Pause and Papers

Oops, the Europe trip took much more from me than I expected. I’m back in Tbilisi now, and I’ll try to get back to the daily writing habit. I have a lot of things to write about, but I’ll start with finishing the post I wrote on the plane back from Berlin.

Approximately 1 in 7 Scientific Papers Are Fake

An interesting meta-analysis provides a new estimate on the share of scientific papers with fabrication and/or falsification to some degree.

The estimate the author (James Heathers) gives us is roughly ⅐, which, to be honest, seems believable to me. There is a lot of pressure to churn out papers, get results, get funding, rinse and repeat in academia, and I always found the current 2% estimate too low.

I think it’s important for us as a society to start discussions on this issue and prioritize battling it in some way or another. This is a problem that will only get worse with time (think of the implications of “GPT-4”-class models).

Dear CTO: It’s Not 2015 Anymore

An observation that applies to any position, but now the “IT guys” are in the same boat as everyone else.

In short: 10 years ago developers were rare, hunted, and had an infinite amount of trust and leeway, but things have changed and now they need to make the case for themselves clear and obvious.

Even though I am initially a developer myself, I feel that this new position is more just and sane.