An update on my AI journey
In my day job, I’m currently working on a new commentary which revisits one that I helped write 16 years ago. In this commentary we proposed 5 pharmacist traits that might contribute to pharmacists’ uptake (or not) of advanced practice opportunities. Since it was written, this has been one of the most read and downloaded papers in the journal’s history and kicked off the work that I went on to do in my PhD. To say I have a soft spot for this paper is an understatement.
Today I’m not going to talk about the content of the new commentary (I’ll link that later in case anyone is interested), rather I want to spend some time talking about the process of writing it. To start, it has been the most fun I’ve had working on a project in a long time. There is something special about getting to work closely again with people you worked closely with in the past. I’ve been able to see how I’ve progressed as a scientist and appreciate how I’m made better by the people with which I surround myself.
There is also something so fun about having dedicated time to just spitball ideas with smart people and witnessing to how they listen and then build off each other to advance ideas. I know it is impossible that any one of us would have come up with the same set of ideas if we hadn’t been talking WITH each other. There are very few things that make me happier than getting to hear smart people talk about their work.
At the same time, we struggled to find the “best” way to connect our previous commentary with the current realities of pharmacy practice. So how did we close this time gap? GenAI. Here’s what I’ve learned in the last month of using GenAI to get this new commentary jump started:
1) Ask google specific and detailed questions. In the past my searches might have contained a couple of key words or phrases which then required that I dig into all the links to find the best ones. Asking questions using complete sentences helped ensure that the results were in better alignment with what we were trying to do.
2) Look at the AI summary and the associated references, before scrolling down into the other links. I “think” that this is a relatively new feature in Google Chrome, but it is so very helpful and often led me to relevant references that I might have otherwise missed.
3) Regular use of these tools makes them easier to navigate.
These practices absolutely helped me to streamline the literature search process. At the same time, I noticed that this streamlining held my brain up in another way. It made my writing slower. Over the last few days, I’ve been thinking more about why this was the case. While I am getting older, I don’t think that was all of it.
Per my request GenAI found EVERYTHING related to what I asked. So, I had to go through it all weighing the pros and cons of EVERY SINGLE ARTICLE. This took time. In the past I would do my best to find 10-15 articles and start writing the paper. This time I had more than 30 relevant papers in my hands that needed consideration.
Despite being helpful for starting with the most relevant references, the AI summary was really distracting. It was so well written and compelling. My lazy brain kept telling how easy it would be to slightly edit it and be done with the whole exercise. But of course, the AI wasn’t in my voice or that of my co-authors. I also couldn’t be sure that it was “right” for what we were trying to argue with this commentary. Nonetheless those well written sentences kept cycling in my mind, making it hard to get to my own thoughts.
So, in the end GenAI sped up one part of my writing process but slowed down another. Am I further ahead now? It’s hard to say. But I have now evolved my writing process one more time and must adapt to new set of circumstances 🤷♀️.
How are you doing with you GenAI learning journey?
(Words: 705)