AI ‘Imperfection’ Keeps Executives Tuned In

A little more than a year ago, when I got acquainted with AI, in particular started training Bespoke AI Assistants/Concierges for the luxury hospitality as a Co-Partner of the BCS, I came across a commonly spread notion of an AI ‘hallucinating’.

‘Hallucinating’ in the context of LLMs and AI Assistants means some errors in judgement presented as facts by the AI or mixing up some facts to produce a certain picture or opinion which can be misleading. This greatly amused me.

And here is why.

Some users were complaining about AI making mistakes yet they did not realise that making mistakes is so very much human and as such makes AI as close to humans in communication as it can be.

AI ‘hallucinates’ occasionally and mostly because the source or training material is confusing, contradictory, and missing vital details. Humans, on the other hand, even if presented with right material or information will still misinterpret it based on their subjective judgement and emotional filter. The AI does not have these filters, but the result might be seemingly the same.

Now, let’s look at Bespoke AI Executive Assistants for Executives and Leaders and the possibility of their ‘hallucinating’ while ‘in service’.

Personally, I believe this is no problem. Firstly, AI hallucinates occasionally. Secondly, the possibility of the AI Executive Assistant making a mistake or an erroneous judgement however rare will keep their bosses — the Executives they serve — on their toes. In a good sense. For, any Executive or Leader should never lose their independent way of thinking, their own sharpness in judgement and also pay attention to what they are told, advised, or ‘served’. The imperfection of the AI that ‘commoners’ complain about trains the Executives to stay alert, switched on and tuned in.

As in real life the Executive should not believe everything their subordinates report or tell them, in communication with their digital confidants they should check the facts and statements too. There is one big distinction though — AI does not intentionally lie, but people do.

Seraphina Michelina Aurelia

Founder and CEO, SMA Crown

Co-Partner, BCS