Is it common for editors to add to a writer's text? My naive assumption was that they often cut and re-structure and maybe suggest re-wording, but I'm surprised to see an editor actively add content that wasn't there at all
People are rightfully pointing out how ridiculous some of the overt substantive changes in that end poem are, but I was pretty amazed by how even the quite minor changes totally alter the feel of the thing too.
"Sit on a couch and look at a wall"
to
"Laying on the couch and looking at the wall"
loses a lot somehow in a hard-to-pin-down way. It's almost an impressively efficient butchering
In a way that proves the point as well though - not only are there inconsistent rules there are also seemingly under-determined rules where you can use one of many options (but only sometimes!)
Especially if enough Amazon internal tools rely on it - would be funny if there were a repeat of the FB debacle where Amazon employees somehow couldn't communicate/get back into their offices because of the problem they were trying to fix
Is there a correlation between wealth and loneliness? I could definitely imagine it being effect for the poorest in society but I wouldn't be surprised if, once you've reached a threshold of being able to keep your head above water, money and loneliness had very little relationship if at all
Unless I'm mistaken (very possible), the existing cubic Newton method they discuss already has this convergence guarantee, but introduces a lot of extra expensive work at each step. The specific contribution of the paper is finding clever regularization tricks to keep this guarantee while sidestepping the extra hassle the cubic method needs
>Scientific progress is amazing today, far faster today than at any point in history
What're you basing this on/how're you defining the growth rate here? Not rhetorical, would be interested to see your data since it seems quite a common argument to hold that it's slowing in lots of areas
How much of that is revenue from actual crypto-related business vs. just appreciation of crypto the company owns? Feels like the impressiveness of this trend really depends on that
The analogy to companies doesn't work because there isn't a legal mechanism by which you can make your short predictions come true. It's illegal to manipulate markets and most ways by which you could destroy a company (without spending more than you hope to gain) are also probably illegal. If you can think of any legal ones I'd be curious to hear
If it were legal to take a short position in a company and then take actions which blew the company up AND there existed cost-effective ways to do so, then you would definitely have seen more legitimate companies taken down by short-attacks. In contrast, here you have an entity where (a) there isn't the same legal safeguards and (b) there exists a claimed cost-effective way to tank the entity after taking a short position
If you disagree with (a) or (b) empirically then cool but it's clearly a totally different scenario to regular companies
Probably a heretical thing to say here but I feel like PG has increasingly fallen into the trap where because he's been smart and insightful in some areas he's gradually convinced himself he's smart and insightful everywhere. IMO the more he veers away from writing on e.g. coding and startups the more he comes across like a very over-confident man who thinks he's figured it all out
It's quite noticeable on Twitter too where he seems to increasingly have very confident diagnoses on everything from geopolitics to culture war nonsense to genetic engineering.
Funnily enough a friend said something lately about SlateStarCodex (or ACX now) - I wonder if getting that much positive feedback on your musings just inevitably starts convincing you everything you think must be just as insightful
I think this is right - my experience as JS and similar places is that really understanding your data-generating process and the nitty-gritty assumptions of your data/simple model makes a huge difference
E.g knowing what's causing missing values in your data and what implications various fixes on that might have on bias in your linear regressor is probably way way more valuable than fitting some shiny non-linear toy
You're right overall that most of these places are a lot less flashy on the inside (have worked at a couple). Linear models everywhere as you said
However I think you're underestimating RenTech here. They're genuinely just in another league compared to what most people consider the "elite" quant shops. You're not getting in unless you're an actually impressive academic with a track-record, so I wouldn't be surprised if they're trying some weird stuff that other shops can't even understand (although I would imagine in small size vs. more vanilla stuff).
IMO places like JS/CitSec do a really good job of bombarding campuses to inculcate this idea they're the absolute apex of mathematical wizardry, but the places that are really doing some dark magic shit aren't trying to get undergrads to apply to them. Ofc maybe I'm falling for the same kind of propaganda for RenTech