For a long time education as been organised in a similar way to fractional distillation. Kids would go through increasing selection similar to a fractional distillation designed in a way that only a small proportion a what enters the system can make it to the top:
However nowadays it seems that education is moving to a fat tail (aka power law): a small proportion of people are on top of leading edge topics and unless one is
coached from the inside it is impossible to get in as the speed of development is too fast. One such topic is
Artificial Intelligence. So many papers are published so
fast that no one can stay on top (unless already inside), as the growth in publication and pre-print is exponential 1 Fig 1.
On the other hand, many topics from changing a light bulb to very simple AI/CNN topics are explaine on YouTube, GitHub or StarckOverflow meaning anyone can solve a simple problem with only a few hours or searching the Internet (aka "Hello World" or "3 fingers coding").
Walsh, Ian & Fishman, Dmytro & Garcia-Gasulla, Dario & Titma, Tiina & group, The & Harrow, Jen & Psomopoulos, Fotis & Tosatto, Silvio. (2020). Recommendations for machine learning validation in biology. Modern biology frequently relies on machine learning to provide predictions and improve decision processes. There have been recent calls for more scrutiny on machine learning performance and possible limitations. Here we present a set of community-wide recommendations aiming to help establish standards of machine learning validation in biology. Adopting a structured methods description for machine learning based on DOME (data, optimization, model, evaluation) will allow both reviewers and readers to better understand and assess the performance and limitations of a method or outcome. The recommendations are complemented by a machine learning summary table which can be easily included in the supplementary material of published papers.