Artificial Intelligence as a jury? But you'd let it pick your employees, right?

Per its proponents, artificial intelligence (AI) holds the promise of everything from eliminating car crashes and medical mistakes to fraud-proofing commodities transactions. Or it might end humanity in a fiery robot take-over. Before the machines take over, can they help us find more justice in our courts and our organizational hiring practices? 

AI promises to be able to gather near infinite data in seconds. We already know that machines are better at playing chess, diagnosing cancers, and reading poker bluffers than humans, why would AI also not beat us at discerning evidence in a trial?

Jack! Are you suggesting that we replace juries of our peers with machines?

Justice is among the highest ethical concepts. While I believe our current USA system offer the best chance at justice, we both know it has flaws. Namely, that humans cannot get past our biases.

Before you stomp out a firm, “No!”, ask a few questions. It’s very likely that your organization already uses a form of AI in hiring. Messing with peoples’ jobs is not as bad as false imprisonment, though both fall under the same ethical category of justice.

Let's play with two scenarios. The first on crime, the second on hiring.

Crime

Let’s say you’re on trial for a crime that you did not commit. Let’s say you’re convicted because influential members of the jury just do not like you. (Yes, this is documented to happen.) Let’s also say that AI has already advanced to the point we know it will in a few years, and that means we have lie detectors that work as well as DNA readers work today.

1. Who or what method would you prefer to determine your guilt or innocence on appeal?

Hiring

Let’s say you’re denied a job for which you are the best candidate because the hiring manager just does not like you—none of us is so naïve to think this never happens. Let’s say that we have machines that can tell us with near perfect accuracy that you are the best candidate.

2. Would you prefer to go through the current system or talk to the machine?

Crime and Hiring (these are often not as far apart as we hoped, are they?)

3. What promise does AI hold for other law enforcement tasks like probable cause street stops and other HR tasks like settling employee harassment claims?

4. What ethical problems do you foresee in the use of AI

We already use assessments (an incipient form of AI) to help us know the job candidate that best fits a company’s desires. (Shameless plug for PDP systems.)

5. Back to ethical seriousness—as AI advances, how far would you go with using it in hiring?

AI will one day—like next Tuesday—be able to mine a job candidate’s online history and pop out answers to how they typically behave (this can be done now, friends).

6. What if AI can tell us how a candidate has developed over the years and their development needs coming in?

7. What effect is AI likely to have on ethical issues in hiring such as determining how well a candidate’s behavior matches your org’s desired culture?