The media has raised consumer expectations for artificial intelligence (AI) by proliferating visions of intelligent assistants and self-driving cars, making the current AI hype cycle louder than ever. But how does this hype cycle compare to others over the last five decades? If the analysts are to be believed, we now find ourselves once again at the ‘Peak of Inflated Expectations’ – with the impending ‘Trough of Disillusionment’ ahead.
However, while these AI visions are some ways off from becoming reality, AI is still a critical technology, one which could contribute $15.7 trillion to the global economy by 2030. Increased development of machine learning algorithms is leading to benefits being felt across industries – from manufacturing to financial services. Organizations in the HCM space are taking great strides towards AI, particularly in the shared service function using smart software, but these organizations also need to be vigilant, and ensure they’re walking the walk before they talk the talk.
Implementing AI requires a new skill set within the HR department to gain greater technical understanding – for instance, considering how virtual assistants can add value to an employee experience and the appropriate moments to deploy them. This means HR must have a clear idea of what constitutes true AI and how it can positively impact their employees, before investing in the infrastructure that allows them to deliver it.
How AI will affect employees
It’s easy to be skeptical about AI, but I’m not. I see huge potential in how the technology could revolutionize the employee experience by transforming the relationship between employee and employer. For example, AI could play a critical role in employee benefits by educating employees on their benefits and helping them make the right decisions.
Our recent Employee Benefits Watch 2016/17 research found that two thirds of employees are eager to hear personalized information about relevant benefits at important life milestones such as marriage or childbirth. AI provides a means of achieving this level of personalization on a huge scale. AI can more effectively use massive amounts of datasets on details such as an employee’s demographic, interests, health, etc., to work out which benefits are likely to be most appropriate and helpful for them. We’re already making steps towards this, using data analytics to help target benefits communications to specific employees’ profiles. A new parent, for example, could automatically be served with information on their employer’s parental leave program.
While integration of AI into HR and benefits systems could improve the employee experience and ensure they’re better served by their organizations, some employees may have concerns. Personal data and how it’s used by organizations, along with data protection and security, is high on the public and political agenda across industries and employers need to be sensitive to this. It’s paramount that employers are able to explain how this is enabling them to create tailored ‘services’ to meet their specific needs and answer all questions on exactly how their data is being used.
The employer benefit
AI wouldn’t be generating such interest if the advantages were purely one-sided – it also presents a huge opportunity for employers. Benefits are emblematic of an organization, signifying how supportive it is of its employees and how committed it is to streamlining their work and personal lives. The way that benefits are delivered plays into this. The speed and accuracy with which AI could deliver or edit benefits (for example, automatically adding new-born children to medical insurance when maternity or paternity leave is taken) would reflect positively on an organization.
Another area where AI could be particularly useful is question management, quickly responding to issues experienced by employees rather than leaving them in a lengthy service center ticket system. Integration with consumer technology could even enable employees to easily ask questions while at home – they could conceivably ask Alexa for information about their wellness program, for example. With Bersin by Deloitte’s latest research indicating that the most mature HR functions spend only 29% of their overall time on transactional work, AI could help those less mature companies that are lagging behind become more efficient and effective.
Finally, AI can lead to improved benefits design. The Thomsons Online Benefits proprietary platform, Darwin, generates data from nearly 2,000,000 users. By crunching this data and learning how employees’ benefits behavior changes over time, AI could anticipate the perfect benefits program for any individual. It also offers us the potential to transition away from predictive to prescriptive analytics. Health insurance premiums are a great example. We can already draw on past data to predict the health needs of our workforce in five years’ time – but to be truly useful, analysis needs to factor in external trends, such as increasing obesity, longer life expectancy, etc. AI has the potential to do this by analyzing larger quantities of data at greater speeds to generate more meaningful insights than a human ever could.
AI’s future: How do we get there?
Those organizations that will be able to deliver on AI and analytics five years down the line are those that are making changes now. We have for the last few years been making changes in our data architecture to make it easier to analyze the rich data that we hold.
While expectations of AI may be inflated at present and tempered by concerns over data-use, the potential in artificial intelligence is remarkable. In the years to come, it will likely surpass doubts and bring about massive, positive change. If I were to ask one thing of my fellow technology professionals, it would be not to over-promise on AI. Doing so will only destroy confidence in the market and delay the advantages it can bring. Instead, companies should make long-term plans for its adoption and build the infrastructure for its success. It’s only then that we’ll realize its super-human benefits.
No comments:
Post a Comment