Likewise here in Victoria, BC. I don’t look at anything other than what I’m trying to sell. Unfortunately, other means of selling locally are virtually dead.
Facebook marketplace is about on-par popularity-wise with kijiji these days. Not too long ago I had to sell a used car and put it on both and received about the same number of genuine interest responses from both sites. Kijiji is a better site in my opinion; better layout, ux, search options and filtering etc. But facebook has more people already on it so network effects.
It seems like maybe a year or two ago this guy was fairly reasonable, and gave what seemed to be genuine opinions on the benefits and drawbacks of lockdown strategies. Sad to see him succumb to audience capture and make what seem to be willful misinterpretations of data (outlined well in jiggawatts' comment). If I as a second year undergrad who has never taken a stats course can see these statistical problems, then a PhD with multiple bachelors definitely can.
Where is the statistical problem if I may ask?
You can twist it however you want but we do have excess death that the very same "authorities" who provides the statistics do not know what causes it.
What is your explanation for why they publish data all over the world that indicates that people die at higher rates than expected but not due to COVID? Because that is what the data shows. Is the data just bad? Then its not willful misinterpretation but just bad data interpreted correctly.
I don't see why you would implement google docs like multi person editing. There's a reason basically every VCS has discrete commits and anybody who has ever coded anything knows why.
Some bars and clubs require you to use the app to scan a QR code for entry if you use a digital license. Other places you can use screenshot of your twin's drivers license because you forgot it and didn't have it set up.
NSW "Clubs" ("Registered Club" or "RSL Club") are required by legislation to record entry. Pubs aren't, and nightclubs would only bother if they've had a major insurance claim and were trying to avoid hikes in their premiums.
What an "RSL Club" is, and the nexus of legislation and idiotic operational rules attached to their tax free status, can be a little too difficult to explain to an American. But operationally it's similar to a Native American reservation casino with the same negative social effects on the surrounding community.
The number of bots misses the point for advertising too. If a company spends $X on Twitter ads and measures a boost of $Y in sales from Twitter leads, how many humans vs bots viewed the ads is neither here nor there.
That said, the question is certainly worth asking beyond the realm of advertising. Media has a significant impact on people and society beyond its profitability.
It depends on whether you're paying per impression or per conversion.
If you pay an advertising company a rate per view and the advertising company charges you knowing that a large percentage of those "views" were fake then you're being ripped off.
Well, that depends. If I pay X per 1k views what I care about is how much I sell as a result of that. If I sell a lot and profit would I care that some huge percentage of my views were bots? Conversely, if I sell very little would I care that there were zero bots?
If you are advertising just to get an idea out, or for something you have no capacity to measure, sure. You need to know accurate numbers. But if you can track conversions on your end, and conversions are good, does the exact number of real views really matter?
Of course it matters if you’re being billed by the impression! No matter how good your conversions might be, your margin would be unquestionably higher without the fraudulent impressions.
not necessarily. It really comes down to how the market sets the price per impression. If the impression cost is ultimately set by conversions, removing the bots will simply raise the price for impressions and you end where you started.
Most Twitter ads are brand advertising, not performance based. And it’s harder than you think to even track conversions from Twitter ads (because of cookies/incognito mode/lack of referral data from some browsers).
On some OpenCL implementations (as in, the ones supporting the SPIR-V OpenCL profile instead of just OpenCL C), this is possible.
Intel has a cleaned up variant of OpenCL for the purpose of running SPIR-V binaries called oneAPI L0, and Zluda can run on top of it. https://github.com/vosen/ZLUDA
AMD used to support SPIR in their driver, but they removed support for that in 2018. As such, you won’t get anything useful out of that one on AMD GPUs.
One reason why OpenCL 3.0 is basically OpenCL 1.2 rebranded, is because no one cared about SPIR profile.
OneAPI is focused on SYSCL, which is targeted at C++ and backend agnostic, it does nothing for Fortran, .NET, Java, Haskell, Rust, or any other language targeting PTX.
CUDA won thanks to embracing GPU polyglot programming and best in quality development tooling.
It is more a failure of the competition to uplevel their game than NVidia's own efforts.
Yeah, all of that goes beyond the OpenCL standard as defined by Khronos, specially since OpenCL 3.0 dropped SPIR-V, and OpenMP was never part of OpenCL.
Can confirm. Beggars and small street paddlers in China all just have their WeChat Pay QR code printed and laminated or you can scan their phone. Hell, summer of 2018 when I was in Hong Kong there was an American tourist(? granted, he did look like one) that was begging with his QR code printed so he could "fly back home".
