Back to Event List

How predictable is the future, anyway?

This is a PAST event. See "Meeting Notes" section for audio, video, documents and other information.

Original event date/time: Friday September 22nd, 2006, 7:00 pm to 9:00 pm

Topic "How predictable is the future, anyway?" Are there inherent limits to how predictable the world can be? Do the laws of physics and probability impose limits on the predictability of the universe? Can technology really be predicted? What about business, the markets, politics, and our personal lives? Do we think we understand the world better than we actually do? Do we foolishly think we can predict the future better than we actually can?

I (Wayne) will be presenting ideas on the limitations of human thinking from the book Fooled By Randomness by Nassim Nicolas Taleb. You are welcome to bring any ideas you have or read about about randomness, determinism, and predictability.

Abstract:

Are there inherent limits to how predictable the world can be? Do the laws of physics and probability impose limits on the predictability of the universe? Can technology really be predicted? What about business, the markets, politics, and our personal lives? Do we think we understand the world better than we actually do? Do we foolishly think we can predict the future better than we actually can?

This is a past event.

Meeting Notes:

Topics discussed were:

• The illusion of knowledge, aka "Pigeons In A Box" - how we form wrong cause-and-effect relationships/beliefs

• Social status from randomness, i.e. lack of differentiation between success due to smarts and luck.

• Lucky fools -- people who don't know their own success is due to luck.

• Anchoring, i.e. recalibrating to a new level of success, e.g. $1 million is great if you used to make $500,000, but sucks if you used to make $10 million.

• Leadership, and how it's attributable to looks and emotional manipulation ("charisma") rather than intelligence

• The Black Swan -- the rare and unexpected event -- unexpected because of the flawed cognitive inference that because it never happened in the past, it can't happen.

• Hindsight bias -- we make up false narratives because we discount the unobserved histories that could have happened -- no "summing under histories" (Feynman's term) -- history can't be repeated like other scientific experiments - why backtesting fails

• Survivorship bias - if you interview the successful people, you won't realize other people tried the same things and failed - backtesting will fail if the data has only survivors.

• The Law of Small Numbers - making wrong conclusions based on one or two observations

• Coincidence - more common than it oughta be - the Birthday Paradox

• News distortion - sensationalism, sound-bites, and "unfiltered noise" causes a distorted perception of risks (you're more afraid of mad cow disease than a car accident)

• The value of "distilled knowledge" vs "unfiltered noise". "Bloomberg machines" -- the news itself reports the wrong cause-and-effect relationships

• Ways randomness can be used for enjoyment - literature, poetry, creativity

• Ways randomness can cause harm - financial risk, medical misdiagnosis

• What randomness means for the future...

These come primarily from Fooled By Randomness by Nassim Nicolas Taleb.

Here's the "black swan" quote from David Hume:

"No amount of observations of white swans can allow the inference that all swans are white, but the observation of a single black swan is sufficient to refute that conclusion."

Note that the issue is the flawed inference, "all swans are white", that makes the black swan a "black swan" -- not the mere rareness of the event. So if you expect asteroid collisions -- even rarely, once every couple hundred million years -- they can't be called "black swans".

And yes, there is an actual historical story behind this term. For thousands of years, the Europeans believed all swans were white. Until they traveled to Australia and saw the black swan.

Wikipedia has some nice pictures of the black swan.

The discussion was cut short with many questions unanswered, in particular the question of whether randomness "dampens out" over sufficiently large time spans.

Additional links:

• The Six Serendipities of Microsoft
• Hedge funds amaranth blames 'unusual behaviour'.