Climate change is making hurricanes more destructive
This article provides a good, brief rebuttal to the claim “actually, climate change isn’t increasing the number of hurricanes.”. While it’s true that the current data since the early 1900s show no discernible trend^{1}, the hurricanes we experience are likely becoming more destructive due to sea level rise, increased rainfall, and greater intensity.
The article also notes that “we’re not sure about monetary damages increasing,” but it highlights that insurance premiums in flood-prone areas like Houston and Miami have skyrocketed. I find this evidence more convincing than citing a random contrarian study claiming “it’s not really that bad economically.”
It’s easy to misspeak about exactly which weather events will worsen as a result of the warming climate, so I find articles like these somewhat helpful. Still, being wrong about 1 out of 35 climate-related predictions doesn’t change the correct course of action: drastically reducing emissions.
How to price an election- a martingale approach , and subsequent discussions:
The video is a true Taleb classic (from the top commenter):
Starts in the middle of a sentence and filmed in a room containing only hard services using a webcam bought from a farmer in Mongolia in 2005.
The video is a discussion with a coauthor of Taleb on his feud with Nate Silver over (what he called) Silver’s incoherent election forecast outputs during the Clinton/Trump 2016 race. He talks at a relatively high level about random walks, martingale processes, and how one can map them to a probability for a binary event like an election. His high level point is that if you are producing forecast probabilities which wildly swing between low and high probabilities leading up to the binary event, it means you are not accurately including future volatility in your forecast.
As usual, Taleb is convincing when explaining his fundamental problems, and he knows how to entertain by throwing in a healthy amount of insults. But in an Andrew Gelman blog post, there is a long discussion in the comments on whether this was actually at odds with Nate Silver’s modeling techniques.
The linked blog post about the disagreement claimed
Silver was treating the numbers as “if nothing changes between now and the election date, here is my probability that Biden wins”.
Taleb believes that every probability number should act as a valid prediction that could be used as betting odds at the time of the prediction. This would make the crux of the disagreement between Silver and Taleb not be a quantitative error, but a philosophical disagreement.
…But even this summary of the disagreement may be wrong (it’s wrong according to Taleb).
Other people have pointed out that since Silver’s election probabilities rose above 90% before dipping back to 50%, someone could have made free money by simply buying Trump at 9:1 odds, then selling Trump back at 1:1 odds:
But this point only makes sense in hindsight: you would have to know, or strongly suspect, that Silver’s model was going to underprice volatility and swing wildly back and forth.
In the end, the only thing I could discern from all the arguing is that Taleb doesn’t trust Silver’s model (and thinks Silver and Phil “the rat” Tetlock are idiots).
This whole back and forth made me think that some online medium made for complex discussions would be better than the comment section of a Wordpress blog. Maybe the discussion quality would improve if the participants didn’t need to jump and search for the comments they’re responding to. Of course, that is assuming that Taleb would care enough to use the features to participate, which is pretty unlikely given his historical Twitter behavior.
She couldn’t think of anything to write about Bozeman because she couldn’t recall anything she had heard worth repeating. She was strangely unaware that she could look and see freshly for herself, as she wrote, without primary regard for what had been said before.
I’m reminded of the first time that I felt that I was really “doing math”: I was trying to remember some trigonometric identity, and I decided to stumble around with equations I remembered rather than just looking it up. I ended up re-deriving the formulas for sine and cosine using the Taylor series for $e^{i x}$.
The irony is that this didn’t happen until I was well into completing a math major in college, and the “discovery” is a simple math fact that students learn in pre-calculus in high school. Still, finding this connection between trigonometry, calculus, and complex numbers on my own was the first time that math felt like play, and not like a set of logical propositions to memorize or homework problems to recognize. It may have been the first moment where I was intrinsically motivated to do math, rather than extrinsically motivated to show how good I was at it relative to peers.
…but this could easily change with longer time horizons of global observations. ↩