EXECUTIVE INSIGHT can wrap their heads around . Let ’ s not talk about running your entire enterprise on AI , it ’ ll be ages before you trust that , but if we bring it down to say : Can I trust AI that generates answers only from these documents ? That ’ s much easier to answer . The answer is yes because you generated that data .
When thinking about AI , think about use cases in three categories . Firstly , augmentation or acceleration use cases – AI can help you do it a little bit faster , which is easier to trust .
Once you develop trust around those use cases , then you go to a second set of use cases which are to replace use cases or avoid use cases . Meaning that AI can do it for you . I ’ m not talking about replacing people , I ’ m replacing jobs to be done . The last one is a creation use case where you ’ re doing something so new that you ’ ve never done that before . That ’ s going to take time to develop trust .
That is the arc which I recommend to everybody . Accelerate something that you ’ re doing already . Build that trust , build that knowledge , build that learning . Then replace something that you ’ re doing today because AI can do it for you , and then do something entirely new with AI .
To me , that is the most important way to think about it because you ’ re thinking about AI from the context of a human . We must align the use and adoption of AI to how trust is built amongst humans .
It is not about technology when it comes to AI , it is about how humans will adopt it . That ’ s the real challenge .
aimagazine . com 21