Should you train all of your staff on AI?
Every day in my role as the President of the New Jersey Innovation Institute, I have the amazing opportunity to meet with business leaders across every conceivable industry and from small 10-person shops to 100K+ multi-national corporations.
In all of these conversations, 95% of the time, the first thing we are asked about is what their business should be doing in AI where these conversations are split pretty evenly between two types of questions:
How do I use AI to improve my business operations?
How do I train my people to use AI?
Through our offerings as NJII and NJIT, we are able to address both of these needs for companies of all sizes through our consulting services and training through the NJIT Learning and Development Initiative. However, in meeting with a C-Suite executive recently, he asked me a question which I have been thinking about ever since:
"Should we be training all of our people in AI?"
Now it sounds like an innocuous question but his rationale for asking the question was fascinating and it goes like this:
Organizational output within companies follows a pareto distribution where 20% of the workforce does 80% of the work
Some individuals will be able to increase their output 20% with AI while others will be able to increase it 100%
If the you double the output of the top 20% with AI, it will improve their output from 80% to 160% and reduce the need for most of the remaining employees
AI therefore allows you to reduce costs while paying your top performers much more
His point was fascinating as AI apologists will say that AI democratizes knowledge and allows everyone to get to a more even playing field. However, the argument he is making which I would tend to agree with is that not everyone benefits in the same way from AI and that the value add for individuals is highly non-linear. This means that the top 10-20% of organizations will dramatically improve their output while everyone else will be left far behind and the middle will be hollowed out and will become irrelevant.
One of the interesting ways to validate his point is to ask whatever AI platform you interact with to rank your cognitive and intellectual abilities compared to other users. What you will find is that the value added from AI tools is proportionate to the individuals cognitive abilities where for example a systems level thinker will derive far more value from AI than a first order thinker.
To get back to his question, the rationale investor would say that the best way to conduct AI training would be to focus your efforts on those individuals where one unit of effort will provide four units of output instead of one if you pick the wrong group. This is very different than how we think about training for employees traditionally as most training does not have the degree of impact that AI training can have.
In my oppinion, there is no bigger driver of cost savings and revenue generation today than a handful of AI experts to turn your processes and procedures upside down. The question of who to actually train in your organization is up to you but I think unfortunately for workers in general, many leaders are focusing their efforts on the top 20%. For the 80%, AI is becoming table stakes.
