Slot machines, retirement homes, and Brexit | Hacker News Reader

Slot machines, retirement homes, and Brexit

knowledge.sparkcapital.com

24 pointseoflazer5 comments

Finding an optimal solution to explore-vs-exploit is the "multi-armed bandit" problem.

https://en.wikipedia.org/wiki/Multi-armed_bandit

> Originally considered by Allied scientists in World War II, it proved so intractable that, according to Peter Whittle, the problem was proposed to be dropped over Germany so that German scientists could also waste their time on it

...though it's actually gotten a lot of great mathematical solutions over the decades :)

I think you could explain the age distribution easily without invoking "explore and exploit." The two generations aren't playing the same game.

Many young people see the value in moving freely across boarders for business and pleasure, they still have the whole of their working lives ahead of them. Many old people feared the changes and regulations looming on the horizon from the EU.

Totally different incentives.

You could also characterise it as a risk, whereby potential downsides are amplified by age.

Young voters will have to live with the consequences for most of their lives, while old voters don't have much time left.

Then again, you could argue older voters were altruistic and voting in the interests of the next generation.

TL;DR: You trade off exploiting what you know against exploring for better maxima in data. Explore is longer term, Exploit is short term. Older people are likely to exploit beause they'll die soon, just as you stick with what you know / already think in restaurants when you're due to leave a city.

Conclusion he infers is that "it could be" that young people are naive and want more data to understand the world (explore), and older people have more data points and wiser choices (though not necessarily correct), and... so, er... experimentation can be a bad investment. It's a little unclear if he thinks Brexit was a good idea, only because older people had apparently more data points If he does, this does beg a lot of questions as to whether these were rationally-considered data, or even consciously processed.

I'm unclear what conclusion or information he wishes to impart abotu Brexit: that part seems very guess-like and vague.

He links to a Woodford Investment analysis that states that Brexit might not be that bad for the UK (the report is from Feb 2016). This goes against every other analyst report that I've read, and the optimism seems to be based on a perception that doing trade deals with other countries is easy. My guess is that the author is a Brexit supporter looking to justify their position. "Oh, I voted with all the old people? They are all actually really wise".

Reply on news.ycombinator.com