Breadcrumbs feature in the fairy tale Hansel and Gretel where two lone children drop them along their journey into the forest so that they can find their way home.
I’ve been thinking a lot about breadcrumbs and paper trails because the research world is rapidly moving to widely using artificial intelligence, including for generating questions, collecting data, and writing papers. Whilst these uses may sometimes be legitimate and make for faster and better research, it’s now also possible for bad actors to use AI to make a decent quality paper in just 30 minutes.
Hard-working human researchers need to differentiate themselves from the instant results of robots. I think breadcrumbs via a paper trail are the best way to do this. Ideally, researchers would leave online crumbs at all stages of the research process. This could include sharing their ethical clearances, pre-registering their plans (e.g., with As Predicted), and sharing data and code (e.g., on GitHub). These are all good research practices, so there’s a potential win–win.
The ideal researcher breadcrumbs would be cheap for those doing the real work and very expensive for those who aren’t. It surely would be hard — although not impossible — to create time stamps in the past.
There are other ways around it, for example mass-registering hundreds of questions and then picking those that suit your instant paper; or just picking any pre-registration and hoping that the journal and reviewers don’t check. This would work sometimes, but it creates another paper trail that would hopefully end in mass retractions.
I was interested to read of an entire novel recently being pulled because the publisher suspects it in to be mostly AI written. Maybe authors will soon need to occasionally video themselves working at their desk? Or hang out at cafes where people can confirm they saw them writing.
My colleague Mike Doran has been talking about the benefits of videos in research for years. Will we soon need videos of researchers running the experiment to prove that they were in the lab? Again, this could have a silver lining, as recording your actions throughout the experiment could help fill in any gaps in the paper, especially for those looking to reproduce or build on your work.
The breadcrumbs worked out badly for Hansel and Gretel, because the forest animals eschewed their nuts and berries and gobbled up their breadcrumbs. Then the children were almost gobbled up by a wicked witch. There are real-life villains in the research world and if researchers want to be the hero of their own story, then they need to share more of their journeys through the woods.