The rise of remote work has prompted companies to embrace social AI to boost employee engagement and productivity. Is this cause for celebration or concern?
Like it or not, remote work is the new normal. Covid-19 has seen to that, forcing organizations to embrace the distributed workforce. The abruptness of the shift has provided little chance for reflection and evaluation. Is remote work a good thing?
The answer to that question depends, of course, on whom you ask. From the employee point of view, the picture is certainly mixed. We’ll consider the evidence, but first, some numbers: prior to the pandemic, 30% of employees, according to management consultancy Gartner, worked remotely at least some of the time. During the pandemic, that number has risen to 75% and Gartner projects that when it is once again safe to go into the office, 48% of employees will continue to work from home at least periodically, if not permanently.
Why, I wonder, will only 48% of employees continue to telework, and not the full 75% who do so today? Could it be that remote work is not the utopia it is sometimes depicted as, and that it does, in fact, leave much to be desired? The evidence of the last several months does not point to any one, clear-cut conclusion. While many workers have spoken about the positive effects of working from home, like being able to achieve a better work-life balance and see their kids more often, others do not share that sunny outlook. Take Reed Hastings, the CEO of Netflix (one of the tech companies that has, incidentally, benefited greatly from global lockdowns), who described remote work as a “pure negative” without any redeeming factors.
As it turns out, many employees share Hastings’ sentiment. A report compiled by IBM’s Institute for Business Value, in which 50,000 people from eight countries were surveyed, found that while in 2019 the vast majority of workers harboured dreams of one day working remotely, the reality has been far from what they expected. Overwork, tiredness, a sense of disconnection and of being left without proper guidelines or support; these have become the common complaints of the remote worker in 2020.
What is an enterprise to do? After all, “a happy worker is a productive worker”, or so the saying goes. Indeed, that phrase is not just an HR platitude but a hard fact. Research carried out by academics from the LSE, MIT and Oxford University found positive correlations, and even causal relationships, between employee wellbeing, employee productivity, and firm performance. The team behind the report writes, “recent experimental evidence suggests that a meaningful increase in wellbeing yields, on average, an increase in productivity of about 10%.”
That 10% productivity boost is not something any business can afford to miss out on. But can it realistically be achieved when remote work throws up so robust a challenge to employee wellbeing? Apparently it can, or so a new wave of tech companies would have us believe. Their solution to the challenge comes as a package of social software and artificial intelligence.
Is this a case of throwing more technology at problems created, in the first place, by a creeping overtechnologisation of our working lives? Or is it a smart, forward-looking response to a problem which will only become more acute with the passage of time? Let’s reserve judgement while we consider what this new wave of companies has to offer.
Employee Engagement Tech and Social Artificial Intelligence
First, a disclaimer: there are many more companies offering software to enhance employee engagement and wellbeing than can be discussed here, so what follows is only a representative sample. With that caveat out of the way, let’s get to the good stuff.
One of the newest entrants into the employee engagement market is Atium. Less than a year old and hailing from Amsterdam, Atium’s purpose, in its own words, is to, “design, build and provide activities that build trust, mitigate conflict, enhance communication & improve collaboration.” Sounds like the complete package. But what does that snappy phrase translate to, in practice?
Atium is still in private beta, so while we don’t necessarily have the full picture, what we can see is that a big part of the company’s offering centres around games. This takes the form, for example, of what Atium calls a “GIF Tournament”, whereby colleagues get together to search out GIFs, then vote for the best one. Trivial though that may sound, like all team-building activities, there is more to it. As Atium puts it,
“By choosing GIFs you think are best and competing against colleagues, you send out signals of a particular culture, values and humour, which helps to strengthen your team’s sense of belonging.”
The other activities baked into Atium like “Fact Bucket” – where colleagues share facts about themselves – are similarly underpinned by the pursuit of high-minded team-building values. And, as is to be expected these days, the whole Atium package is delivered with the help of a chatbot named Athena, who provides automatic prompts to employees on how and when they should be socializing with their colleagues.
The chatbot approach has also been taken by LEAD, a San Francisco-based company that is also seeking to stop social distancing from turning into social isolation. The essence of LEAD’s offering is a bot that operates within work communication apps Slack and Microsoft Teams to match employees and introduce them to one another. Unlike Atium, once two workers have been introduced, there are no games to play or set rules for how an interaction should proceed. Coworkers are simply left to set a time to talk, before the LEAD bot later checks in to ask how the meeting went. That data is fed back to the HR department – HR can tweak the rules for which employees are matched, and how often.
“Remote-friendly, purpose-built programs that create human connection” is the promise from Donut. This might be starting to sound all too familiar, but Donut does differentiate itself through its offer of no less than 12 different programs to connect employees. Mentorships, coffee meetups, code reviews, “mistake parties” for sharing stories of life’s mishaps and missteps; the range of options on offer is perhaps a reflection of Donut’s comparatively longer time on the market (the company was founded in 2016).
As with the other pieces of software we’ve discussed, Donut offers admins the power to tweak programs according to their preferences of who should meet with whom and how often. One of the company’s most recent innovations is Donut Watercooler, a feature that uses – you guessed it – a chatbot to post random questions in a Slack channel. The idea, as Donut puts it, is to elicit responses from employees in such a way as to “recreate office banter and bonding.”
Wellbeing and Social AI
So, three companies, three pieces of software to break through the digital divide and foster connection, one employee to another. We here at WAIRI have not yet hands-on tested any of these products and so have no grounds to evaluate how effectively they fulfill their stated mission – we will keep you posted as we do!
What we can ask, from a human wellbeing side, is: to what extent and in what circumstances it make sense, fundamentally, to use software and artificial intelligence to try to influence human relationships inside an organization? After all, what is common to all three of the companies we have featured (and to the “employee engagement” industry as a whole) is a shared commitment to algorithmic social matching or what we might term “social AI”.
Is social AI problematic? Proponents of the technology will argue that algorithmic matching can deliver big benefits to companies as an automated team-building and culture-creation tool that works at scale, provides measurable results, requires minimal human oversight or intervention and is effective for both in-office and remote workforces. For HR executives who need to meet budgets and demonstrate progress towards targets, that is surely a difficult proposition to ignore, whatever the downsides may be.
But am I the only one (Rahim speaking here) for whom AI-powered employee engagement tools conjure up a picture of HR managers in a control room, observing workers as scientists might observe experimental subjects, all the while tweaking algorithms to try to achieve the “perfect” – happy, motivated, productive or insert your own adjective here – employee? For all the positivity and problem-solving intent of employee engagement companies’ marketing, one could read here a subtext that is all about social engineering.
On the other hand (says Ana), what if social AI could tackle and truly relieve some of the burden of social isolation, and foster human connection, which is known to be one of the most important factors of human wellbeing?
This mixed reaction goes to show the delicate balance social AI systems have to face, because they are addressing, in this case, such a human problem. A mix of human+AI solutions might be the answer when it comes to social AI.
For example, instead of two colleagues being introduced to one another by a chatbot that has determined based on keywords that they might be a match, that introduction could be made by a real person from within the company who is already a friend to both parties? (the common human link could still be searched via technological means) Would those two newly introduced colleagues not then invest more in that relationship and try harder to make it work, knowing that they have a friend in common who is also invested? And couldn’t that mutual friend who made the introduction also make an active effort to sustain the relationship between the two colleagues? (perhaps being reminded by an automated system).
Of course, the essentially timeless human problem social AI aims to solve is sound: how to forge a group of strangers of disparate backgrounds and preferences into the single, cohesive units that we call teams, departments and companies.
We are all more productive and perform better when we work in teams with people we know and actually like.
But to ignore the consequences of introducing artificial intelligence into the delicate, human ecosystems that all organizations are would be shortsighted.
We believe all companies aiming to develop or implement Social AI solutions would do well to dedicate some deep thought to its implications. Rapidly choosing a start-up offering or, as a start-up, jumping on the easiest technical solution to be an early AIaaS arrival to the new Social AI market might not make for a human wellbeing friendly experience.
A true scientific approach needs to be applied to some such solutions – a deep understanding of what the research literature on human bonding has to offer, consulting sociologists, running experiments and refining towards human wellbeing. This scientific approach has to also be mindful of requirements for fast and agile product development, testing and validation, which can make the difference between start-up success or failure.
We at WAIRI can support you with this, and work with you towards solutions that put the human wellbeing principles in the center of the technology.
Because this is such a delicate topic, we are curious about what you believe – will Social AI provide value in the future, helping you find those special someones that are your tribe, your family away from family, or does it remind you of Orwellian scenarios?
Subscribe and let us know ….
Authors: 93% Rahim Rahemtulla – 7% Dr. Dr. Ana-Maria Olteteanu