I hope to preprint my paper for your review on arxiv next week titled:
"A Novel Bridge from Randomness in Stochastic Data to, Like, OMG I'm SO Randomness in Valley Girl Entropy"
We will pay dearly for overloading that word. Good AGI will be capable of saying the most random things! Not, really, no. I mean, they'll still be pronounceable, I'm guessing?
You kind of have to read the article to understand the big bold headings.
He is postulating that the way we deal with physical memory (an example of L2 and L3 caches is provided) demonstrates that, as we proceed with trying to form AGI out of our classical computer architectures, there are some fundamental problems for example larger caches are slower. With human intelligence, this doesn't always seem to be a problem for some humans. If you understand the "attention" part of recent developments in this field, he's saying that transformers are the most efficient way we got to achieve that and it's starting to look like a problem from a "physicality" standpoint as the author puts it.
The "physically" here is that larger caches are not just computationally larger banks of data but are actually physically larger and, by Euclidean distance, further away. Yet paradoxically the elephant nor blue whale is not the smartest brain on the planet, the distance from the center of my head to broca's region seems to have no effect on my elocution. Etc. Studying Einstein's brain doesn't do much (I guess the insulators are somewhat important?) for understanding Einstein's intelligence ... but that is 100% critical when understanding L2 and L3 caches on a die.
> Who is to say biological computers won't be built?
No one is saying that. We're just pretty sure it's not happening in our lifetimes.
I think you misunderstand the author's intent. He's not saying "The things I can't imagine are not going to happen". He's trying to argue that, "Look, the way things are going, the diminishing returns we are already seeing, the way our hardware works, this isn't going to get us to AGI." Of course, if you had some new architecture or weird "wetware" that somehow solved these problems I'm sure this article would concede that that's not the point.
You do realize that before ansible there was a whole generation of scripters thinking "yay! some framework that is just going to be endless proprietary bugs I can't fix!"
I asked if you tried it, it sounds like you have I guess. I'm sorry you did not find another tool for your toolbox. I did.
What if you took those scripts that you have used to automate your life, dumped them into something like cursorAI and asked the model to refine them, make them better, improve the output, add interactivity, generalize them, harden them, etc?
Sometimes when I have time to play around I just ask models what stinks in my code or how it could be better with respect to X. It's not always right or productive but it is fun, you should try it!
So you're effectively arguing that quid pro quo isn't quid pro quo if you can't prove the counterfactual (that this contract would have happened without drilling rights)?
Doesn't that seem ... stupid? To put the onus on the people who have no access to any communications or details about what looks like corruption? It's my opinion that without forcing them to do everything in the open and prove to us that there is no quid pro quo, you're going to end up with Soviet USSR style government real fast.
I'm sure this is "business as usual" but ... maybe it shouldn't be?
It could be referring to the the underpinnings of how these things are used.
Use race as a dimension for something and that ends up as a value in a vector that packs a human into a discreet set of pigeonholes. Then take many of those and stack them and you've got a matrix ready for things like principal component analysis or CNN training.
You might say "oh come on, that hasn't been done since WWII by IBM" and you'd be wrong. It still happens today with things like calculating insurance premiums and approving bank loans. And your response might be "no way, nobody records someone's race" and while that might be technically correct, we frequently harvest things like income and interest in products that are highly correlated with spacefic races (some innocuous others much less innocuous). This can be harvested through cookies in websites like facebook or they can be self reported income on credit card applications.
You can disagree that it's the same as saying "matrix multiplication is racist" but that is just a boiled down way of saying "we are very good at hiding racism in our algorithms and then acting super surprised when someone points them out and our defense is that we just did some math."
> Why don't you quote his article and then the wikipedia "interpretation" that is "very biased"? Let's take concrete examples.
That's a request for concrete examples. Regardless, why don't you provide concrete examples of the liberal bias in those articles to back up your claims?
Again I requested concrete examples. Please post what he said and what's actually word for word in the summary you are claiming is biased. It's really easy to talk abstractly and say whatever you want. I think you'll find that once you try enumerate this, it's hard to defend Sanger's assertion that the articles are clearly biased towards liberal positions.
Why don't you quote his article and then the wikipedia "interpretation" that is "very biased"? Let's take concrete examples. I'll start, here's an excerpt from the section in question in Larry Sanger's article:
> The rest of the article—which, I confess, I did not read entirely, as it is very long
He cares so much about the bias that he can't be bothered to find it. The only thing I saw (and I mentioned in my original post) is that he took issue with the description of statue removal phrasing. So he skips through three long articles, doesn't read them all but finds what he's looking for. This is largely how people operate today, they start with what they want to believe then they go looking for it.
> A neutral treatment would, of course, give broad factual coverage of such things as where the rioting took place, how many people were arrested, and numbers of injuries and deaths attributable to the rioting. The main Wikipedia article actually seems to do a good job there, as far as I can tell.
Sorry he said "good job" and I said "great job". I suppose there is a difference there and I apologize for that.
Nowhere did I claim that there were equally slanted liberal compared to conservative changes. Are we going to play the "reality has a well known liberal bias" game? Am I supposed to find articles on wikipedia that claim Biden lost the election or that no one should get a vaccine? I mean, this is a political statement: "Now that we have tons of vaccines, every death due to covid is a travesty." You know at least in the USA it's hard to find mainstream media that is conservative because those views have become so asinine. Is that my problem? Is that Wikipedia's problem?
What next? Are you going to complain that 98% of college professors are liberal? Okay, go change it. Enjoy trying to get people who enjoy liberal endeavors to promote conservative views. Am I shocked that the people who want to freely edit and debate articles on a website for no profit happen to slant liberal? Not really.
> (I don't think this is possible, therefore I think you are wrong)
What part exactly was I wrong about? I suggested you look at articles from 30-50 years ago and point out the liberal slant. Instead you invented a claim that I did not make so that I would be incorrect. Strawman.
If we cared enough about this, we'd all be reading conservapedia.com.
This is a very poorly written article that myopically focuses on only the most extremely recent events in the culture war. Wouldn't a better article look at how wikipedia handles, say, the the civil rights act of 1964 and the rise of Evangelicism? Things that are, you know, many years ago and should be much more stable articles than these fly-by-night everyone is having a conniption over at this very moment?
Reading this article is kind of like I'm watching Jerry Springer except it's not as entertaining. As things stabilize the articles get better and as time stretches to infinity so too does the editing work eliminate bias on Wikipedia. Is it the greatest thing ever that you should unquestioningly read? Not even close. The author thinks it should be impossibly neutral. Nirvana fallacy.
Honestly the section of the article on "The Antifa/BLM riots" was absolute garbage. He frames it that "National Democrats generally supported the rioters; portrayed them as “mostly peaceful” activists against fascism and racism, even contributing money to their defense; took seriously the notion that we should “defund the police” or backed similar police “reform” proposals; and stubbornly minimized the months of bloodshed, danger, and destruction the riots caused." But then he notes that at least one article included all this information and also agrees that the main article does a great job in neutrality. Then he has a problem with how monument removals was framed. sigh okay this guy is clearly cherry picking.
Trash article from someone trying to stay relevant.
Nowhere in the article does it mention the cost to the healthcare system per innoculation. It's also an article written to make it sound like Merck just wants to save lives but you need to also consider the billions off dollars this would net them. Also, buried in the article was this gem:
"Though most experts have little doubt that the vaccine prevents HPV infections anywhere in the body and therefore the resulting cancers, technically, this has not been proven."
Yeah this is crazy neat but the unnecessary precision is a real bummer that could be solved by just restricting things to three decimal places. For example when I use a utility like https://jakearchibald.github.io/svgomg/ to optimize, I see a 3.53k -> 2.21k reduction for the sphere with lighting. This is not just from precision reduction but also using simple paths over polygons ... there could likely be increased size savings by "rolling up" the fills in the final output.
Still a crazy neat tool but targeting edges as paths instead of polygons would save quite a few bites when hosting this stuff.
This was discussed in a really interesting way in the Pearl/MacKenzie book "The Book of Why" which I heavily recommend for people interested in cause & effect. Really opened my eyes to a lot of things I had been doing statistically but never known formally what was going on. http://bayes.cs.ucla.edu/WHY/
Criticism against Tesla is not rooting against Tesla. Playing options as a result of a CEO shooting his mouth off doesn't make you against the environment or against green technology. In this case, it just makes you a realist.
I'm an avid environmentalist. Elon Musk has lied so many times that I'm not rooting against him, I'm just afraid that his house of cards is going to come tumbling down and he will be another Solyndra where a good concept is ruined by a conman who has said that within a year you can make $30k/yr letting your tesla be a robotaxi and that in a year and three months there will be a million robotaxis on the road[0].
When criticism is valid, it's okay. We've seen Elon Musk lie about production time and time again. It's okay to distrust him. That's not "rooting against the environment."
"A Novel Bridge from Randomness in Stochastic Data to, Like, OMG I'm SO Randomness in Valley Girl Entropy"
We will pay dearly for overloading that word. Good AGI will be capable of saying the most random things! Not, really, no. I mean, they'll still be pronounceable, I'm guessing?