Thinking Fast and Slopes
drorpoleg.comThis is an incomplete analysis; it looks like he's read Kahneman but only read about Taleb.
Similarly,he does pay homage to power law distributions, but fails to note that normal distributions are assumed ubiquitous because everyone learns them in Stats 101. They don't apply to most of the interesting world.
> In most industries, the key to success is controlling demand (getting attention) rather than controlling any specific resources or factors. Output is non-linear — a single person with a computer can outgun a giant company and produce a billion-dollar app or a billion-views video.
Is it also the case that larger organizations have not adapted to the new communication mediums? They seem far more risk averse and often handcuffed by their legal departments on what they can and cannot do.
One parallel that I could draw is the current state of streaming and the entertainment industry. Traditional artists and labels say it's hard to make money streaming and the industry if having trouble creating stars like it traditional did in the past. (https://www.nytimes.com/2021/05/07/arts/music/streaming-musi...)
The large movie industry was traditionally a supply driven model. There are only so many summer Hollywood blockbusters they can put out and the article points out that we are shifting from the mass-produced for the masses media to niche products, that means moving away from overdone one size fits all marvel superhero smash the bad guy movies to bespoke unique content (TikTok). This article (https://www.profgalloway.com/struck/) points out the problems facing the industry and according to a consumer's trend report in 2022, more than half of gen z and millennials prefer TikTok to streaming services. Services like Netflix, paramount, Disney, and the like are jacking up prices and still losing money.
This is prima facie plausible. The best way I know to get some empirical pricing on this would be DOOM option chains all the way back to 87 and the “volatility smile”.
If true, it’s an accelerant on all the worst things about the modern economy: value accrues not to sound fundamentals but those already well-enough capitalized to bet the tail in a diversified way, which is of course not an option for the working person.
When I first learned the properties of a Cauchy distribution it sent chills down my spine.
I'm super curious about this statement. What was profound about the properties of a Cauchy distribution for you? I watched this series of videos: https://www.youtube.com/watch?v=9j398FbirQY&ab_channel=Ellio... and then google practical applications of it. Is it just that you are able to use this approach to map the behavior of something that seems so complex as to be beyond distillation/mapping?
I first read about the Cauchy distribution in a book titled The Physics of Wall Street by James Owen Weatherall. Before then, I was mostly familiar with the normal distribution. The normal distribution shows variance as contained around mean, median, and mode. The Cauchy distribution can look like a normal distribution but it actually has no mean. What is unsettling is how real world processes or events might seem like a normal distribution, but actually be a Cauchy distribution. Which means a variance you believed to have reasonable bounds would suddenly go haywire.
https://www.itl.nist.gov/div898/handbook/eda/section3/eda366...
> overestimating low-probability events is not irrational — it's optimal
It can't possibly be more optimal than estimating them correctly.