I had the same reaction to the headline, but in this particular article it actually means exponential.

(It does make a leap from computation to intelligence though.)

They do make a point of saying that they are answering only the very specific argument of "geniuses can solve NP problems with greater efficiency than muggles". Apparently they intend to argue against muggles thinking at n^2 and geniuses thinking at n^3 in future essays.

Exponential in what?

Or when they mean quadratic or super linear.

or HN's favorite, "orders of magnitude"

I have not read the article but in general:

The problem I have with exponential is that you can’t have an exponential difference between two data points. Exponential is a measure of growth which can’t be seen with just two data points. However orders of magnitude can accurately describe the difference between two data points. Even though it is somewhat ambiguous as well (powers of 2,10,16?) orders of magnitude makes more sense than exponential.

Are non-base 10 orders of magnitude actually used anywhere? I haven't encountered neither in either physics, CS or any other field (and I'm reading whatever I can get my hands on).

> Are non-base 10 orders of magnitude actually used anywhere? I haven't encountered neither in either physics, CS or any other field (and I'm reading whatever I can get my hands on).

In computer science, using a base of 2 (or a power thereof) for magnitudes is not uncommon.

I agree that it might make sense in theory, but in reality people use it extremely inflationary to make their comments sound more „science-y“.

Instead of saying „X is orders of magnitude bigger than Y“, why not simply say „X is much bigger than Y“?

If you're discussing a type of object that has such a wide range of magnitudes, describing it as orders of magnitude difference is a good descriptor. For example, the size of stars. One star may be ten times the size of another. Another may be 100 times the size. The next may be 1000 or 10,000 times the size. You wouldn't simply say that stars have a wide range of sizes, you would say that they have orders of magnitude difference. Likewise, you could say that one star may be several orders of magnitude larger than another.

9 is much bigger than 1.

900 is much bigger than 1 and (two) orders of magnitude greater.

Or "two football fields"

(without specifying which kind of football)

I don't really think 5% error margin is really worth getting upset about for a rough visualization guide.

An AFL field is about three times the size of those used in most other codes.

The only people likely to think of an AFL field are us Aussies, and only the subset of us that don’t follow rugby union.

And we’re statistically insignificant.