Yves here. Tom Neuburger has a short post with Twitter Files updates, plus evidence that Google is obstructing searches that ought to lead to his site.
In a bit of synchronicity, I stumbled upon another Google gimmick for messing with search results. I had seen a tidbit on Interfax that is relevant to a new John Helmer post, wanted to include it in an introduction, and so went to find it.
This is what I got:
I had not clicked on anything. If you have very good color vision, you might be able to detect that the second title is a tiny bit pinker than the others. That is the one I had looked at earlier. That behavior holds across all searches. I don’t know why Google has implemented a barely perceptible color difference. We have a very clear color difference for read v. unread links at Naked Capitalism.
But that is not the striking part. Google has subjected me to a feature in beta telling me Interfax is Russian. I tried the same search with site:washingtonpost.com and got no such interference.
This is what you see when you move the scroll bar down:
Moreover, in the bottom field, immediately elow “Send feedback,” you see “Remove result”. So at a minimum, it looks as if Google is hoping you’ll remove sites it deems unsanitary from your search results.
Now based on this surprise, my surmise is Google is doing this only to Russia sites. But I wonder how and when this started. I have sometimes used Google to search en.kremlin.ru, since its search function isn’t very good; Firefox remembered those searches and Google did not put up an interfering warning when I ran one of those old queries again just now.
I would be very interested to hear from other readers who have gotten similar results. And yes, Google is terrible but Bing is pretty similar and I have not found Brave or DuckDuckGo to be much better. Qwant is also a disappointment. Yandex give much better first pass results but does not have date or date range filter, so most of the time when I try Yandex, I wind up back at one of the bad boys.
By Thomas Neuburger. Originally published at God’s Spies
This week has seen a number of developments, including the release of four new Twitter Files, several of great significance.
- Twitter Files 13 — Twitter and Covid
Alex Berenson, January 9, 2023 - Twitter Files 14 — Russian Bots and #ReleasetheMemo Hashtag
Matt Taibbi, January 12, 2023 - Twitter Files 14a — Supplemental — More Adam Schiff Ban Requests,
and “Deamplification”
Matt Taibbi, January 13, 2023 - Twitter Files 15 — How the pharmaceutical industry lobbied social media to shape content around vaccine policy (Intercept version: “Covid-19 Drugmakers Pressured Twitter to Censor Activists Pushing for Generic Vaccine)
Lee Fang, January 16, 2023
(Additional brief comments here.)
The final file, by Lee Fang, is especially troubling. It documents how Pfizer and BioNTech attempted to use Twitter to protect its profit from the activist campaign to get them to make their vaccines available free or at low cost to developing nations, especially in the Global South.
From his Intercept write-up:
The campaign they were concerned about was the launch of an international push to force the drug industry to share the intellectual property and patents associated with coronavirus vaccine development. Making the patents available, in turn, would allow countries across the world to swiftly manufacture generic vaccines and other low-cost therapeutics to deal with the ongoing pandemic. [emphasis added]
That’s evil in its purest form — choosing profits over people whose lives are at risk — with Big Pharma caught choosing its profits and consigning many brown-skinned others to death.
Thanks to the Twitter Files, we now have proof of this, not just suspicions. And only thanks to the Twitter Files, I hasten to add.
Unfortunately, this week I’m in transition. Next week I’ll send several posts covering these developments.
In the meantime, I have updated this file (“Twitter File Links, Aggregated”) with links to the above Twitter Files (along with a few comments), and will continue do so whenever a new file is released. It will always be up to date, so feel free to bookmark it.
Finally, Google may be getting into the content-flagging act. As you know, I’ve been covering the Twitter Files pretty extensively. If you search on my name, “thomas neuburger”, and narrow the search to include only “Previous week”, you see this flag ahead of the results:
Changing the filter to “Past month” removes the flag. Is this meaningful? I don’t know. But asking if I’m “trusted on this topic” is worrying.
I’ll have my own Twitter tale to tell, perhaps, if I can find data to support my suspicion. But for now, if you get the same result from a Google search, please let me know, either in the comments or by email.
Thanks again to you all. More soon.
40 comments
Comments are closed.
Just FYI, DuckDuckGo uses Bing for it’s searches.
Correct, but I think it didn’t some time back….
I thought that duckduckgo originally had used google but without the tracking?
As I understand it, duckduckgo did originally do its own indexing and did not have any usage agreements with google. Then in an effort to shortcut to better indexes, duckduckgo made a deal with yahoo!, and after that is when they seem to have quietly made a deal with microsoft to use bing.
Overall I tend to search with javascript off, which tends to curb pop ups and annoying inserts, although of course, that does not prevent search result manipulation.
The CEO of Duckfuckno made a public statement last year that “we will do everything we can to stop Russian disinformation from showing up in our search results and we’re fully online with the the U.S. government in standing with Ukraine”. So definitely don’t use this search engine for anything related to the Ukraine war. You’ll only get State Dept approved propaganda.
My understanding is that DDG rents access to Bing’s web index, but that the filtering algorithm which narrows the web index into search results is DDG’s own. It’s been a while though, so DDG could’ve changed.
Getting a viable index is a barrier to entry for new search providers for at least two reasons. The obvious reason is bandwidth. Paying for the server farms to run the crawler bots to index the web is a substantial cost for any search provider, and it must be done before the search results are good enough to put advertising eyeballs on the search page.
The other reason is site access. Google’s crawler bot especiallly, but Bing’s also, get priveliged access to websites to build their indexes. That is Google and Bing bots are whitelisted by site owners so those sites can be featured prominently in search results. If DDG were to run their own crawlers they would be stopped by innumerable Google Captchas. (ugh)
Yesterday I searched the dot ru version of yandex (as opposed to the dot com version) for “Украина вертолет” which is “Ukraine helicopter” and above all the search results yandex displayed a message reading “Некоторые материалы в интернете могут содержать недостоверную информацию. Пожалуйста, будьте внимательны.” which put into google translate gives me “Some materials on the Internet may contain false information. Please pay attention.”
I thought that was a nicely moderate and decent way to approach things, if they must be approached even at all.
Last night I had a second email bounced by Google for linking to an insecure site. That site was a Yandex-translated Russian webpage. Same thing happened last week with another email containing a Yandex-translation link.
Then I took a proofreading quiz that former internet star Waxy had linked to a while ago and the sample article to be proofread was ripping on Elon Musk’s management of Twitter!
They’re about to offer us Biden’s head to assuage the mob. That’s not enough, and they damned well know it.
To the moderator: thanks for deleting my well intentioned post. I know I’m not welcome at this site any longer :-)
We did not delete anything. Your comment went into moderation. And it was just released, where you put it, in a different spot.
We have very clearly stated Policies, in a tab in the header bar area. I suggest you read them before attacking us on fabricated grounds.
Remarks like yours are an invitation to be blacklisted.
Shadow banning exists on Meta/Facebook. Bots proliferate and the AI suggest that they may be increasing in number geometrically?
Genuine journalism is truth and that is a dangerous act, especially now! The courts will delay and ensure that the changing paradigms mean Google will change slowly. Huge fines imposed in Europe, Russia, India and China.
Glaring lack of accountability in the nation of checks and balances.
Just say no to censorship. Or … put another way … support … defend … demand … freedom of speech.
It’s that important.
Doesn’t work if your voice is suppressed.
Or if your internet “demonstration” is kettled off in some obscure corner far away from the “First Page Results” of a search.
Riffing off of the maxim that “All Politics Is Local;” I’ll assert that the trend is towards fully ‘curated’ content on the internet. Thus, we are going to have to relearn the ‘gentle arts’ of direct political organizing; face to face, or perhaps over iron sights.
Stay safe. Organize.
It’s why ‘misinformation’ is such a clever word. It’s not necessarily incorrect information.
But no-one cares about product deceptions like “No added sugar!” when there’s raisin or apple juice instead as a sweetener.
Control the discussions, control the people.
Google can be handy if you want to, for example, look up stuff like the area of Denmark. But over the past several months I have noticed how their searches can be garbage if you are searching for something that does not fit the ‘narrative’. It gets very stubborn in refusing to show results for events that you know damn well happened, no matter how you word the search term. And of course if you flip over to the Brave search engine, it will show you want you want on the first page. In the same way that Twitter had the FBI, the CIA, politicians (e.g. Adam Schiff), corporate execs, etc. all having a say on who was and who was not on Twitter, it seems that the same is happening for Google in their searches. But what really needs to happen is to have all those people calling for censorship being called out on that, no matter how ‘nice’ a person they seem to be. If they are for censorship, they are not your friend even if they are doing it for your ‘own good.’
There’s a way around this. Leave the platforms. Begin doing all business on federated platforms. If enough people move their activity over, sooner or later the federated approach will have the swing to begin establishing their own web crawlers.
Ohhhhhh, but the “switching costs” cries the crowd! As they allow fascist convenience to trump ethics and morality.
Thank you for your impressions of Bing, Qwant, etc. When DuckDuckGo went to the dark side about a month ago I switched to SwissCows and so far it has been satisfactory.
I had high hopes for Swiss Cows. Not to bad if you simply ignore the first page or two.
I too switched to SwissCows on all platforms on which I could do so. I have no complaints, and no results of the sort Yves cites.
“The final file, by Lee Fang, is especially troubling. It documents how Pfizer and BioNTech attempted to use Twitter to protect its profit from the activist campaign to get them to make their vaccines available free or at low cost to developing nations, especially in the Global South..”
The jury is still out, but that could end up being a blessing in disguise for the Global South.
That was my thought as well. Let Big Pharma raise prices on those shots (not vaccines for sure) and maybe have to sit on millions of un-used and quite possibly un-safe SHOTS.
The incident is so egregious, my suspicion is that the word was out that Twitter was a rent-a-censor for anyone with influence/money/clout. A mere 3 years after they reluctantly rolled over for Senator Warner, they had given the mouse a cookie and the rest came calling for cake.
Yesterday.Turley put up this piece about NYT publisher Sulzberger and his Orwellian censorship=freedom pitch.
https://jonathanturley.org/2023/01/18/sulzberger-disinformation-in-the-most-existential-problem-facing-the-planet-today/
Of course the Times has always been a defender of the status quo through a variety of Sulzbergers* and one could argue the brief 1970s liberal period was as much a rejection of unfashionable Nixon and the by then indefensible Vietnam War, not a change of plan. When it comes to US imperialism they are for it.
Therefore the argument of the current A.G. Sulzberger that they are defending journalistic “independence” is itself disinformation. Their real goal is to block the independence of others.
Meanwhile it’s harder to see what Google is up to since nepotism doesn’t play a role. But being threatened by powerful Congress people and arguments such as the above may have a lot to do with it.
*https://en.wikipedia.org/wiki/Special:Search?search=sulzberger
Exactly. We need honest brokers for search, but we’re not going to get that as long as people embrace censorship as somehow a good thing.
Aren’t Sulzberger and the NYT walking an extremely fine line here? He is in effect, calling for his commercial competition, big and small, to be suppressed, or legislated away. For all their faults, you don’t see Google or Facebook calling for their competitors to be banned (maybe the Parlor incident aside). There are ‘free market’ appearance to keep up — for the present.
My understanding is that Google maintains a detailed profile on you, using cookies and your login ID. Many people have a gmail account, if only as a junk address, and remain logged in. Google thus knows who you are across search sessions, and has all kinds of stalker-ish profile information that gets used as ‘shadow variables’ any time you do a search.
E.g., if you search for “pizza near me”, Google uses not only your IP address, but other info from its stalker profile to decide what “near me” means. This means that your search results may differ somewhat (maybe a lot?) from those received by others. The censorship and rank-swizzling algos get added into the mix as well.
You can perhaps get around this profiling somewhat by using a separate browser that flushes all cookies and is never logged into your gmail account. Browsers like Firefox also have add-ons that can be used to spoof the User-Agent string, so you can disguise the identity of at least your browser.
If the main reason we’re using Google instead of Yandex is the ability to filter dates (I also use this feature regularly), I would say we should try to find some Russian speakers to nag/petition Yandex to add this functionality.
Yandex filters dates if you go to advanced search options (slider icon just to the right of the search field). I’m not registered (#$%! ph# required!) so I can’t permanently adjust my search settings which are different from Google but not dissimilar in function.
Anecdotal only, but I tend to notice that Russian sites are much more functional and much less cluttered than Western cookie-loving websites. Almost as if by design as if primitive Russian minds cannot grasp the all the wonderful business opportunities (I think they understand the tracking part somewhat well, Yandex “knows” I’m in St Paul which is only two hours by car and nine years ago by residence).
Mentioned here before but you can eliminate a lot of clutter and produce other surprising results by turning off style sheets. This is still possible on Firefox by turning on the top text menu and then View/Page Style/No Style.
And re Yandex–I’ve never used it. Is it in Russian and requires Google Translate? Doesn’t that defeat the purpose?
You can adjust the language settings, location, and time range to be searched. They also have search functions for video and images and a good translation page. They offer email services, sell ads and basically are duplicating all the useful parts of Google. They also seem to skew search results but far more lightly than I’ve become used to.
Bottom line: Are you more at risk by sharing data with Russia or with corporate America? I trust neither but for the life of me cannot imagine Russia using my data to suppress my social media presence or bounce my emails.
I recently downloaded Yandex and can’t figure out how to clear their cookies. Do they also leave stuff in your cache? Maybe one of the brethren can help me with this issue.
Alphabet has been unsuccessfully trying to diversify their income stream for years. Search is still the only major profit center for the company, and they are vulnerable to consumers making the simple choice to use another search engine.
Google does not have an advantage in search results anymore. Not because the others have gotten good, but because they have gotten much worse. Simply choosing another search engine consistently is the best way to hit them where it matters – their advertising revenue.
I personally use startpage.com because duckduckgo has chosen to go with maximum search result manipulation as well.
All the alternative engines, except for Yandex, use Bing, but how they use your data and how they add a layer of additional manipulation onto the raw Bing results matters
Only time I read TWITter is if someone I trust links it. I’m not registered on it, and never will be. The volume of noize on the Russian-Ukrainian propaganda end is the worse I’ve ever read/heard. Thanks for the heads-up. No surprise.
Morning Yves,
I’m pretty sure this question has been asked before, but has there been any consideration among the naked capitalism crew to set up a federated server? The NC crowd is pretty much legion these days. I’m sure it would populate rapidly given what this site represents to so many of us.
And, I didn’t see anything per se in the rules that keep us from disclosing our mastodon info for finding one another out in the federated universe. Is that ok?
As Thomas demonstrates, the problem is just getting out of hand and it feels like we need to make a concerted stand someplace. Somehow?
Thank you.
By “federated”, do you mean a Mastodon instance? Is there a federated solution for search, à la Google?
I have been an occasional Twitter user, but at this point I’m mulling Mastodon. There’s a learning curve, I’d imagine. Twitter will survive, though I’m doubtful that Musk can un-hose it and he’s also spook-adjacent, albeit not in the same woke configuration as those he ejected. Musk seems determined to crapify Twitter in other ways, e.g., shutting off the API for third-party apps with no warning, which in effect killed Twitterific overnight.
Mastodon is one example of federated architecture. Here’s a wiki article that explains more in depth: https://en.wikipedia.org/wiki/Federated_architecture
I’m far from qualified to comment deeply on the matter. But as I understand it the takeaway is that the architecture is distributed so that no single entity can dominate the experience. Google scholar provides a template by way of analogy. For example, one instance of a federated architecture might center on news. Another on entertainment. Many others on the different academic disciplines … ad infinitum.
The trick is that they all share their information under one interconnected roof. But, cut the head off one, all the others continue to exist. And, there are no rules specifying how many “entertainment” or “science” servers can exist.
Using Mastodon again as a reference, one has three feeds: i) Home feed is everyone you follow, ii) Another is the local feed that is all the members of the particular server one parks themselves under, and iii) the federated server that is a fire hose stream of all the servers that exist under the mastodon universe.
It’s not too difficult a leap to make from twitter. I did. I completed my migration just yesterday and closed down my twitter account. Most of the differences are trivial, and there is no reason someone couldn’t open a server that provides basically the exact same functionality as twitter. I’m not aware that one exists, however. That said, the more users climb aboard, the more motivation to continue to evolve the architecture.
Thanks for clarifying. Agree with all your points. I have some background in distributed systems (e.g., writing applications), so the architecture is all very familiar. For some years, I’ve been focusing elsewhere, though, so haven’t been keeping up with what’s actually available. As you say, I can easily see an alternative to Twitter being built. Scaling it and building out a similar ecosystem would be difficult, but not impossible. I’m less sure about search, since that requires vast amounts of memory and server power to index the web. However, it’s clearly something that is needed (I tend to think Google cannot be reformed), and a federated architecture does seem like the way to go.
If you want a date filter in yandex then use:
date:yyyymmdd..yyyymmdd
date:>yyyymmdd
date:yyyymm*
More yandex tricks: http://www.ph4.org/ya_yakod.php
Google really isn’t a surprise. The search engine has been quietly doing this for some time. It has probably gotten to the point where this is a big reason for the search quality going downhill in the last few years.
Remember, the Twitter files also revealed that Google, Facebook and others were also being “invited” to FBI lead meetings. In short the censorship was not just restricted to Twitter, at all. The rest have also been slowly implementing government lead censorship projects. If they were clever, these have been coordinated by, say UK or other foreign based companies to skirt around the 1st amendment. But I suspect the FBI have not been very clever here.
Also, recall when nakedcapitalism appeared on the ‘disinformation watch’ site “PropOrNot”? If I recall correctly, Google and/or Facebook were being “advised” by those sites on what to restrict/ban from their social networks. I suspect the entire apparatus is completely intact. A bit like the crew who organised the Syrian campaigns being shifted over to Ukraine. Plus ca change.
For what it’s worth I used startpage.com to replicate Yves’s search, I got the same top 5 results (4 and 5 being swapped in my case) without anything warnings re the provenance of these results.
startpage.com uses google as the underlying search engine but anonymizes the query so that the results are not customized based on your search history/browser/device/IP address etc.