Lord Clement-Jones, former chair of the House of Lords Select Committee on AI, was the key note speaker at our event on 8th May, Assessment in the age of the algorithm. Below, you can watch a 20-minute video of him in conversation with Jeremy Tipper, Strategic Advisor, Sova, and read his tips on incorporating AI for HR professionals. The broader discussion covered how business leaders can embrace new technology and be aligned with the workforce, setting them up for success and to be ready for change.
HR managers have a massive stake in introducing AI and need to introduce proper risk management to the process. AI is not just like putting in a new machine. It has huge implications. Broadly speaking AI is assistive or augmentative, not generally substitutionary, but HR needs to decide how far to go in replacing tasks or jobs and be aware upfront which it will take over. It’s important to go into the process with your eyes open when losing a percentage of your workforce.
I don’t believe that tech developers quite get the diversity issue yet. It’s of course important to build AI assessment processes that have diversity and social inclusion as parameters at the start, but it’s also important to look in the right places. Check which universities you are sourcing from – do they have a diverse student body?
Our qualitative research in the tech industry said that it’s not just STEM subjects that are important. Skills we’re going to need in the future are creativity and critical thinking – skills that are often developed by doing the humanities. We need people who can use, apply and develop AI and its applications, not just write an algorithm or code – machines will be coding soon!
Although cognitive diversity is an important part of ensuring a future-ready workforce, and tech and AI developments can present a huge opportunity to increase cognitive diversity in the workforce when done in the right way, other types of diversity such as gender, age and culture, are just as, if not more important. A truly diverse workforce has a direct impact on business success.
Ensure you build explainability into the specifications, algorithms and processes of the program. You need to be able to explain how the machine arrived at its conclusions to a discrimination case in court or at a tribunal. Ensure HR and lawyers are involved in designing the requirements right from the start. The conventional rules such as the Equalities Act still apply, bias in recruitment is still subject to the same laws. This isn’t going to change, and any recruitment process will need to be transparent and explainable.