Understanding the limitations of mathematics to predict the future: Ronald Meester at TEDxLeiden


Before you watch the TEDx talk, here is one important question. What is the similarity of:

1. A dike designed to withstand a very huge, catastrophic flood with a probability rate 1 : 1000, or which means, a catastrophe can only happen once in a thousand years?

2. A news article which reports that many children born in 21st century will live up to 100 years old?

3. A computational model which can ‘accurately’ predict that a political party’s program, on certain measurements, can either increase or decrease 50,000 votes for every step it takes?

The answer: all these mathematical models are thoroughly false. Full stop.

Reality is not as simple as mathematical models can always predict about every decision we make. One of the most fundamental flaws in it, despite its overpowering usefulness in modern technologies (like Google), is that it can’t make uncertainties certain. Engineers can proudly say that the dike they design can withstand a flood for a millennium, but who knows if God, universe, or aliens you name it, decide to play dices with our fate? That’s where, even our most sophisticated knowledge of mathematics, becomes greatly fallible.

Ronald Meester, a statistics expert, gives this TEDx talk with one very simple, yet abyssal, message: we can’t predict the future. Accept the uncertainties. Full stop. Listen to his talk and think deeper.

An algorithmic method to win Oscars (or Razzies, I fear)

random oscar


Ambitious enough to win an Oscar? This Time Entertainment’s algorithmic system can let you build any possibilities.

Taking over 240 Best Picture winners and nominees from 1970s to present, altogether with over 12,000 different tags (as found on Internet Movie Database, or IMDB), this so-called ‘Random Oscar Winner Generator’ enables you to imagine billions, even trillions, of possible plot twists resulting from the combination of all those different tags, although some results may turn out to be irrational (or more possibly, Golden Raspberries-worthy), for example:

1. Genocide, sex, and writing collide for a gay lover in 17th century Brooklyn

2. Love, sex, and sadism collide for an evil emperor and a German gay general in 1940s Washington, D.C. (like something out of a gay porn, I suspect?)

3. Anti-semitism, friendship and prejudice collide in England for an American prime minister, a Jewish witness and an African-American

But some results are pretty much Oscar-worthy, for instance:

1. A piano player in 1950s Rome faces existentialism, infidelity, and unfaithfulness

2. Hope, honor and self discovery resound in Rome for a warden, a prisoner and a mistress

3. Hate, death and love confront a Chinese-American domineering father and a child in peril in the 1960s

Start the game here: The Random Oscar Winner Generator

With billions and trillions of probabilities, I wish you sincere good luck!