Thursday, 30.01.2020 – Can Statisticians Forecast Better than A guy With A Pencil?

Link of the day: Delphi Epicast

I am currently working on a master thesis that tries to predict cases of Ebola in Congo. The models used for forecasting are quite complicated. But in the end it all kind of boils down to drawing a more or less straight line from the last point and trying to continue that into the future. I always wondered: what would happen If I drew that line with a pencil instead of generating predictions with … more

Cognitive inequality

It has by now become a well-known trope that automation of low-skilled jobs hits certain parts of the population much harder than others. And that this trend is likely to continue in the future. But most people think about this inequality in terms of education, not in terms of cognitive abilities. Few people are aware of how pervasive the problem of cognitive inequality is. And even fewer think about what to do about it.

The standard IQ test with mean … more

The Pharma Industry is Broken and Medication Should Be Directly Sold To States

The Problem

The pharmaceutical industry clearly is not working as it should be. Not only did an industry that makes people healthier, happier and helps them to live longer manage to become the least popular industry in the US (worse than the US government and a whole 12 percentage points worse than the oil and gas industry). They are also not developing and providing the products that would most benefit society and the world as a whole. While the world … more

Saturday 25.01.2020 – Tracking Cell Phone Data

Link of the day: Twelve Million Phones, One Dataset, Zero Privacy

Locations of cell phone users are not only tracked by governments, secret services and telecom companies. New York Times Opinion writes about a data set of 12 million phone users generated by private phone location companies. Data sets like this one are regularly sold to private firms. The authors managed to identify individuals like Johnny Depp and Tiger Woods with high probabilities. An interesting read.

Phenoptosis – Are We Programmed to Die?

In 1999 the Russian biochemist Vladimir Skulachev coined the term Phenoptosis in his paper. In analogy to apoptosis (planned cell death), where cells undergo a death program to protect the organism, he hypothesized that sometimes individual organisms die to protect a population or larger group of organisms. Wikipedia lists numerous examples:

  1. E. coli undergoes programmed death once infected by a phage (a virus that infects bacteria).
  2. Male Antechinus die after their first mating season as their body essentially disintegrates

Monday, 20.01.2020 – How to Create Disruptive Research Teams

Link of the day: Ingredients for creating disruptive research teams

A friend of mine, Stefan Torges, has written up a quite extensive summary of the characteristics that tend to be present in very successful and disruptive research teams. In his summary he draws conclusions from research evidence as well as from various historical examples such as the Bell labs (they invented, among many others Unix and C), the Manhattan Project or Lockheed Martin. If you are building up … more

Sunday, 19.01.2020 – Why Published Science May Systematically Underestimate the Effects of Climate Change

Link of the day: Scientists Have Been Underestimating the Pace of Climate Change

Scientific publications may systematically underestimate the effects and the speed of climate change (and there appears to be evidence they have done so in the past). One basic argument is this: Public dissent and fighting is detrimental to the cause behind most climate research. In order to avoid this, scientists are incentivized to publish conforming to a consensus view, i.e. to publish only claims that almost all … more