The most important AI skill leaders can borrow from pilots - opinion
AI succeeds when leaders act like pilots, by reading conditions and diverting before trouble hits.
A few weeks ago, a pilot faced a critical decision while navigating bad weather. They had three options: land at a preferred airport, fly to an alternate airport through adverse conditions, or land at a third airport temporarily before continuing. Ultimately, the pilot chose to land at the third airport, wait out the weather, and resume the flight safely once conditions improved.
This experience taught the pilot that good risk management doesn't involve asking if a risk can be mitigated, but rather if the risk is worth taking at all. In aviation, pilots learn a framework for making such decisions in real-time, which can be applied to managing the deployment of AI systems. The core principle is that risk needs a corresponding benefit, and managers should question whether the specific advantage gained by deploying the AI system justifies the associated risk.
If the answer is no, then the risk should be avoided. When assigning risk decisions within an organization, it's crucial to identify who has the authority to say no and ensure they have complete visibility into the deployment. If this responsibility is not clearly defined, the decision-making process is likely to break down. Just as pilots prefer clear weather conditions for flying an unfamiliar plane for the first time, organizations should assess whether the environment is suitable before deploying AI systems.
A generative AI pilot in an internal, low-stakes workflow with human oversight is akin to a clear-weather flight, whereas integrating the same model into an autonomous decision loop without human intervention, touching critical assets, represents flying in challenging conditions. Risk management should be integrated into planning from the outset, rather than being retrofitted as an afterthought.
Early detection of risks leads to more cost-effective solutions, while attempting to retrofit oversight and controls onto an already deployed AI system can be expensive and risky. Good risk management is about making disciplined, well-informed decisions that avoid unnecessary risks, much like the pilot who successfully navigated bad weather and landed safely.
Written by urgent.news from Jerusalem Post's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
