Paperless offices. Robots to do the heavy lifting. Databases full of information that will make life easier, more pleasant, healthier, and longer. These were the promises made (or at least implied) when society went down the road of all-out digital tech.
But the innovations digital technology has wrought have somehow made our jobs, our social relationships, and even our lives, in some respects harder, not easier… certainly more hectic. For many of us, it feels as if the algorithms are in control.
For example, instant, free, IP-based communication with customers on the other side of the world has been a great boon for many companies – but the employees who find themselves having to conduct business at all hours of the night (to match the customer’s time zone) might have their misgivings.
Another example: The huge amount of data a corporate employee has to contend with these days is just overwhelming. Data has exploded in the past five years – more of it was created during that period than in all of human history – and that data needs to be controlled and corralled in order to enable organizations to take advantage of it. The more of it there is, the more complicated sorting it out is, and the more likely we are to lose focus on what we really need to be doing.
In short, the digital lifestyle, in the office and outside, can be a mess. We should be looking at ways to fix the situation and ensure that we reap the rewards digital technology can bring us. One way we can do that is by using artificial intelligence.
In businesses today, huge amounts of data are being generated from all sorts of sources. AI machine learning-based systems are analyzing and extrapolating it to develop new insights, new information, and new data. More data should be a good thing; the more you can analyze, the more precise your focus should be on the goals your organization has, whether it’s in sales, opening new markets, product or service development, etc.
But it also means new and more tasks that team members have to go through in order to get their jobs done – and for many, the huge amounts of email, messages, files, PDFs, presentations, and much more that they have to process has reached a point of diminishing returns. In other words, the rush of data is not helping them do their jobs – it’s hindering them.
How do we know? It’s a fact that 73 percent of data that organizations collect is never used – and that’s because organizations don’t have the time or resources to deal with all of it. Instead of getting analyzed, the data just sits around in a repository, clogging up the works. Some of it may be very relevant to important projects – but because of that lack of resources, there’s no way to ferret out what’s relevant and what isn’t.
Is there any way out of this digital bind? To the rescue comes artificial intelligence; specifically, machine learning, the branch of AI that can find valuable connections and insights from large sets of data. Used properly, machine learning can help organizations make data management easier – ensuring that relevant data gets the attention it deserves, while extraneous data gets sent to its fate in the circular file.
Machine learning set loose on a project, for example, would study the ways team members interact with data and determine based on those actions what the best practices are for those teams. A machine learning system evaluating email use would observe several team members for a period of several months and see which messages they retain and which ones they delete. For example, when preparing a sales presentation for a client, the ML-based system would check out all the email messages relevant to the sale – customer purchase history, complaints and compliments, supply chain issues affecting the product or service being offered, and anything else staff has to know in order to ensure that they hit all the relevant points and prepare to address objections, problems, and issues in advance of the presentation.
Now, imagine that system being applied to all components of an organization’s data resources – not just email, but files, folders, social media sources, etc. A machine learning-based system could parse through all the data in all those resources, presenting the relevant information in a single screen. The company wins twice – first, all the relevant data for that sales presentation gets highlighted, including the “hidden gems” of data in storage areas that staff may not even be aware exists. And, staff members don’t have to spend hours searching for relevant data.
Did we say “hours?” Actually the situation is far more dire. A study by RingCentral shows that employees lose up to 32 working days a year just switching between windows on their screens, as they open and close folders, database windows, documents, spreadsheets, and more just looking for the information they need. Any company looking to justify the outlay for a machine learning-based system to deal with data density doesn’t need to look much further.
Are there downsides to ML? Of course; there are downsides to almost anything, if it isn’t applied intelligently. In the case of machine learning, some experts are concerned that it could make workers lazier as they come to rely on the “powers” of their intelligent systems.
One of the big benefits of an ML-based system is that it becomes more intelligent as it “learns” patterns of use, actions, and storage. The objective of data scientists who build ML-based search and analysis systems is, of course, to get as close as possible to zero errors and produce answers in as short a time as possible. Of course, nothing is perfect, and each type of machine learning system has its foibles – biases, bad data inputs, clustering issues, and others – but good segmentation, proper training sets, can improve the situation. But ensuring that staff remains alert when the machines begin to augment their work is likely to become more of a challenge, if our experience with technology in the past is any indication.
While machine learning can’t solve all our problems, it could be just the thing to help organizations get in control of their data – and of their time and resources. Ultimately, getting control is what it’s all about for organizations. When we lose control of the data, the data tends to control us. With machine learning-based artificial intelligence systems, we have an opportunity to take control of the data – and make it work for us, instead of the other way around.