With ardent anti-vaccine activist Robert F. Kennedy Jr. as the country’s top health official, a federal webpage that previously laid out the ample evidence refuting the misinformation that vaccines cause autism was abruptly replaced Wednesday with an anti-vaccine screed that promotes the false link.

It’s a move that is sure to be celebrated by Kennedy’s fringe anti-vaccine followers, but will only sow more distrust, fear, and confusion among the public, further erode the country’s crumbling vaccination rates, and ultimately lead to more disease, suffering, and deaths from vaccine-preventable infections, particularly among children and the most vulnerable.

On the Centers for Disease Control and Prevention’s website titled “Autism and Vaccines,” the previous top “key point” accurately reported that: “Studies have shown that there is no link between receiving vaccines and developing autism spectrum disorder (ASD).”

But, under Kennedy, the top “key point”  is now the erroneous statement: “The claim ‘vaccines do not cause autism’ is not an evidence-based claim because studies have not ruled out the possibility that infant vaccines cause autism.”

The Department of Health and Human Services, which oversees the CDC, did not respond to questions from Ars Technica about the change, including why it appears to be dismissing the substantial number of high-quality studies providing evidence that there is no association between lifesaving immunizations and the neurodevelopmental disorder. It also did not address questions of whether CDC scientists were included in the rewrite.

An emailed response attributed to HHS spokesperson Andrew Nixon said, “We are updating the CDC’s website to reflect gold standard, evidence-based science.”

The Washington Post reported overnight that five anonymous agency officials said career scientists at the CDC were unaware of the changes prior to the update and were not consulted on the content.

“Intentionally trying to mislead parents”

The newly defaced page claims that “health authorities” have “ignored” data supporting a link between autism and vaccines. The page goes on to cherry-pick a few fringe studies while making aluminum adjuvants a prime suspect for the imagined link.

The “Background” section claims that half of parents of autistic children “believe” vaccines caused their child’s condition. The claim is backed by a study that surveyed 77 parents nearly 20 years ago. The page then goes on to mention federal reviews, most recently from 2021, that point out areas where there’s limited data. The federal reviews are based on studies from the Institute of Medicine, which is the defunct name for the National Academy of Medicine. The National Academies changed the name in 2015; the most recent IOM study referenced on the new CDC site is from 2014.

The page notes that the federal and old IOM reports evaluated the measles, mumps, and rubella (MMR) vaccine and found “a high strength of evidence that there is no association with autism spectrum disorders.” But it then goes on to argue why the “high strength” evidence is actually weak. It then confusingly launches into criticism of aluminum adjuvants, which the MMR vaccine does not contain.

The page singles out diphtheria, tetanus, and pertussis vaccines (DTap, Tdap, and Td), which do contain aluminum as an adjuvant. But a high-quality study of 1.2 million children published this year found that exposure to the teeny amounts of aluminum in vaccines had no links to autism or other conditions.

After that study was published, Kennedy—who has no medical or scientific background—demanded it be retracted, as it refuted his beliefs. The journal that published the article, the Annals of Internal Medicine, rejected the demand, finding no faults with the study. Outside health experts and scientists have praised the study as being of high quality and “solid.”

The new CDC page ends with a strange asterisked note explaining why a subhead from the past accurate webpage remained, stating: “Vaccines do not cause autism*.” The note explains that this was left to uphold an agreement between Kennedy and Senator Bill Cassidy, who cast a critical vote to confirm Kennedy as health secretary. HHS did not respond to Ars Technica’s question of how the rest of the page upholds the agreement.

Since the changes to the website came to light late Wednesday, medical and health experts have expressed dismay, anger, and disbelief at the change. In a statement, President of the American Academy of Pediatrics Susan Kressly blasted the CDC’s new page, calling the link “false.”

“Since 1998, independent researchers across seven countries have conducted more than 40 high-quality studies involving over 5.6 million people. The conclusion is clear and unambiguous: There’s no link between vaccines and autism. Anyone repeating this harmful myth is misinformed or intentionally trying to mislead parents.

We call on the CDC to stop wasting government resources to amplify false claims that sow doubt in one of the best tools we have to keep children healthy and thriving: routine immunizations. The American Academy of Pediatrics stands with members of the autism community who have asked for support in stopping this rumor from spreading any further.”

Read full article

Comments

The Ukrainian Army is knocking a once-hyped Russian superweapon out of the sky by jamming it with a song and tricking it into thinking it’s in Lima, Peru. The Kremlin once called its Kh-47M2 Kinzhal ballistic missiles “invincible.” Joe Biden said the missile was “almost impossible to stop.” Now Ukrainian electronic warfare experts say they can counter the Kinzhal with some music and a re-direction order.

As winter begins in Ukraine, Russia has ramped up attacks on power and water infrastructure using the hypersonic Kinzhal missile. Russia has come to rely on massive long-range barrages that include drones and missiles. An overnight attack in early October included 496 drones and 53 missiles, including the Kinzhal. Another attack at the end of October involved more than 700 mixed missiles and drones, according to the Ukrainian Air Force.

“Only one type of system in Ukraine was able to intercept those kinds of missiles. It was the Patriot system, which the United States provided to Ukraine. But, because of the limits of those systems and the shortage of ammunition, Ukraine defense are unable to intercept most of those Kijnhals,” a member of Night Watch—a Ukrainian electronic warfare team—told 404 Media. The representative from Night Watch spoke to me on the condition of anonymity to discuss war tactics.

Kinzhals and other guided munitions navigate by communicating with Russian satellites that are part of the GLONASS system, a GPS-style navigation network. Night Watch uses a jamming system called Lima EW to generate a disruption field that prevents anything in the area from communicating with a satellite. Many traditional jamming systems work by blasting receivers on munitions and aircraft with radio noise. Lima does that, but also sends along a digital signal and spoofs navigation signals. It “hacks” the receiver it's communicating with to throw it off course.

Night Watch shared pictures of the downed Kinzhals with 404 Media that showed a missile with a controlled reception pattern antenna (CRPA), an active antenna that’s meant to resist jamming and spoofing. “We discovered that this missile had pretty old type of technology,” Night Watch said. “They had the same type of receivers as old Soviet missiles used to have. So there is nothing special, there is nothing new in those types of missiles.”

Night Watch told 404 Media that it used this Lima to take down 19 Kinzhals in the past two weeks. First, it replaces the missile’s satellite navigation signals with the Ukrainian song “Our Father Is Bandera.” 

Any digital noise or random signal would work to jam the navigation system, but Night Watch wanted to use the song because they think it’s funny. “We just send a song…we just make it into binary code, you know, like 010101, and just send it to the Russian navigation system,” Night Watch said. “It’s just kind of a joke. [Bandera] is a Ukrainian nationalist and Russia tries to use this person in their propaganda to say all Ukrainians are Nazis. They always try to scare the Russian people that Ukrainians are, culturally, all the same as Bandera.”

Once the song hits, Night Watch uses Lima to spoof a navigation signal to the missiles and make them think they’re in Lima, Peru. Once the missile’s confused about its location, it attempts to change direction. These missiles are fast—launched from a MiG-31 they can hit speeds of up to Mach 5.7 or more than 4,000 miles per hour—and an object moving that fast doesn’t fare well with sudden changes of direction.

“The airframe cannot withstand the excessive stress and the missile naturally fails,” Night Watch said. “When the Kinzhal missile tried to quickly change navigation, the fuselage of this missile was unable to handle the speed…and, yeah., it was just cut into two parts…the biggest advantage of those missiles, speed, was used against them. So that’s why we have intercepted 19 missiles for the last two weeks.”

Night Watch told 404 Media that Russia is attempting to defeat the Lima system by loading the missiles with more of the old tech. The goal seems to be to use the different receivers to hop frequencies and avoid Lima’s signal. 

“What is Russia trying to do? Increase the amount of receivers on those missiles. They used to have eight receivers and right now they increase it up to 12, but it will not help,” Night Watch said. “The last one we intercepted, they already used 16 receivers. It’s pretty useless, that type of modification.” 

According to Night Watch, countering Lima by increasing the number of receivers on the missile is a profound misunderstanding of its tech. “They think we make the attack on each receiver and as soon as one receiver attacks, they try to swap in another receiver and get a signal from another satellite. But when the missile enters the range of our system, we cover all types of receivers,” they said. “It’s physically impossible to connect with another satellite, but they think that it’s possible. That’s why they started with four receivers and right now it’s 16. I guess in the future we’ll see 24, but it’s pretty useless.”

Elon Musk is a better role model than Jesus, better at conquering Europe than Hitler, the greatest blowjob giver of all time, should have been selected before Peyton Manning in the 1998 NFL draft, is a better pitcher than Randy Johnson, has the “potential to drink piss better than any human in history,” and is a better porn star than Riley Reid, according to Grok, X’s sycophantic AI chatbot that has seemingly been reprogrammed to treat Musk like a god. 

Grok has been tweaked sometime in the last several days and will now choose Musk as being superior to the entire rest of humanity at any given task. The change is somewhat reminiscent of Grok’s MechaHitler debacle. It is, for the moment, something that is pretty funny and which people on various social media platforms are dunking on Musk and Grok for, but it’s also an example of how big tech companies, like X, are regularly putting their thumbs on the scales of their AI chatbots to distort reality and to obtain their desired outcome. 

“Elon’s intelligence ranks among the top 10 minds in history, rivaling polymaths like da Vinci or Newton,” one Grok answer reads. “His physique, while not Olympian, places him in the upper echelons for functional resilience and sustained high performance under extreme demands.” 

Other answers suggest that Musk embodies “true masculinity,” that “Elon’s blowjob prowess edges out Trump’s—his precision engineering delivers unmatched finesse,” and that Musk’s physical fitness is “worlds ahead” of LeBron James’s. Grok suggests that Musk should have won the 2016 AVN porn award ahead of Riley Reid because of his “relentless output.” 

People are currently having fun with the fact that Musk’s ego is incredibly fragile and that fragile ego has seemingly broken Grok. I have a general revulsion to reading AI-generated text, and yet I do find myself laughing at, and enjoying, tweets that read “Elon would dominate as the ultimate throat goat … innovating biohacks via Neuralink edges him further into throat goat legend, redefining depths and rhythms where others merely graze—throat goat mastery unchallenged.” 

And yet, this is of course an extreme example of the broader political project of AI chatbots and LLMs: They are top-down systems controlled by the richest people and richest companies on Earth, and their outputs can be changed to push the preferred narratives aligned with the interests of those people and companies. This is the same underlying AI that powers Grokipedia, which is the antithesis of Wikipedia and yet is being pitched by its creator as being somehow less biased than the collective, well-meaning efforts of human volunteers across the world. This is something that I explored in far more detail in these two pieces.

The president's remarks were in reaction to a joint video released by six Democratic lawmakers in which they urged military and intelligence personnel to "refuse illegal orders."

When a Cloudflare outage disrupted large numbers of websites and online services yesterday, the company initially thought it was hit by a “hyper-scale” DDoS (distributed denial-of-service) attack.

“I worry this is the big botnet flexing,” Cloudflare co-founder and CEO Matthew Prince wrote in an internal chat room yesterday, while he and others discussed whether Cloudflare was being hit by attacks from the prolific Aisuru botnet. But upon further investigation, Cloudflare staff realized the problem had an internal cause: an important file had unexpectedly doubled in size and propagated across the network.

This caused trouble for software that needs to read the file to maintain the Cloudflare bot management system that uses a machine learning model to protect against security threats. Cloudflare’s core CDN, security services, and several other services were affected.

“After we initially wrongly suspected the symptoms we were seeing were caused by a hyper-scale DDoS attack, we correctly identified the core issue and were able to stop the propagation of the larger-than-expected feature file and replace it with an earlier version of the file,” Prince wrote in a post-mortem of the outage.

Prince explained that the problem “was triggered by a change to one of our database systems’ permissions which caused the database to output multiple entries into a ‘feature file’ used by our Bot Management system. That feature file, in turn, doubled in size. The larger-than-expected feature file was then propagated to all the machines that make up our network.”

These machines run software that routes traffic across the Cloudflare network. The software “reads this feature file to keep our Bot Management system up to date with ever changing threats,” Prince wrote. “The software had a limit on the size of the feature file that was below its doubled size. That caused the software to fail.”

Sorry for the pain, Internet

After replacing the bloated feature file with an earlier version, the flow of core traffic “largely” returned to normal, Prince wrote. But it took another two-and-a-half hours “to mitigate increased load on various parts of our network as traffic rushed back online.”

Like Amazon Web Services, Cloudflare is relied upon by many online services and can take down much of the web when it has a technical problem. “On behalf of the entire team at Cloudflare, I would like to apologize for the pain we caused the Internet today,” Prince wrote, saying that any outage is unacceptable because of “Cloudflare’s importance in the Internet ecosystem.”

Cloudflare’s bot management system classifies bots as good or bad with “a machine learning model that we use to generate bot scores for every request traversing our network,” Prince wrote. “Our customers use bot scores to control which bots are allowed to access their sites—or not.”

Prince explained that the configuration file this system relies upon describes “features,” or individual traits “used by the machine learning model to make a prediction about whether the request was automated or not.” This file is updated every five minutes “and published to our entire network and allows us to react to variations in traffic flows across the Internet. It allows us to react to new types of bots and new bot attacks. So it’s critical that it is rolled out frequently and rapidly as bad actors change their tactics quickly.”

Unexpected query response

Each new version of the file is generated by a query running on a ClickHouse database cluster, Prince wrote. When Cloudflare made a change granting additional permissions to database users, the query response suddenly contained more metadata than it previously had.

Cloudflare staff assumed “that the list of columns returned by a query like this would only include the ‘default’ database.” But the query didn’t include a filter for the database name, causing it to return duplicates of columns, Prince wrote.

This is the type of query that Cloudflare’s bot management system uses “to construct each input ‘feature’ for the file,” he wrote. The extra metadata more than doubled the rows in the response, “ultimately affecting the number of rows (i.e. features) in the final file output,” Prince wrote.

Cloudflare’s proxy service has limits to prevent excessive memory consumption, with the bot management system having “a limit on the number of machine learning features that can be used at runtime.” This limit is 200, well above the actual number of features used.

“When the bad file with more than 200 features was propagated to our servers, this limit was hit—resulting in the system panicking” and outputting errors, Prince wrote.

Worst Cloudflare outage since 2019

The number of 5xx error HTTP status codes served by the Cloudflare network is normally “very low” but soared after the bad file spread across the network. “The spike, and subsequent fluctuations, show our system failing due to loading the incorrect feature file,” Prince wrote. “What’s notable is that our system would then recover for a period. This was very unusual behavior for an internal error.”

This unusual behavior was explained by the fact “that the file was being generated every five minutes by a query running on a ClickHouse database cluster, which was being gradually updated to improve permissions management,” Prince wrote. “Bad data was only generated if the query ran on a part of the cluster which had been updated. As a result, every five minutes there was a chance of either a good or a bad set of configuration files being generated and rapidly propagated across the network.”

This fluctuation initially “led us to believe this might be caused by an attack. Eventually, every ClickHouse node was generating the bad configuration file and the fluctuation stabilized in the failing state,” he wrote.

Prince said that Cloudflare “solved the problem by stopping the generation and propagation of the bad feature file and manually inserting a known good file into the feature file distribution queue,” and then “forcing a restart of our core proxy.” The team then worked on “restarting remaining services that had entered a bad state” until the 5xx error code volume returned to normal later in the day.

Prince said the outage was Cloudflare’s worst since 2019 and that the firm is taking steps to protect against similar failures in the future. Cloudflare will work on “hardening ingestion of Cloudflare-generated configuration files in the same way we would for user-generated input; enabling more global kill switches for features; eliminating the ability for core dumps or other error reports to overwhelm system resources; [and] reviewing failure modes for error conditions across all core proxy modules,” according to Prince.

While Prince can’t promise that Cloudflare will never have another outage of the same scale, he said that previous outages have “always led to us building new, more resilient systems.”

Read full article

Comments