There are a lot of emotions among doctors and academics about what AI will mean for their practice and lives. What is less often discussed is the concrete things you should have already stopped doing because of AI. (note: our post yesterday partly inspires this piece). Here is a partial list:
You should never ever write a review article. AI is an all-you-can-eat review article buffet. Why on earth would you write one, when you can generate an up to date one any given Sunday.
You should not be writing grants. First, if you are still writing federal grants, consider scratch lotto tickets. Second, if you decide to do it anyway, at least get AI to write all those pages— skim and edit as you see fit.
Game over for all descriptive research using publicly available data. I made a career— 500+ papers— often using publicly available data (like all FDA approvals in the last decade sort of thing). Sadly, all this research is dead (or should be). You can just ask AI: make a histogram of the median overall survival gain for all drugs approved in the last week, month, decade, etc. There is no need for some people to publish paper with these questions gated to their preferences.
Medical guidelines. We don’t need guidelines written by 12 academic cardiologists with above average conflicts of interest. AI makes tailored guidelines updated to the minute.
Meta-analysis. Your meta-analysis is pointless. AI has already defeated you. It can generate continually updated meta-analyses.
Personal statements/ letters of rec. First, question the judgment of anyone asking you for a personal statement or letter of rec, second, have AI draft it for you.
CV - AI can keep your CV. Check lightly.
Notes in conversation. Just have AI record and summarize. All the more reason to look at patients and not the computer. As I said before…
Reading long boring things. Put it into AI and have AI summarize.
Drafting emails. Have AI take first stab at responding to emails. When in doubt about the right response, ghosting is the best solution.
For anyone making a career in the knowledge economy, AI is coming for you. But in the meantime, there's no reason why you should be doing any of these tasks. Things that require a little bit of intelligence, and a little bit of elbow grease, should already be outsourced to AI.
There's going to be someone in the comments who says for one of the items on my list : But there's a real value in writing a review article. Pulling those references, reading through that data. Blah blah blah.
Well, there's also real value in writing your review the old fashioned way. Use note cards, and put pen to paper. No computer. While you're at it, use a fountain pen. Better yet a quill.
The printing press was also a net negative for books made by hand. Do you know how much better a handmade book is?
And, things were better before running water and indoor plumbing. Go ahead and crap into a bucket and throw it out the window.
Anyway, they'll be plenty of time to write the review article after you get done spinning cotton to make your own clothing. Join me, luddites!
Fun article, makes sense for adults. But AI is successfully making a lazy young generation lazier by destroying problem solving-independent thinking-critical analysis and process skills. Careful which horn you toot.
Luddites destroyed machinery because the industrial revolution was destroying their livelihood, not because they eschewed progress. Sounds about right.