Article · 7 minute read
‘It’s Still Day One...’
Dan Hunt looks at the ethics of using AI in Assessment
By Dan Hunt – 23rd May 2023
“It’s still day one…’
This will be a familiar phrase for anyone who has worked at Amazon, as founder Jeff Bezos used this regularly to remind Amazonians not only of the potential for the business to grow into a far larger and more complicated beast than it was at the time, but also to consider the long-term implications for what the internet could develop into and how it might impact our lives. Interestingly, for David Bowie fans out there, as far back as 1999 he took a similar point of view and was pretty vocal about the unlimited potential for the internet to alter society in ways it was hard to imagine (never waste an opportunity to shoehorn Bowie into a conversation).
Why am I quoting Jeff Bezos and David Bowie? Because there’s a new technology in town you may have heard of – Artificial Intelligence. It is barely out of the daily news cycle at the moment and really seems to have exploded into the public consciousness over the last 12- 18 months through the burgeoning popularity of Chat GPT in particular. Therefore, I thought it made sense to put some thoughts down not only on the potential ethical implications for Talent Acquisition and assessment, but rather to consider more broadly some of the implications of AI on the world of work.
As with the introduction of any new technology, there will inevitably be someone somewhere holding a placard suggesting the end of the world is nigh! And, human beings will have no use! Equally, there will be many commentators who will state that this technology will be of huge benefit to humankind, with trillions generated for the global economy and human beings finally unshackled from the daily grind to live their lives eating fruit by the sea – or whatever you’d like to be doing if you weren’t reading this at your desk.
The question then is, does AI represent a genuine existential threat to society? Or perhaps it is another in a long line of technologies, such as the printing press, the automobile, and indeed the internet, who’s impact on society certainly caused the decline of certain industries while creating new ones, but after an initial period of adjustment and disruption were more or less integrated into the existing social order and we will learn to get along with it.
AI in the Workplace
In relation to the potential impact of AI on the workplace – there is a whole dissertation’s worth of discussion to be had on this topic, and you are welcome to read mine. Coincidentally, as I sat down to write this, BT have announced a reduction of its workforce by as much as 55,000 by 2030, including 10,000 jobs to be replaced by AI. The Chief Executive, Phil Jansen has stated that “For a company like BT there is a huge opportunity to use AI to be more efficient”.
PWC have also suggested that while the potential boost from AI to the world’s economy by 2030 could be $15tn, also by the mid-2030s, 30% of jobs could be at potential risk of automation, while 44% of workers with low education could be at risk in the same timeframe. I highly recommend Daniel Susskind’s book “A World Without Work” for more research into what impact AI could have on the global workforce, if the topic is of interest.
…. So, what are we to make of these developments? Are we all going to be redundant in the next 10 years? Perhaps we will become a nation of hairdressers, carpenters and gardeners – and this is in no way meant to be condescending towards those employed in these roles. Not only do they frequently stand at the top of jobs with the highest levels of job satisfaction, but the combination of dexterity, manual skill and creativity required to do them means they could be far less likely to be impacted than anyone employed in the knowledge economy. Who knows how powerful instruments such as ChatGPT could be in the next 5 to 10 years?
One thing I do find encouraging is that it seems we have learnt some lessons from the recent growth of the social media industry, where ethical questions about its impact on mental health in young people, for example, were arguably considered far too late. Ethical concerns about the growth of AI are far more front and center of the conversation than seems to have been the case for other recent technological developments and that is to be welcomed.
AI in Assessment
Perhaps the fact that even many AI companies are requesting world governments to introduce regulation and controls suggests just how revolutionary this technology might be. In fact, from an ethical perspective the implications for AI are so profound in how it might impact human life in the next 100 years, that it is extremely difficult to consider just one argument about whether the growth of AI is a ‘good’ thing. Rather, it is perhaps more useful to consider the many question of ethics and AI specifically in regard to its potential impact on different aspects of human life, especially when considering how AI may influence different industries.
For example, in the world of Talent Acquisition and assessment, should AI be used to screen applications and reject candidates, and what mechanisms do we introduce as to how to make positive or negative decisions on applications? A critical consideration is what inputs are fed to enable the AI system to make decisions, and therefore the questions of unconscious bias arise and how to mitigate against this. But also, what about whether AI is able to screen a candidate’s social media profiles as part of the process to determine whether they are rejected for an application or progressed? Is that ethical?
At first glance, most people reading this might decide that this is an intrusion of an individual’s privacy, and there is an argument for that. But perhaps not, for each situation leads us to more questions; does an individual posting on social media for all to see forego a right to privacy? What do we mean by privacy in the 21st century? What if the social media profile showed an individual doing something illegal, should that be used as a determining factor? Should AI be considering social media feeds as an input into decision making on suitability in the first place? Should human beings be doing this? Because undoubtedly some do.
As with anything related to ethics, there are no simple answers, and often any answers that do arise lead to even more questions. Perhaps a judgement could be made on whether it would be ethical for a human completing the same task to use that approach. Inevitably the question of the ethics of AI is inextricably linked with the ethics of how human beings behave, and, as we all know, that is complicated enough!
Anyone who has seen Black Mirror will know that the show’s overarching theme is that technology isn’t inherently bad, it’s more what humans choose to do with it, and that will always mean that there are positive and negative implications arising from any new technology. From this perspective it is positive that the heads of many AI firms are already flagging that some regulation is needed, and hopefully, from the perspective of the workplace, careful consideration will be given to the efficiencies and profitability that can be gained from using AI, versus the impact that those efficiencies may have on the global workforce.
In summary, there are no easy answers to this whatsoever, but two things that stand to mind – a quote from my previous article that always sits with me-
“Change is faster now than it ever has been, but the slowest that it ever will be.”
And…
‘It’s still day one.’
About the Author
Dan Hunt
Dan is Managing Client Partner at Saville Assessment. He has a wealth of experience working as a Talent Leader at some of the world’s biggest brands, including BP, Amazon, BSkyB and Johnson & Johnson.
You can connect with Dan on LinkedIn here.
If you’d like to discuss the use of AI in assessment further, or have a project you require help with, get in touch.
The views and opinions expressed in this blog post do not necessarily relfect those of Saville Assessment and its employees.