>Imo, the bigger question is why Hungary, Poland, Spain, Greece, and Cyprus (all countries part of the recent EU Spyware Scandal) [1] continue to allow their Interior Ministries to attack the phones of both domestic and foreign opponents, and are abusing "Spyware for political gain" [2].
It's also a national security issue given that Israel may be piggybacking on the spying.
The management console for a lot of these products tend to be hosted on-Prem behind fairly restricted network controls with traffic closely monitored by customers.
While this doesn't mean that there probably isn't some phoning in, it's extremely difficult to obfuscate.
That said, the act of purchasing a product like this can absolutely be used as leverage, but that's any sort of weapons sale (which this functionally is)
>web3 is a culmination of late capitalist grift, ironic meming, libertarianism and mansplaining, all resting on bad tech that is damaging the planet ... Proof of Stake is an alternative to validating transactions that doesn’t involve setting fire to the earth
The very first sentence of this piece is dishonest: blanketly condemning "all" related tech as environmentally devastating. Later on in the piece it's admitted that innovations like Proof of Stake solve the problem of needless energy consumption.
>To actually use this stuff is quite difficult, even the most simple thing, like buying something.
Yeah, scanning a QR code with your phone - as one needs to do to buy something online with a mobile wallet - is super hard. ;)
This article seems to be a cavalcade of shallow and dishonest criticism.
There's lot of edgy people on Twitter... lots of folks with hammers and sickles, ironic or otherwise, in their handles, despite the many documented atrocities committed by communist states.
Tankies are true believers, just like Trumpists, or the typical soccer moms who picket outside Planned Parenthood, but when their daughter needs it they obviously use it, then back to picketing the next day.
"This time it will be different."
There are also various explanations of how and why the Holodomor just "happened", and it had nothing to do with communism/socialism/USSR/Stalin, or it was Nazi propaganda, etc. ( https://www.reddit.com/r/communism101/comments/izcxbk/i_saw_... - with reputable sources such as sputniknews :| )
If the cracker were being marketed by their creators as “the easiest way to steal money” would it be any different? Intent does matter. In fact for a crime to have been committed one has to prove intent. If the intent is to facilitate the robbing of banks, that’s different than “pen-testing” as a means to help/protect.
Ooh "diverse", is it because "homogeneous" countries would have the disease better in control, so like many things, it's a problem attributed to "diversity"?
Whenever I read "homogeneous nation" I take a glance at my racism-o-meter...
I think the problem isn't heterogeneity in ethnicity, it's heterogeneity in values. Many western European nations have a good amount of ethnic diversity, but they don't have half the population saying that nationalized healthcare (like in the UK) is "socialist" and they don't want it. The different ethnic groups there seem to be able to all agree on what kind of things are necessary to achieve an advanced, modern nation. We in the US simply cannot.
Where did you get those numbers? It certainly isn't seen in election results. Approximately half the population reliably votes GOP in every election, and they certainly don't want Medicare for All. On the Democrat side, MfA is seen as "extreme"; even in the current Primaries (which is mostly composed of voters who are reasonably strong Democrats), Bernie is only getting about 25% of the vote so far.
I have no idea how reliable these polls are, or how much things have changed since they were conducted, but I've heard the statistic before and I at least got the impression that it wasn't considered a controversial one.
From what I understand, there are similar statistics about various other policies that could be considered 'socialist' (but are more European-style social-democratic). While many Americans are strongly opposed to anything with the label socialism, they seem to often actually be in favor of the actual policy that these 'progressives' propose.
I haven't taken any kind of deep dive into this though, so for all I know this is more controversial than I assumed.
This article is making claims about several polls that Americans supposedly want a bunch of things that very progressive Democrats (not just regular Democrats) want. It sounds to me like their polling methodology is severely flawed, because their poll results are completely different from our election results.
This reminds me of the 2016 election. All the poll results said that Hillary would easily win the election. So much for those polls.
It's possible that the polling methodology is flawed, but I don't see the problem with there being a big difference between what people claim to want and who they vote for.
In fact, I'd say it's rather unsurprising that, whatever people might actually want, they would vote against their interest because 1) socialism is evil, and 2) "my team should win, no matter the policy".
Furthermore, there's a difference between a poll that predicts an election winner, which is notoriously difficult to do, and a poll that simply asks people whether they want <thing x>.
It's possible the wording of the questions is manipulative, or that they happened to poll an unrepresentative sample of the population. I don't know at this point.
But the fact that 1) the results of an 'issue' poll don't align with voter behavior and 2) that the results of an election winner poll doesn't in itself make a good case that the poll must be flawed.
>they would vote against their interest because 1) socialism is evil, and 2) "my team should win, no matter the policy".
If people are honestly this stupid, then why even bother polling them to see what policies they want? This is a serious question: if people don't bother voting based on policies they want, then why is that even important? If this theory of yours is true, then it really seems to call the voting population's basic intelligence into question, not to mention the fundamental idea of democracy itself. If the people are too utterly stupid to vote for their own best interests at least most of the time, then why wouldn't authoritarianism be a better system?
Honestly, I find it easier to believe that the polls are bad. I've seen way too many polls in my lifetime where the questions were manipulative, or just plain poorly worded or chosen. On top of that, the other problem with polls is that they only capture results from people who actually bother to respond to the polls, or are in a place where the pollsters are. Usually, when the latter is brought up, we're supposed to somehow believe that the pollsters have some great methodology that lets them determine what a larger population wants just by sampling a much smaller number of people in that population, but this is completely subject to bias and error. And finally, I point again to the 2016 election: predicting a winner is not as hard as asking if they want <thing x>. In this case, the poll is asking exactly the same thing as the election itself will: who do you plan to vote for? Asking about <thing x> is not as easy, because you can get wildly different results by asking the question differently. Do I want "Medicare for All"? Well, the answer might depend on many things, like how exactly it would be implemented. Is private insurance still allowed to exist as a supplement or alternative? Are non-citizens/"illegal aliens"/"migrants" (pick your term) covered? etc. You would probably get very different answers depending on those points in a large poll in America, so if you just ask a simple "do you want it? yes/no" question, that might not align at all with what actual candidates are proposing. Asking "who do you plan to vote for?" avoids that. And in the 2016 election, the polls were nowhere even close to the election results. That tells me the polls were either garbage, or rigged to try to influence the election.
> If people are honestly this stupid, then why even bother polling them to see what policies they want? This is a serious question: if people don't bother voting based on policies they want, then why is that even important? If this theory of yours is true, then it really seems to call the voting population's basic intelligence into question, not to mention the fundamental idea of democracy itself. If the people are too utterly stupid to vote for their own best interests at least most of the time, then why wouldn't authoritarianism be a better system?
All kinds of things play a role in this 'stupidity': tribalism, decades of anti-socialist propaganda, legitimate concerns based on history, and disengagement. It's completely understandable that one could reject 'socialism' and yet support most of what actual socialist argue in favor of.
> Honestly, I find it easier to believe that the polls are bad. I've seen way too many polls in my lifetime where the questions were manipulative, or just plain poorly worded or chosen.
If you're honestly engaging with me on this, and I'm hoping you are, I'd challenge both of us to do some research and further discuss this. I'm inclined to believe that in this case, with concrete policy questions, the questions and answers are not too manipulative, but of course I don't "know" this. Let's look into it.
> On top of that, the other problem with polls is that they only capture results from people who actually bother to respond to the polls, or are in a place where the pollsters are. Usually, when the latter is brought up, we're supposed to somehow believe that the pollsters have some great methodology that lets them determine what a larger population wants just by sampling a much smaller number of people in that population, but this is completely subject to bias and error.
True, but afaik this generally skews in favor of older, more conservative respondents and, crucially, more socalism-phobic, so to speak.
> And finally, I point again to the 2016 election: predicting a winner is not as hard as asking if they want <thing x>. In this case, the poll is asking exactly the same thing as the election itself will: who do you plan to vote for? Asking about <thing x> is not as easy, because you can get wildly different results by asking the question differently. Do I want "Medicare for All"? Well, the answer might depend on many things, like how exactly it would be implemented. Is private insurance still allowed to exist as a supplement or alternative? Are non-citizens/"illegal aliens"/"migrants" (pick your term) covered? etc. You would probably get very different answers depending on those points in a large poll in America, so if you just ask a simple "do you want it? yes/no" question, that might not align at all with what actual candidates are proposing. Asking "who do you plan to vote for?" avoids that. And in the 2016 election, the polls were nowhere even close to the election results. That tells me the polls were either garbage, or rigged to try to influence the election.
I think with the election polls one of the big issues is exactly the problem you describe earlier. Getting an accurate, representative sample is very difficult. Especially when voting for Trump was, for many people, a kind of social suicide.
That said, you might be right. I am inclined to believe that policy questions, even with all the hidden complication behind implementing said policies, are probably answered honestly and I'm assuming that the sampling is relatively 'honest' too. But I don't know this, and it's absolutely fair to challenge me on that.
What I'll do, one of these days, is look more into this because while I don't think my beliefs depend on the accuracy of these polls, I do think it's important that my beliefs are truthful (as well as my claims). If you're not just interested in defending your beliefs or attacking mine (which I'm not saying you are), I'd very much like it if we could both look further into this and at some point continue the discussion with more than our respective assumptions :). I'm happy to be corrected if the facts are on your side.
the point is the headline and the tone. there is absolutely no empathy in it. you don't need to be for or against authoritarianism to spot that. and I feel sorry for anyone who doesn't frankly.
It's also a national security issue given that Israel may be piggybacking on the spying.