Quick Notes on Nate Silver's: On the Edge

a mix of poker/crypto/EA

I saw this post a while ago (feel like it was paul graham, but couldn’t find it) where you stick to reading books > 5-10y old and for recent content you follow the mediums of podcasts, twitter/x, and blog posts/substacks. I try to follow this and do pretty well, I think it’s works good.

However my X timeline, podcasts, and substacks were all recommending nate’s new book On the Edge. After reading book The Signal and the Noise, I was pretty excited to check this out.

It focused on about 1/3 poker, 1/3 crypto… mostly sbf, and 1/3 ea and rationalists.

You can read all of my highlights and notes via Sublime here. Below is a set of highlights with some thoughts on why they are interesting to me.

Byrne Hobart at the first Manifest conference (here are some notes on the second one) gave a presentation called Prediction Markets v. Financial Markets. He lays out the case that currently prediction markets are extremely skewed towards smart money. Nate similarly makes this point with the following

You’re essentially indirectly competing against the smartest, most informationally savvy groups in the world.

Byrne and Nate both continue to say that the prediction market sharps are extremely

dependent on recreational bettors to make their profits

and that in order for prediction markets to succeed they need to slowly shift from being very academic to more fun. I think this contrast is very visible when comparing Metaculus and Manifold. This shift can also be seen as various prediction markets are starting to shift focus on offering sports betting and books are slowly offering event trading (for more checkout Towards Event Markets).

As news organizations scrambled to correct their coverage, for instance, traders at Manifold determined that the IDF probably hadn’t been responsible for whatever had happened on that particular night at the Gaza hospital

This was probably one of my favorite notes in the entire book, even though it was almost just an aside comment. It gives a nice example of how forward looking prediction market driven news can be effective (side-note I am building a prediction market driven media company at adj.news).

I feel that nate nicely captured the state of prediction markets, how they are useful, what they can improve. Also covered in The Signal and the Noise, I think the following quote shows the widespread acceptance of the need for political polling but acceptance of single numbers for other forecastable things like GDP.

Political polls are dutifully reported with a margin of error, which gives us a clue that they contain some uncertainty. Most of the time when an economic prediction is presented, however, only a single number is mentioned. The economy will create 150,000 jobs next month. GDP will grow by 3 percent next year. Oil will rise to $120 per barrel.

In general I have a very mixed view on EA. It’s interesting to me for a few reasons, (1) Tons of people that overlap in my interests at some point move into EA, this is prevalent with poker players having a large role in the movement, of course crypto, and a large majority of forecasting community being rooted in it. (2) I find many of the forums and discussions that take place to be very differentiated from many other parts of the internet and I like that.

Overall I think the quotes below sum up how I view many people in the EA community, I also think that the sometimes get hyperfocused and dedicated to specific causes (I am still not sold on P(doom) / AI risk and there has been insane amounts of funding for it).

The EAs were too “woke” and too concerned with appearances.

and

In practice both EAs and rationalists have a catholic appetite for involving themselves in all sorts of controversies

Some other quick notes

• I was very surprised with how manual nate described the practice of bookmaking the initial lines being and how they welcome sharp betters to make the first bet so that they can adjust the lines. This is described in the section on market making and retail driven books. I had always thought that bookmaking was extremely quant adjacent and there was almost no “feel-based” lines being set

• I kinda didn’t really say much about the poker portion of the book given its something I am least familiar with. I thought it was well done and an interesting intro into the sections of other risk takers

• I focused on specific highlights in the book but the main concept of the book was that there is this group of people nate calls “The River”. Groups of people that generally fall into this group are poker players, crypto traders, vcs, hedge funds, EA, rationalists, forecasters, etc. Basically people that quantify risk and take it.

  • I think the reason I liked the book so much was simply the large overlap between all of these groups