Hacker News Reader

Comment by timr | Hacker News Reader

After listening to this, I'm more convinced than ever that Nassim Taleb is the most overrated phenomenon in business/economic literature. His books are vague and trite, and while his arguments touch on currently popular memes (i.e. "debt is bad, because it exposes you to downside risk"), his thoughts go no deeper than the level of analysis you can glean by watching five minutes of CNBC.

It was still interesting to hear this interview, because it's clear that when he's pressed by a moderately intelligent interviewer, Taleb has nothing to fall back on other than repetition of sound bites. By the end of the interview, his logic degenerates into the repeated invocation of a weak metaphor comparing US spending with the Madoff ponzi scheme.

Hopefully this guy's five minutes will be up soon.

View on HN

His style tends to be hyperbolic but he does have interesting things to say. For example how he can't predict the (financial) future with respect to what will happen or when, but he can predict to whom it will happen to.

He makes some reasonable points on debt and leverage. He didn't seem vague to me - reduce debt, prefer equity over debt, build robustness by spare parts or by functional redundancy. That all seems pretty reasonable, specific and not really a view that's advocated by the main stream financial press.

His comments may be reasonable (for some value of the word), but you and I have very different definitions of "specific". I can reduce the entire talk to a one sentence summary ("debt is risky; redundancy is less risky."), with little loss of information.

More importantly, nearly everything the man says about debt or deficit spending revolves around a cartoon charicature of statistical reasoning, and boils down to the thesis that debt is always riskier than surplus, because the downside is larger. That's fine, but it's basically just folk wisdom. Even the notion of the "black swan" is just a goofy way of saying what every third-rate insurance agent knows: given a long enough outlook, improbable events are guaranteed to occur. But the important question (the relevant question) is: do you actually care? That's a far more interesting discussion (one that plays a prominent role in everything from insurance contracts to option pricing), but one that eats away at the simplistic appeal of his thesis.

His remarks about "spare parts" and "functional redundancy" are so vague as to be meaningless...sometimes, "spare parts" means cash surplus, but other times, they mean having more than one marketable skill. Maybe he means both. When you speak in glittering generalities like Taleb does, it's easy for other people to project their thoughts on to your words. I suspect that's what makes him so popular.

You realize your comment exactly describes your criticism? Any specifics or just vague and trite?

Okay, here's a specific: if I wrote my comment the way that Taleb writes books, I'd expand my remark with a bad metaphor, re-state the metaphor ten different ways, and then use those re-statements as proof by example of a broader thesis concerning the decline and fall of English literature.

The man uses ten-dollar words where five-dollar words would suffice, he can't organize an argument to save his life, and his book chapters consist of endless assertions and re-stated assertions of the chapter titles with little further elucidation. Those are but a few of my specific complaints about his writing.

Could you point out specific holes in his arguments?

I find your comment to be too much relying on metaphors and not backed up with example or data and without any specifics?

E.g.

One of the recurrent themes in his books is how forecasters fail, and how when they fail they fail by huge margin.

In My experience I have found this ideas to be quite insightful.

Why do you think that his main Thesis about events with large impact and low expectation and their effect on our ability to make predictions is incorrect?

I agree that he himself gets involved in narrative fallacy and that his arguments could be more succinct and sans the Tony and other character that he uses in his book. But Maybe those are there for the hoi polloi and not for people who understand Machine Learning, Search and stuff.

"Why do you think that his main Thesis about events with large impact and low expectation and their effect on our ability to make predictions is incorrect?"

I didn't say that. I said that he's a bad writer, that he argues in flimsy metaphors, and that he tends to make sweeping generalizations based on little more than his own previous metaphors and generalizations. Don't put words in my mouth.

It's axiomatic that low-probability events must occur, given a long enough outlook. What's not axiomatic are the broad, overwrought conclusions that Taleb reaches from this starting place. Reasonable people who grasp the basic tenets of probability theory as well as Taleb can (and do) arrive at drastically different conclusions about US fiscal policy than does Taleb.

The Black Swan was a black swan.

Reply on news.ycombinator.com