Real-world use case: AI in the world of job applicants
In 2018, about 67% of hiring managers and recruiters used AI to pre-screen job applicants. By 2020, that percentage has increased to 88% .Poorly constructed and monitored AI models introduce biases in the decision process, lack accountability and transparency, and do not necessarily even guarantee to be effective in hiring the right applicants. These AI hiring models fail by:
- Accepting (hiring) a candidate whose resumes and job experience closely matches the AI model metrics, behavioral characteristics, and performance assumptions, but for other reasons not captured by the model does not work out and most be fired (False Positive)
- Rejecting highly-qualified candidates whose resumes and job experience don’t closely match the metrics, behavioral characteristics, and performance assumptions underpinning the AI models (False Negative).
The good news is that these...