It might “address” the criticism but it doesn’t rebut it. It’s absolutely true that MNIST can be misleading and the results on it might not be representative.
In addition to removing NSFW images from the training set, this 2.0 release apparently also removed commercial artist styles and celebrities [1]. While it should be possible to fine tune this model to create them anyway using DreamBooth or a similar approach, they clearly went for the safe route after taking some heat.
They are right. Climate change is absolutely not an existential risk.
“In a high-emissions scenario where little is done to curb planet-heating gases, global mortality rates will be raised by 73 deaths per 100,000 people by the end of the century”
It would be very bad for the tech progress in the West if these types of lawsuits succeed, while China can do this. The legislature would have to act to explicitly legalize it and that could take years and it could be too late by then. The US is already falling behind China in the AI progress.
She opines on a lot of things she has very little knowledge about and her writing style is terrible. But it’s crack for HN readers who mistake skepticisms and snark for insight.
There was another fun example of how you can’t rely on supposed experts when a University of Nebraska professor Ana-Rhodes Short embarrassed herself by weighing in on the poker cheating controversy. She created a whole game tree assuming that the player who claims that he got cheated could only have one possible hand instead of a range of hands and also somehow thought that running it twice would change the odds. She also claimed that she used it as a lesson for her unfortunate students. Then she got extremely defensive, blaming toxic men, etc. when poker players pointed out her errors. (For the record I’m leaning towards no cheating and extremely poor handling of the case by the player who claims he got cheated).
Exactly. By adding more and more ads, their own widgets, and pushing down smaller sites in the search results to favor brands, Google made it very hard for smaller sites with user-generated content to grow and profit organically. And then fewer people searched for them or expected to find them.
I’m kind of enjoying the fact that newer sites and apps like Discord don’t let Google index them after Google killed off a large part of the independent Internet. We’re missing out on a lot of good content that would be put up in an alternate universe with a less greedy and shortsighted Google.
What struck me when reading through many outraged tweets of artists was how it was all about the rules, copyright, the art critics, the other artists, the financial aspects, and most of all personal attacks on the submitter. Nobody gave a damn about the art consumers and their preferences. Sure, it probably should become a separate category in competitions and it could suck for lower-level artists. But it makes me think that this whole industry needs a realignment.
1/4 mile became the standard measure for a good reason. 0-60 is just as much a measure of traction as it is of acceleration and you can go to a drag strip and test your car against others legally in 1/4 mile.
I’ve seen this type of an argument before and it seems not just skeptical for the sake of it but super lazy. The main reason for progress is that people with outdated ideas die and if they don’t, the progress will end - that’s the best we can do? A bastardization of Max Plank’s idea, ad hominem, (justifiable) criticism of western approach to death, and anti-capitalist motives all blended into some kind of pro-death philosophy? Maybe let’s reform how science operates to promote innovation in hundreds of possible ways instead of telling billions to die because of an idea if we don’t have to? And is the quickest innovation necessarily even a good goal? AGI presents serious dangers.
Not too be overly negative but Hossenfelder herself is a good example. She posted many authoritative-sounding videos about topics where she has little knowledge and it shows quickly. She’s a walking illustration of Gell-Mann amnesia. But clicks are good for her business, just as they’re good for the business of newspapers.
Nah, it’s a real thing. You know how some NBA player can be top 10 (but not #1) in scoring, assists, rebounds, and steals? Good chance he’ll win the MVP.
I’ve driven multiple Ferraris and Lamborghini on various tracks for many hours. I agree that Ferraris are better on the track in general: historically they’ve been lighter and RWD vs AWD. You’re still more likely to see them during track days than Lambos and Ferrari as a company does more racing. The track-oriented models like 458 Speciale (I haven’t driven it) are usually even more visceral because of a louder exhaust note, harder suspension, being lighter and lower, racing seats, etc.
A world where nothing but conscious experience exists and what you think as the world is simulated would be indistinguishable from your experience. Ergo, the “real world” does not exist.
There are only two ways they can be losers: they are not GTO or the rake is higher than their winnings. Hold’em is pretty much solved (almost because approximations were used for bet sizing). It is possible that you can do better by deviating from GTO to better exploit bad players but that makes you vulnerable to be exploited yourself.