Things in recruitment we hoped AI would help with...but didn't

A Reality Check on the Hype

Artificial intelligence (AI) and machine learning are often sold as revolutionary forces in recruitment, especially in IT, promising matching, automation, efficiency and accuracy. AI isn't there yet, but as they say; it's 'on the way'.

Predictive Analytics

Claim: AI-driven predictive analytics can forecast a candidate’s job performance and fit by analyzing data, including past employment and social media profiles & activities.

Reality Check: Predictive analytics, while a very great tool with insights, must be 'handled with care' in this domain. The biggest issue lies in the quality and integrity of the data employed. Incomplete or biased datasets can skew predictions and the fundamentally unpredictable nature of human behavior adds another layer of complexity. People are not just 'cogs' in a machine; they are complex beings whose actions are influenced by an array of unforeseen personal and environmental factors.

Generalization cannot be overlooked. Data Models that rely on past data might break down when faced with new scenarios or diverse candidate pools. The reliance on algorithms, scraping personal data such as social media activity, raises concerns about privacy and ethics. Humans aren't simple variables that can be plugged into an algorithm; they're complex individuals influenced by a myriad of unpredictable factors.

In recruitment, we depend on understanding the behaviors of human potential and fit, so an overreliance on data-driven predictions without a tempered approach can make wrong decisions.

Candidate Matching

Claim: Move Beyond Boolean

Reality Check: AI can aid in matching candidates more efficiently than traditional methods, the technology is still limited by the data it is trained on as well as the platform it is grafted on to. Biases in training data can lead to biased outcomes, and the subtleties of candidate suitability, such as cultural fit and team dynamics, are difficult to quantify and often shit on algorithmic assessments.

Automated Tasks

Claim: Automating tedious and routine tasks such as interview scheduling and sending updates to all parties

Reality Check: Automation of administrative tasks works until it doesn't. Recruiters often find themselves managing the automated systems or addressing the errors they produce, which consumes even more time than a simple phone call. Ghost scheduling and bombing interviews as well as spam robots are also a big issue. Confirming interviews over the phone is still the 'golden tradition' in recruitment. AI has been helpful but produces another layer of issues.

The Human, AI powered Recruitment

Claim: Augment human capabilities, not replace them

Reality Check: The claim that AI augments human capabilities is valid to an extent; however, the integration of AI can sometimes lead to over-reliance on technology. This can influence human intuition and decision-making in recruitment, which is critical for assessing intangible traits and potential that AI can't tell the difference.

Workforce impact?

Claim: Automate Everything.

Reality Check: While AI does automate certain tasks, the transformation in blue-collar jobs and the broader job market is gradual and uneven. The integration of AI varies widely across different industries and locales, with many sectors still reliant on human labor due to economic, technical, or social barriers with automation.

Almost there

Technology hasn't quite caught up with our ideal future yet. Companies need to recognize that while AI can be a very powerful tool, it comes with its own set of challenges and limitations. There are also serious ethical concerns, including issues of privacy and bias, that need careful handling. AI should be seen as one part of a broader strategy, where human judgment plays an irreplaceable role. It's not the catch-all solution it's sometimes made out to be, but with the right approach, it can significantly enhance recruitment processes in all industries other than tech.