The pace at which collaboration automation tools and platforms have proliferated over the past decade has been breathtaking. But sometimes it takes all the running you can do just to stay in the same place. Even as these tools have made collaborating easier, the sheer volume of data we manage has grown even more exponentially. Even with the best tools, people can hardly keepup. The solution? Augment the humans with artificial intelligence!
How many emails do you receive each day? Do you consider that volume to be a heavy load, or a reasonable number?
Go back five years and ask the same question. Then go back ten. Fifteen. If you kept records of it, you’d likely find that the number has increased significantly steadily along. What seems like “a light day” today probably seemed like a mountain not long ago.
A May 2018 Forbes article tells us that 2.5 quintillion bytes of data are created each day and that over 90% of the data in the world was generated over the previous two years. Not only are we managing more data, we’re also communicating more. The same article reports that we send 16 million text messages every minute and 156 million emails, plus over 103 million spam emails. The article forecasted that by 2019 the number of emails sent every moment would increase to 2.9 billion.
The Tools Can’t Keep Up with the Traffic
In the last 35 years we’ve gone from brief, text-only email to rich content messages. From console-to-console text communication to global social networks. From kilobytes to megabytes togigabytes to terabytes to petabytes and more. From dBase II to Hadoop. We’ve gone from text to hypertext to voice to video to app sharing and whiteboarding with more to come.
Today’s collaboration platform integrates everything. You almost never have to leave the collaboration platform to use notetaking, calculating, document preparation, presentation production, and other applications. And you can do it together, with many participants all working on the same document together at the same time. The unified communication (UC) tools that support superior collaboration just keep getting better and better.
The question we now find ourselves facing is whether those UC tools are growing fast enough, or if the volume of data is exceeding our ability to keep up.
The Answer is in Measuring Productivity
A March 2017 McKinsey report tells us, “In the United States, productivity growth has declined sharply since 2004, yet digital technology has been widely apparent during this period.” It goes on to mention that productivity growth skirted negative territory in 2016.
Fifteen years in which the collaborative tools designed to increase the growth of productivity, yet that growth declined. Similar reports from the UK and elsewhere cite similar shortfalls.
The growth in sheer data volume simply outpaces the improvements in productivity created by better technology. This begs the question as to what we do next to overcome this.
Augment the Humans
While it sounds somewhat like the plot for a science fiction movie, today’s artificial intelligence (AI), machine learning (ML), and cognitive technologies make it a reality.
In any processing of information there are simpler preparatory and foundational tasks that must be performed before any actual analysis can take place. Collation, aggregation, correlation, interpolation, extrapolation, sorting, and other basic steps are necessary to enable higher-level work to be performed. Why not assign these simpler tasks to AI engines? Any proportion of the overall work that is defrayed by AI results in productivity gains.
Evidence abounds. Users of Microsoft Office 365 send and receive email using the Outlook application. Long-time Outlook users remember when everything came into their inbox and they had to first separate the spam and insignificant messages from the important ones. Only then could they start to read and act upon the remaining email.
Today’s Outlook mobile user enjoys the AI services of the “Focused Inbox”. Using AI and ML techniques, Outlook identifies messages it assesses to be low value and places them in the “Other” inbox. Only significant messages, according to the AI, go into the “Focused” inbox. Since users have the option to reassign messages from one inbox to the other, the ML engine quickly learns more about what constitutes “significant” based on the user’s behavior and improves the validity of the inbox assignments. Users save tremendous amounts of time not having to perform this repeated filtering themselves.
Citrix Systems has taken this much further embedding AI throughout their intelligent WorkSpace environment to provide as much augmentation as possible to eliminate the need for humans to perform many preparatory and routine tasks. This has significant implications for users of every data stream that courses through WorkSpace, including email, collaboration platforms such as Microsoft Teams and Slack, social media networks and more.
These may seem like simple examples, but they illustrate the overall strategy. Outlook and Citrix WorkSpace have become augmentations of the human user’s judgment. Many data workloads can be similarly assessed and evaluated by AI and ML, reducing the time required for human intervention to be slashed by filtering data it deems to be less useful and prioritizing the more important elements of the workload.