Testing times - what do you do when knowing the answer is no longer the question?
Recruitment and selection in the age of radical uncertainty - over my career I have seen huge shifts in testing, from pen and paper testing of 10s of 000s of people to the introduction of CBT and now unproctored/remoted proctored testing.

Over 10 years ago at ATP India I talked about how the education system had survived largely unchanged for over a century - now, the democratisation of knowledge and exponential technological change are upending the paradigm, with huge impacts for testing and
certification.

The same forces are shaping the future of just about everything - including the world of work and, by implication, HR and the talent and recruitment marketplaces.

This has major consequences for selection - both WHAT we test and HOW we test it.

We continue to test knowledge, when the half-life of knowledge has never been shorter.

We continue to test proxies for intelligence, behaviours and skills when it is better to test their application in real world scenarios- which is why Assessment Centres have the highest levels of predictive validity.

The educational system continues to churn out square pegs for holes which will soon no longer exist, and the same will probably be true for recruitment selection, which has traditionally been an extension of this model - it is increasingly failing employers and students.

And in a world of paradoxical talent shortages, these models continue to exclude or leave behind large parts of the global population - who don’t have access to formal education or who don’t fit the production line - eg the neurodivergent.

Yet these are the people whose abilities and skills may be just what employers need - with resilience, problem solving and big picture thinking skills.

Case study of UNDP’s radical graduate programme.

Consequences of technological and societal change for recruitment:

We are seeing both critical talent shortages and job shortages.

The application and screening processes has been upended by AI - on the applicant side and the employer side.

Increasingly the future is likely to be one of AI agents talking to AI agents and AI driven internal/ external talent marketplaces (UNDP case study).

Increasing use of AI in evaluating applications and marking assessments, raising issues of bias as well as the potential of greater objectivity.

AI introduces the possibility of moving from standard MCQ testing to more complex simulations. Assessment centres have already shifted from in person to virtual, next shift to AI : eg interviews with agents, video generated scenarios and group discussions etc

Possibility to move from proxies for intelligence and EQ testing to embed testing of more complex constructs in simulated environment at scale, evaluating the application of skills and competencies in job-related contexts

Testing for skills we don’t have for jobs that don’t exist yet!

The new skill is prompt engineering - basically asking a question

So perhaps the future of testing is no longer to find the answer, but to find the people who know what question to ask?