We can't type in front of our clients
Musicians have to perform live to prove that they are not AI.
I came up with this reflection after listening to one of Rick Beato’s YouTube videos about AI-generated music.
What is the equivalent for freelancers in medical communications?
I initially thought that conferences were the answer. They are the usual way we meet and learn from other professionals in this type of work. But I think we have to go deeper and be able to show how we work. We can’t type in front of our clients, so what do we do?
I think when a client can’t see how the work gets done, it's easy to assume another black box can do it, so we see teams that offload tasks to AI, basically substituting a black box for another black box.
To solve this, I have been working on an AI Use Policy that I share with clients and a way to keep track of each step required to perform my work. But even more importantly, I'm helping my clients understand where judgment goes by explaining the workflows that used to happen behind the scenes. For example, some of the CME providers I work with have brought part of the localization pipeline in-house. LSPs have been cut out, so I'm offering guidance on which steps should not be automated and where clinician involvement is key.
So, when a client comes our way, I think we can be very transparent and explain the parts of the work that needed our judgment. “I did X here because of Y.” Do this especially when something seems obvious to you, because it might not be obvious to them.
Recently, a client was sending me documents to review without approved terminology or a style guide. I flagged that they needed both, because the decisions were staying with me instead of becoming institutional knowledge. They weren't sure how to use them, so I'm slowly helping them adopt both at scale.
Which parts of your work that stay behind the curtain do you wish your clients understood better?