this mass volume of data and signals. It could be as simple as a bot (short for Robotic Process Automation – RPA) or an upscaled, state-of-the-art AI system. Today’s cutting-edge systems include self-healing AI, which manages itself and only approaches an IT admin when it needs guidance or a decision to be made.
There’s been some backlash when it comes to automation, however, with many believing (erroneously) that autonomous technologies are taking away jobs. While it’s no secret that CFOs and CIOs are very interested in how automation will help save them money, it’s not that cut-and-dry. If history is anything to learn from, there’s an example of how the technologies and changes that arose during the Industrial Revolution were met with the same skepticism. But in hindsight, these innovations created a lot more jobs than they took away, and it freed up people to work on more meaningful tasks instead of rote manual labor. This seems to be the case for the ‘automation revolution’ that is upon us, as well. A Gartner report seemed to think that this year, 2020, is going to be the tipping point when AI starts creating more jobs than it takes away.
The nature of jobs will change, but automation isn’t the threat to employment that some are making it out to be. It isn’t that AI is taking jobs away; rather, it’s introducing a new smart colleague that frees IT staff to work on more meaningful, complex issues that truly require the ingenuity of a human mind.
Getting Started with Autonomous AI
Most driverless cars today still use humans in the driver’s or passenger’s seat, ready to take over if need be, but that’s not as easy with today’s automated or AI-enabled technologies. Today, the best way to proceed with automation is via a controlled trust-building exercise, deploying AI capabilities one application at a time, ideally starting with one that is not business-critical. As you start getting a grip on how the technology works, you can work your way up course-correcting when needed. This way you can see what insights it delivers, and make sure you’re comfortable with it before expanding it to other business towers.
Gartner analysts predict that AI and related technologies will replace almost 69% of the manager’s workload by 2024 – and that companies that aren’t using AI and automation in some form by 2022 will fall behind their competitors. Every organization may have a different appetite for AI-driven technologies based on their size, the industry they’re in, how outdated their IT is and other factors specific to their circumstances. But the question when it comes to adopting AI is no longer if or when but how.
Two best practices for successful implementation are:
- Have a change management strategy in place. IT professionals may feel like they’re being asked to give away the keys to the kingdom, even handing over their jobs to an AI. As noted above with the example of the industrial revolution, that’s often not the case. But to allay fears, it is a good idea to put change management in motion.
- Start small. Trust and visibility are important for stakeholder buy-in in any endeavor, and that is certainly the case for AI adoption within an organization. Most organizations are wary of the “big bang” approach – they don’t want to put all eggs in the same basket. Starting small in one area is often a wise approach. Go for the small wins, don’t expect a return on investment in the first year, and always have a plan to expand into high value use cases.
The Opportunity is Out There
The promise of autonomous AI is great, but there have been enough AI failures in the wider world – like with autonomous vehicles – that some are reluctant to test the technology out. But as Gartner points out, AI is rapidly becoming a business necessity rather than a “nice to have.” Rather than taking jobs away, treat AI as a digital colleague that handles repetitive work so IT professionals can focus on more valuable, high-level work. The best practices and recommendations above will help you get started on your AI automation journey. There will likely be some bumps along the way, but like most things in life, the journey is the destination as you move towards a better, smarter and more productive enterprise.