Yeah I mean, I don't think advertisers are interested in specific people per se, certainly not enough so to risk going to prison. It's more like, "hey can we show the version of the beer ad where they're watching football to people who watch football?" No one cares *who* that is specifically
Just gonna drop some relevant publicly-available information here that many people are not aware of support.vizio.com/s/article/Sm...
lowkey goated when aligning on a roadmap and solidifying key deliverables is the vibe
Oh, nice!
Yeah I think we gave "if we just LISTEN to antivaxxers we will be able to persuade them so much more effectively" a good college try. Turns out all it did was give measles a three year head start. The idea was wrong, that happens, time to change course.
all songs should start by explaining what color world the guy lives in, what color everything he sees all day and all night is, whether this applies both indoors and outdoors, etc.
Hmm I might end up getting regular use out of this one
bsky.app/profile/greg...
I am willing to accept that I could have phrased the question better, consequently I have been owned, and that the underlying point I was responding to was itself incorrect due to the point I was trying to make but, again, could have phrased better
Interesting I didn't know that. Thanks for that one fellas
yeah I understood it better when I learned that if you just order a regular cup of coffee and don't specify they're going to give you cream and sugar. mostly people just want to be given cream and sugar without having to ask for it or think about it
it was never, like, handwringing about specific good cause X being overshadowed. it was always working backwards from supporting donald trump to find a justification.
who is putting guns to regular people's heads, making them go to chatgpt.com, and use more AI month after month after month?
seekingalpha.com/news/4505254...
no, that possibility went out the window when he got the Nazi tattoo and went on the Nazi podcast etc. He appears on the surface (of his literal skin) to be a Nazi. Your view is the one that requires an explanation for why he's not who he appears to be
the answer will surprise you
it is 2026 and people still don't understand that the answer to "why was this article written" is "because thousands of people like you reposted it"
Sure thing, and thanks for the post! It's a good illustration of the broader point that many independent researchers + low precision + selecting on significance is a bad equilibrium for generating valid discoveries
Second, if we do just look at expected values, it looks to me like you're asking not exactly for unbiasedness but for something like calibration: $\E[Y(1) - Y(0) \mid \hat{\tau}] = \hat{\tau}$.
Quote from blog post with highlighting: "I think a useful starting point in these discussion is to say what you want out of the published literature. I don’t have a full answer, but one thing that I would like is for published research coefficients to be unbiased. If a published result has an effect of 1, then I would like to (correctly) believe that if I were to somehow average across all (published and unpublished) research conducted on this topic with similar setups then I would get something like 1. Ideally, this would also be informative of future replications of the research too." The highlighted portion is the sentence beginning “If a published result has an effect of 1 …” and ending “… then I would get something like 1.”
This is interesting. First of all, I think one could argue that minimizing squared error (and therefore caring about the bias/variance tradeoff) satisfies this. It depends on what your metric for "like" is.
Perhaps we could learn something by taking the difference of these changes. A "difference of differences," if you will
she's right, there's a good chance there's toxic mold, carbon monoxide etc. merid.org/insights/env...
They didn't receive any lawful orders to follow
over time people forget what the $10k is paying for
**You're absolutely right**—that wasn't the right target. That's on me.
## The Problem
When I was selecting a target for the precision bombs, I forgot to double-check
I feel like he used weasel wording on the autonomous weapons that gives him room to say later that he never said their models wouldn't be used for autonomous weapons (just that humans would be responsible for it). If he wanted to say "never using our models in autonomous weapons," he would have
the fucking smol bean pfp, to top it all off
Cool paper! I'd like to see how the results look if they tune hyperparameters
yeah at least the bundle of sacrificing the environment and stomping on labor in service of actual real business interests has benefits for the economy
Waymos are operating autonomously every day in SF, LA, and Phoenix. People count on them to get to work, get home safely, etc. There are human operators to help with some edge cases, but that's largely stuff like "I left my wallet in the Waymo" or "I just got in a Waymo and someone threw up in here"
In some sense I think they must--if their drivers don't get paid the cost to operate their cars plus a market-clearing wage then they don't have drivers. The fare you pay when you ride an Uber has to be the cost of operating a car + driver earnings + Uber's cut
