Reading headlines one will marvel at the abilities of generative AI, but looking at the comparatively limited usage of generative AI in the industry, which is pretty basic, one is left wondering. The key question is if AI is so capable why hasn’t it already taken over everywhere?
To understand this, imagine you are the coach of a soccer team. Think of AI as a professional soccer player. His highlights reel is jaw dropping, he can score from any distance at any part of the field. What’s more, he has no transfer fee, can start tomorrow and his salary demands are the same as the waterboy. Naturally you give him trial before putting him on the team. Then you find out that sometimes he just misses the ball completely. Other times but rarely he scores own goals instead. Now you are not so sure to put him on the team anyway, right? Nobody wants to be the coach with the player that scores an own goal every other game.
That is similar to how GenerativeAI is treated today. There is much excitement and impressive showcases but in the end very little makes it to production. Most is “Stuck in Beta.”
But that’s too bad. There are ways that AI could be part of the team. You just need a new way of managing AI. Let’s look at a few tactics
- Keep AI away from the goal. You’d want to reduce the risk that AI scores own goals. That can be done by putting him in attack or in the midfield. On defensive set pieces you direct AI outside the field. The equivalents of own goals are data leaks, inappropriate answers leading. The way you handle this is by putting guardrails on AI or using dedicated instances that are not shared with others.
- Make sure AI always has someone close by. This curious tactic is the equivalent of man marking your own player. this will make sure that when AI drops the ball there is someone to pick it up and put it back into play. The equivalent of this is having a human in the loop. For all AI tasks someone is there to review and solve harder problems.
- Train the team to play together with AI. It is clear that AI does not act like the other players on the team so that is why you need special training regimes on how to play with AI to make the most of its abilities. By understanding the dynamics of its game it can be used to a fuller extent. The analog to this is AI training for employees. Courses and training in how to prompt and use in work related tasks helps leverage generative AI to a higher extent.
- Rework your gameplan. With AI on the team you probably want to think about how the whole team plays and approaches matches. Perhaps you want to play more offensive with high pressure rather than parking the bus in front of the goal. You might want to rethink how you market and run your team and the trainers you want to employ. In the real world this amounts to rethinking your operating model.
Thinking of generative AI as an interesting new signing to your team that needs special attention to bring exceptional value to your team might help you employ some simple tactics that can get you started. But if you don’t get started, remember that soon the league will be full of teams that did and you will be facing AI in competition.
Danish 
English