Some people are very concerned that the automation enabled by AI will do away with jobs. This concern seems to make sense on the face of it. For instance, when is the last time you actually went into your bank and spoke to a human cashier, instead of just using the ATM outside?
Fortunately, the impacts of automation on human staffing are not quite that cut and dry. Sure, ATMs automate many routine tasks today that human cashiers used to exclusively perform. But when they were first introduced, the machines augmented the customer service and convenience capabilities of the bank. Their value was promoted as offering things like after-hours service or convenient money access where a physical branch didn’t exist — rather than as a replacement for humans. Bankers were then able to focus on more complex transactions, such as loans, mortgages, merchant deposits, and the like.
This increase in capabilities actually led to an increase in the number of human tellers, even during the rapid roll-out of ATMs:
At the dawn of the self-service banking age in 1985, for example, the United States had 60,000 automated teller machines and 485,000 bank tellers. In 2002, the United States had 352,000 ATMs—and 527,000 bank tellers. ATMs notwithstanding, banks do a lot more than they used to and have a lot more branches than they used to.
Contrary to early sci-fi depictions of humanoid robots capable of replicating the full range of mankind’s talents and capabilities, what we actually got was a variety of single-purpose machines, each one only capable of performing a single task. This did indeed replace some human jobs, but generally the jobs that were replaced were the ones which had already been broken down into atomic tasks, such as on assembly lines.
As technology continues to develop, new areas of employment are being affected. Automation is getting into much more customer-facing industries, such as the front desks of Las Vegas hotels:
As we eliminate front desks altogether and … you’re able to check in over your phone. But the people that were at those front desks are still there. They’re there to customize your journey, there to greet you at the car, and they’re there to escort you to your room to make sure you have everything you need. You can’t replace that and that engagement.
So checking people in is only one very routine task of that job. Front desk agents also act as adjunct concierges: making restaurant bookings, giving tourist advice, and performing hundred other little service tasks that provide added value for guests.
It turns out that the process of issuing guests room keys is a mechanical task ripe for automation. The hotel would far rather all of that were taken care of without tying up expensive human employees and a chunk of prime real estate in the lobby. Meanwhile, the guests are also happier, because they don’t have to wait in a line after a long day of travel, as their room key is already on their smartphone.
Enterprise IT groups would do well to learn from these consumer industry examples to avoid being replaced by automation or outflanked by shadow IT. The specifics of each situation are going to be different, but here are some suggestions.
Humans Are Valuable; Don’t Waste Them
Employees are too expensive to waste on single-task jobs. Back in the day, one of the routine tasks of junior systems administrators was to swap backup tapes. This task took up maybe an hour of time spread over a week. However, I recall one company’s central tape vault wasn’t staffed by an army of tape jockeys; it had a very impressive tape robot.
Automate any simple tasks which are taking up too much of people’s time
Don’t Forget to Plan for the Impacts of Automation
Going back to that hotel example for a moment, let’s think about how all those automated systems are going to work a few years down the line. One hotel in Japan found out the hard way:
Japan’s Henn-na “Strange” Hotel has laid off half its 243 robots after they created more problems than they could solve. Many of the robots that have been retired were in service for years, making them outdated. The hotel decided it was easier to fire them than to replace them, citing high costs. In the end, a lot of the work had to be left to humans anyway, especially when it came to asking more complex questions.
Don’t just consider day one; think about long-term operations. Bringing this back to IT, even if you could freeze everything and take a perfect snapshot in your CMDB, what is your plan to stop that snapshot from drifting further and further from actual reality? If you commit 100% to a single vendor across your entire stack, what is your backup plan if that vendor changes its roadmap, gets bought, or goes out of business? And what resources have you set aside to support all of this?
Become Visibly Proactive & Inventive
No savings will be achieved by a hotel that just takes a bunch of employees who are no longer needed on the front desk and sticks them in the back office. Instead, clever hotel managers will extend the concierge desk, the valet desk, or other front-of-house jobs. These all add customer value and apply the savings of automation in time/budget/resources to help build brand loyalty. AI in the workplace provides the scope for out-of-the-box thinking. One idea: give employees iPads and set them roaming the property, actively looking for guests who might welcome assistance.
The same thing can be applied in enterprise IT. Every IT manager has a long backlog of things they wish they could do – or even things which they know they should be doing, but simply can’t get to. The right question is not “which jobs can you eliminate?” but “which additional jobs can you do?”. A big part of that is making IT proactive instead of being reactive. Don’t wait for users to come to you with a specific request; instead, engage with them to understand what they need to achieve, and propose solutions that they might not even have considered.
Forget the Next Big Thing; Look for the Next Little Thing
Once we have automated away the low-hanging fruit, what can we do next? Staying with our hotel metaphor one last time: checking guests in on smartphones only recently became possible, but card keys provided value long before. Instead of waiting to replace mechanical locks until the whole ecosystem was in place to support mobile check-in, hotels adopted card keys and embraced all their possibilities: custom branding, access to different areas of the hotel (gym, members’ lounge), and even payment for in-house services. Now all of that is moving to mobile, but how much was saved in the meantime by using the stepping-stone solution?
There is a saying: “the perfect is the enemy of the good”. This is particularly true for IT professionals, who are prone to searching for a Grand Unified Theory of Everything, instead of focusing on piecemeal improvements that can be made quickly. Take a leaf from developers, who work with toolchains of diverse products, and upgrade or replace individual links in the chain as and when it makes sense. The savings from an improvement in one place can be applied to another part of the chain that needs attention.
The bottom line is that there is plenty for humans to do, in hotels as well as enterprise IT. Deployed right, AI will help them do it, and do more of it, better. The results can be impressive, both in hard economic savings, and in softer terms of user satisfaction.
AI deployed for the wrong use case and without thought for the longer term can be counter-productive, sucking up scarce resources without much return. AIOps is AI focused on a particular task – in this case, IT Operations – in a way which augments human capabilities and opens up new possibilities for engagement with end users. That is the way to achieve lasting value.