Should Twitter ask you every hour or two what you want to see?

The Silicon Valley religion is that robot software can figure out what you want based on what you’ve done in the past. The result is that Twitter shows you tweets that are a lot like tweets with which you’ve previously interacted. But what if you’re in the mood for something different? Suppose that you’ve spent a lot of time on Twitter condemning hate, ridiculing Deplorables for their refusal to wear masks and accept experimental vaccines, exhorting young people to get out and vote for Democrats, and demanding additional investigations of the January 6 insurrection. If you open up the Twitter app as you’re settling down for the night, the application will show you political tweets. But what if you don’t want to see the same stuff at 9 pm that got you riled up at 9 am?

Suppose that Twitter had a “laugh” emoji option for a reaction to a tweet. Then it would be possible for a user to click “entertain me” and see the recent tweets that other users thought were funny. Based on text analysis, the system could respond to a “teach me” click with tweets that were educational in nature and/or linked to thoughtful tutorials. With a bit of merging of ChatGPT into Twitter search, perhaps this could be done as freeform text rather than a set of predefined moods, but I think the moods/interests button would be better (less effort) for most users.

Note that this could be done without cooperation from Elon Musk & Co. as a skin on top of Twitter by a search engine that had ingested at least a significant subset of tweets.

How does it work now? Spectacularly badly. If I type “funny” into the search box, the results are mostly people fighting. “funny tweets twitter.com” works on Google to find Twitter accounts that offer hand-picked items.

Here’s a site that tries to do the third-party skin… funnytweeter.com. Some of the tweets were a lot funnier than people hitting each other, but they can’t take advantage of anything known about an individual user, e.g., that he/she/ze/they likes knitting or aviation or whatever. Examples:

Readers: What do you think about this idea? Twitter should track your moods and figure out your mood by asking you… What are you in the mood to see?

Full post, including comments

How is Twitter identifying offensive content?

Here’s a puzzler from a Twitter exchange in which I was a passive observer. Twitter says that there are offensive replies:

If a person were to click on “Show”, he/she/ze/they would be exposed to a world of Nazis, anti-2SLGBTQQIA+ hate, Faucism-denial, Trump-support, etc., right? Here’s what was hidden:

How were these replies “offensive”?

Here’s the full thread, started by Dr. Karen (profile says pronouns “he/him” and “Vote Blue”):

Check out the full beard that this physician chooses to wear. How could he possibly achieve any kind of effective seal with a mask, be it cloth, surgical, N95, or N190 (my personal choice: double N95)?

Full post, including comments

Gender studies graduates at Twitter fact-check the Harvard Medical School professor

Here’s a great example of gender studies graduates at Twitter practicing California-style Science. A statement is scientifically false if it “goes against CDC guidelines” (as it happens, Professor Kulldorff’s March 2021 heresy of recommending COVID-19 shots primarily for older folks is today the official policy of the Danish government, informed by MD/PhDs).

What was the result of the censorship? The journalist explains in a tweet later in the thread:

After Twitter took action, Kulldorff’s tweet was slapped with a “Misleading” label and all replies and likes were shut off, throttling the tweet’s ability to be seen and shared by many people, the ostensible core function of the platform:

In my review of internal files, I found countless instances of tweets labeled as “misleading” or taken down entirely, sometimes triggering account suspensions, simply because they veered from CDC guidance or differed from establishment views.

Separately, my rage against long-form argument in a Twitter thread is somewhat reduced because I’ve found the “read like a book” icon at the top of the screen. Here’s how it then renders:

Full post, including comments

Matt Taibbi’s story about Twitter’s suppression of the Hunter Biden laptop story

Have folks tried to follow and understand the story about Twitter’s pre-election-2020 suppression of the Hunter Biden laptop story and other material that politicians asked them to deep-six?

I think there is supposed to be a narrative in here, but it is so chopped up by the presentation as individual tweets that it is tough to follow.

Has Matt Taibbi mostly proved that Twitter needs a substantial re-thinking to be suitable for long-form text? (I think tweets should be allowed at any length up to the standard relational database CLOB (character large object) limit of 2 billion characters, but a reader sees only a short summary (that long-form authors are forced to craft) until he/she/ze/they clicks “more”)

Readers: Have you figured out whether there is anything of interest in this reveal of internal Twitter machinations?

Update: In the official NYT version of history, Twitter’s shaping of what viewpoints people could express (or send to each other in private messages) never happened. The front page of the NYT time has space to talk about “notable diversity” of the U.S. World Cup team, but there is nothing about the Twitter files reveal. (Separately, I dispute that the US team is diverse. There are no gender ID requirements for World Cup players and yet for some reason players of only one gender ID have been selected.)

(Joe Biden’s granddaughter also does not exist according to the NYT. A search for plaintiff “Lunden” Roberts or granddaughter “Navy Joan” yields no results on nytimes.com.)

Full post, including comments

Twitter is 99.99 percent hate-free

“Hate Speech’s Rise on Twitter Is Unprecedented, Researchers Find” (NYT, today):

Before Elon Musk bought Twitter, slurs against Black Americans showed up on the social media service an average of 1,282 times a day. After the billionaire became Twitter’s owner, they jumped to 3,876 times a day.

Slurs against gay men appeared on Twitter 2,506 times a day on average before Mr. Musk took over. Afterward, their use rose to 3,964 times a day.

And antisemitic posts referring to Jews or Judaism soared more than 61 percent in the two weeks after Mr. Musk acquired the site.

This sounds plausible. After all, even Jews plugged into the ruling political party (#2 Joe Biden campaign donor after George Soros) were forced to take refuge in the Bahamas.

Yet, as the social media censors would say, this story is “MISSING CONTEXT”, i.e., that there are more than 500 million tweets per day. A few thousand that contain “hate words” (possibly used by those quoting or complaining about perceived hate) is a tiny percentage. The cited study actually shows that Twitter is more than 99.99% hate-free!

Here’s an example of a “slur against gay men” that came up when I searched for “fag”. This meets the criterion described in the NYT, but the tweet is not hate speech.

Here’s another example in which the author is not expressing a negative opinion about members of the 2SLGBTQQIA+ community:

And from a user with a profile of “white, trans, bi, she/her, … Marxist-Leninist-Maoist”:

Separately, who went to a high school in the hate-free pre-Trump days? My memory is that the word “fag” was used more than 3,964 times per day just by the 2,000 students at our high school. We were under the benevolent leadership of Jimmy Carter (folks who were nostalgic for the Carter administration got a replay with Biden after all, at least of the main feature: inflation) and it was in mostly-Democrat Bethesda, Maryland where even the Republicans believed that bigger government was generally better (since everyone in that company town worked for the government in some way).

Speaking of hate, here’s Pfizer spewing hatred for people who pay taxes that flow straight into its corporate pockets, i.e., the 50 percent of Americans who are without a Ph.D. and who yet want to ask a few questions before getting injecting with experimental drugs:

Note that the Deplorables cannot comment on the post!

Full post, including comments

Who will join me on Mastodon?

I can no longer tolerate the environment of tolerance for a wide range of speech that Elon Musk has established on Twitter. It is time to leave for Mastodon. Who’s with me?

It looks as though there is already a “Philip” using the service. So my username will be @PhilipTheSecond@Of.Mastodon

Related:

Full post, including comments

Anyone on Twitter can be canceled by Twitter, Apple, Google, or the European Union?

FastCompany says that Apple and Google will kill Twitter by removing the app from their respective monopoly stores if they see anything on Twitter that they don’t like.

Twitter, obviously, will remove anything and anyone that the company deems objectionable. Historically this was people who did not follow the Democratic Party line, e.g., the New York Post for the Hunter Biden laptop story.

“Twitter must comply with Europe’s platform rules, EU digital chief warns Musk in virtual meeting” (CNN Business):

… the social media platform must take significant steps to comply with EU content moderation laws, …

Twitter has “huge work ahead” to meet its obligations under the Digital Services Act, Europe’s new platform regulation, said Thierry Breton, the EU’s digital chief, in a readout of his meeting with Musk.

“Twitter will have to implement transparent user policies, significantly reinforce content moderation and protect freedom of speech, tackle disinformation with resolve, and limit targeted advertising,” Breton said…

I would love to learn about this law! To “protect freedom of speech”, it is necessary for a service to prevent anyone from speaking in a way that the European Union bureaucrats don’t like (“reinforce content moderation”)? And who decides what is “disinformation” that violates EU law? Sticking with the Hunter Biden laptop story, above, all of the Washington, D.C. expert insiders said that it was disinformation. “Hunter Biden story is Russian disinfo, dozens of former intel officials say” (Politico, just before the 2020 election on 10/19/2020):

More than 50 former senior intelligence officials have signed on to a letter outlining their belief that the recent disclosure of emails allegedly belonging to Joe Biden’s son “has all the classic earmarks of a Russian information operation.”

The letter, signed on Monday, centers around a batch of documents released by the New York Post last week that purport to tie the Democratic nominee to his son Hunter’s business dealings.

While the letter’s signatories presented no new evidence, they said their national security experience had made them “deeply suspicious that the Russian government played a significant role in this case” and cited several elements of the story that suggested the Kremlin’s hand at work.

While there has been no immediate indication of Russian involvement in the release of emails the Post obtained, its general thrust mirrors a narrative that U.S. intelligence agencies have described as part of an active Russian disinformation effort aimed at denigrating Biden’s candidacy.

Facebook hasn’t faced the same scrutiny, perhaps because they are still censoring in accordance with the ruling politicians’ wishes. But they could also be taken down by the smartphone duopoly or the EU.

In summary, there are now at least four filters through which content must pass before it can be distributed via the only practical modern means of reaching a substantial number of people. Folks in China might have more practical freedom of speech because there are only two filters: the operator of a service and the government.

This is an interesting illustration of how the early Internet nerds’ predictions turned out to be 100 percent wrong. None of them would have imagined a world in which there was no practical way to sell a book if a single bookstore (Amazon) didn’t like it and in which multiple bureaucracies exercised veto power over the online existence of any individual user and his or her (there was just two gender IDs back then) speech.

My question is why the same standards aren’t applied to web sites and email. Google and Apple can program their browsers to reject attempts to visit web sites that contain disinformation, e.g., that COVID vaccines do not prevent infection and transmission. Or at least augment web pages with context, as Twitter and Facebook already do. Google and Apple also control email systems. Why allow Deplorables to share misinformation and disinformation via email? The idea is that the companies, out of the goodness of their corporate hearts, will ensure online safety when on Facebook and Twitter, but will make zero attempt to prevent people from being led astray when reading email? How does that make sense? At a minimum, shouldn’t Gmail add context? If a personal friend gets an email from Rochelle Walensky about becoming sick with COVID a month after receiving the bivalent booster, Gmail could display “MISSING CONTEXT. The latest bivalent COVID boosters have been proven to protect against all SARS-CoV-2 variants. Visit cdc.gov for more information about COVID.” If someone is reading about how New York State is #1 in the nation in percent of residents’ income taxed away to fund state and local government, Safari could add a banner “Visit www.governor.ny.gov to learn how New York provides abortion care and protects you from gun violence with the taxes that you pay.”

Sculpture from the Louvre below titled “EU online safety expert deplatforms Nick Fuentes.”

Full post, including comments

When will the Ministry of Truth get to work?

It has been a month since Elon Musk closed his acquisition of Twitter. Changes are happening at a fast pace, potentially proving SR-71 pilot Paul Crickmore’s point “You’ve never been lost until you’ve been lost at Mach 3.”

According to President Biden, this is a platform for lies. FoxBusiness:

President Biden mentioned Elon Musk and Twitter during a fundraiser Friday night …

“Elon Musk goes out and buys an outlet that sends — that spews lies all across the world,” Biden remarked. “There are no editors anymore in America. There are no editors.”

The president seemed to suggest there was no longer any moderation on Twitter, a claim echoed by several others on the platform who are also critical of Musk.

“How do we expect kids to be able to understand what is at stake? What is at stake? So there’s a lot going on, a lot going on. But we have an enormous opportunity, enormous opportunity,” Biden added.

Biden’s comments soon after a White House tweet bragging about the Social Security cost-of-living increase was flagged by Twitter for lacking “context.” The White House later deleted the tweet.

If there are lies there must also be truth, right? Here are some lies that ultimately resulted in Marjorie Taylor Greene being “permanently suspended” from Twitter:

The Truth at the time (August 2021) was that vaccines do prevent infection and spread of COVID-19. As of November 5, 2022, the above tweet remained inaccessible via a search for “the fda should not approve the covid vaccines (from:mtgreenee)”. Twitter returned the following screen:

For those who sought to read something critical of the Federal government, in other words, Twitter reminded them to cease their thoughtcrime and read only official Federal government sites.

I guess it was fair to say that there was a Ministry of Truth operating within Twitter in August 2021 and even earlier this month.

But if Joe Biden and the rest of the U.S. government want Twitter to prevent lies from being published, does it makes sense for each social media platform to decide what is a lie and what is the truth? Advertisers are supposedly demanding that Twitter eliminate lies from its platform (yet the cash-seeking Democrat running against Ron DeSantis was allowed to lie with impunity; see Twitter won’t suspend a politician who lies to get money?), but how can they have confidence unless Twitter follows an official source of truth and flags anything that contradicts that source? We don’t want Facebook and Twitter to have different concepts of lie vs. truth, do we?

Related:

Full post, including comments

Did Twitter manage to fill its diversity quotas via the recent mass layoffs?

In a document written, I think, well before the Elon Musk takeover, Twitter wanted to hit the following quotas:

(Grammar police: Not “WHOM you work with?”)

They already had 46 percent “women”.

So they just needed to fire men and non-binary workers in order to achieve the 2025 dream in 2022. (Note that the quota of 50% for “women” necessarily meant that “men” would be less than half of the workforce because at least some of the workers would identify with the other 72 gender IDs recognized by Science.)

I wonder what Twitter meant for “At least a quarter of our US Tweepforce will be under-represented populations”. Would “over 60” be “under-represented”? How about 2SLGBTQQIA+? Members of the LGBTQIA+ were 13.5 percent of the workers before Black Friday:

If LGBTQIA+ qualifies as “under-represented”, Twitter could have filled its 25% quota simply by firing only workers who admitted to cisgender heterosexuality.

Who wants to guess what the diversity stats will look like the next time they’re released? (Or maybe part of Dark Elon’s plan was to fire the people who prepare these stats?)

Speaking of quotas, the United Nations is working toward a quota of 0% for the killing of journalists identifying as “women” (alternatively, a quota of 100% for the killing of journalists identifying with gender IDs other than “women”).

Full post, including comments

Did the Twitter search programmers deserve to be fired?

Elon Musk fired a lot of people from Twitter today (CNN). But the advertisers are leaving even faster, now that there is the potential for users to utter misinformation, e.g., that COVID-19 “vaccines” don’t prevent infection and transmission.

Appeasement doesn’t work, apparently. Government-funded General Motors and Pfizer, for example, will no longer advertise on a platform that allows people to say things that are contrary to what the Biden administration wants residents of the U.S. to hear.

Let’s look at the overall quality of the product that the pre-rightsizing crew developed. Suppose that we are searching to see if Kanye West has been unpersoned. A search for “kanye west” does not yield Kanye West’s profile:

How about “@kanyewest” as a search string?

What if we type “kanye west twitter profile” into the Google?

There is an account and it exactly matches one of our search strings (“@kanyewest”), but Twitter couldn’t find it! Maybe this is because Google hired everyone competent? Bing also shows the @kanyewest account as the number 1 result. So does Duck Duck Go.

How soon do we think we can start seeing the kind of big improvements in Twitter that a good product manager could drive, e.g., “If you type ‘Kanye West’ it has to return Kanye West’s account as the first result.”

How did Elon Musk and team zero in on the low-productivity workers? You’d think it would take a few months for new management to figure out which programmers on the search team were the good ones.

Also, when do we get a high-quality remake of the following:

Elon Musk can play himself.

Some lines:

  • Put that kombucha down! Kombucha’s for accepted check-ins only.
  • The good news is you’re fired. The bad news is that all you’ve got is just one week to regain your jobs. Starting with tonight’s git. .. First prize: a Tesla S. Second prize: a set of pronouns. Third prize is you’re fired.
  • f*ckin’ f*ggots -> f*ckin MAGAs
  • ABC = Always Be Censoring
  • I made $970,000 last year -> I lost more than $100 billion this year.

Update, November 5:

Full post, including comments