In this case, because they all had to read and write English. The Indian and American programmers probably grew up speaking English. The Russian and Chinese programmers who took the test were selected for preexisting proficiency in nonnative language.
Yeah, but if USA and India had the majority of contestants, the bias is in their favour
That's not how statistics work. If the samples are random, but different in size, the smaller sample will have a more biased average. So in this case, if we assume the samples are random, we just know that the Russian and Chinese averages are much less representative than the American and Indian averages.
However, we also know that the samples were biased. American and Indian programmers speak English. The testing company is based in Palo Alto, CA. Their website is totally in English. The Russian and Chinese programmers were selected not only for motivation to take an online test, but for a preexisting ability to speak a nonnative language.
The old tired argument that there is a lack of qualified workers in this country is pretty much not in line with reality.
He's right. Show us the wage spike and we'll believe in your worker shortage.
I can't help but wonder if this is only a measure of publicly known hacking
Hacking? TFA is just about test scores.
On a side note, the sample size of 1.4 million doesn't matter if the sample is non-random. Many more Indian and American programmers took the test, and their average scores are most likely lower for that reason, even if there is no additional bias in the Russian and Chinese scores.
That's a big "if" though...what India and the United States have in common with the testing company is the English language. The Russian and Chinese samples are sampling Russian and Chinese people who speak English.
they did a bad thing
Not really, apparently he had permission.
if the bot used Twitter to build its responses
Actually it's more basic even than that - from what I've read today the bot would obey requests to parrot incoming content. Most of the crazy things it said were literal repetition of such inputs, though I guess eventually whatever pretraining it had was overwhelmed by the new inputs.
There is plenty of space for more highways on the peninsula, even apart from replacing BART by roads
You admit that "replacing BART track with roads" was at best a non sequitor, and then refer to subregions not served by BART. OK, then.
the greedy, privileged minority
Right, it's the people who don't own cars that are the rich and privileged. You need to get out more.
half their transportation costs paid for by other tax payers
Does making up arbitrary numbers to suit one's ideology fly in your line of work? Interesting.
Ripping out BART and replacing it with roads and buses would be a start. There is plenty more space.
You're shockingly ignorant of the greater SF Bay region if you think that statement makes any sense. It would mean multi-deck city streets, extensive car tunnels under existing roadway, and so forth.
I have a problem with rent seeking, corruption, waste, and forcing people into poverty.
Right...accessible transit is forcing me into poverty...by making it at all possible for me to get to work.
I find this somewhat amusing given my experience when Metro Transit went on strike. I found that when they were on strike traffic improve slightly during my commute times. That may have been better planning on people's part or because of the lack of giant mostly empty buses getting on and off of the highway.
Dunno which MT you're referring to, but the Bay Area sees crippling traffic slowdowns during BART service interruptions such as the strike three years ago.
The argument is that adding a few more outlying stops over the last 10 - 20 years has caused doubled ridership at downtown stops in the past 5 years, which happen to coincide with major economic and population booms in the region. It's pretty clear that the latter, rather than the former, is causative for the increase in system load.
The whole idea that BART ridership increases in the past five years have been caused by limited service expansion over the past two decades is pretty dubious. Public data and statements from BART make it pretty clear that it's the core downtown SF/Oakland stations, and certainly stations between Concord/Fremont and Daly City (that is, original stations from the '71 - '72 openings) that have seen the most ridership growth. That growth in ridership has occurred because of wide spread economic and population growth in the Bay Area, not expanding service to past residents.
One man's "magic" is another man's engineering. "Supernatural" is a null word. -- Robert Heinlein