The Weekly Flail- August 31, 2024
OVERVIEW
The Trump fascist party is disintegrating in the wake of the Harris-Walz Democratic ticket launch. Neither fascists nor criminal organizations, in general, can tolerate being mocked and laughed at. I note this from the international perspective of The Flail, unlike the national media, which is still struggling to ignore the absurdity of the criminals, couch humpers, and brainworm sufferers surrounding Trump.
While celebrating Harris and Walz, people may have noticed that Harris’s hyped interview revealed a firm commitment to zero changes beyond a return to normalcy with humor. With growing climate disasters and direct support for the Zionist genocide in both Gaza and the West Bank, this is not normalcy that anyone wants, but that does not seem to be a problem in the 2024 election.
We have a growing list of potential pandemic viruses, but pandemic is a word, similar to genocide, that does not exist in the Harris lexicon.
Is there an AI disaster looming with Model Collapse?
WAR
Palestinians Seek US Appeals Court Review of Biden Genocide Complicity Case
Common Dreams August 31, 2024
By Brett Wilkins / Common Dreams
Six weeks after a three-judge panel of the 9th U.S. Circuit Court of Appeals dismissed a lawsuit filed by Palestinians, Palestinian Americans, and rights groups accusing senior Biden administration officials of complicity in Israel’s Gaza genocide, plaintiffs in the case on Thursday asked the full federal appellate court to revisit their suit.
The plaintiffs’ petition—which was filed by the Center for Constitutional Rights (CCR) and the law firm Van Der Hout LLP—requests an en banc rehearing of their case, in which U.S. President Joe Biden, Secretary of State Antony Blinken, and Defense Secretary Lloyd Austin are defendants. To win such a rehearing, a case must involve a matter of “exceptional importance” or be inconsistent with previous court rulings.
“With unconditional U.S. support, Israel has killed about 40,000 Palestinians, injured more than 90,000, forcibly displaced 2 million, and pushed large segments of Gaza into famine,” CCR said in a statement. “Israel’s actions, which followed numerous expressions of eliminationist intent by its leaders, have led many legal experts and scholars to conclude that it is committing genocide, the most serious human rights crime.”
“With unconditional U.S. support, Israel has killed about 40,000 Palestinians, injured more than 90,000, forcibly displaced 2 million, and pushed large segments of Gaza into famine.”
. . . .
Ukraine war: after Telegram founder arrest, Russians fear loss of ‘main information source’
The app, used by both the Kremlin and its opponents, is a vital source of uncensored news in a country where dissent is increasingly suppressed
France’s arrest of Telegram chief Pavel Durov has raised fears in Russia that the popular messaging app – used both by the Kremlin and its opponents – could be blocked, depriving them of one of the last sources of critical, uncensored news.
Since the start of its offensive in Ukraine in February 2022, Russia has cracked down on dissent and protest, leaving Russians without independent news outlets or access to Western social media such as Facebook, Instagram and X.
In that climate, Telegram – which was itself blocked for a period by the Kremlin for refusing to cooperate with Russian law enforcement agencies – has become one of the last bastions of free speech and uncensored information.
Moscow now fears for the fate of the messenger and its Russian-born founder, Durov, charged in late August with failing to curb extremist and illegal content on the platform.
Though he has been released on bail, he cannot leave the country and the Kremlin has warned France against turning the case against him “into political persecution.”
Durov’s arrest is not the only headache the privately-owned service faces.
The European Commission is also investigating whether Telegram has more EU users than claimed and must therefore comply with more stringent rules.
CLIMATE
Downslope
The Honest Sorcerer — August 26, 2024
The fossil fuel bonanza is slowly coming to an end. Not with a bang, mind you, but with a whimper: presenting us with a long winding road back to a much simpler life. If you happen to live in a region consuming half of the world’s oil supply, though, the way downhill will be one hell of a ride… And if you think we can innovate ourselves out of this hot mess, then think again and buckle up instead.
One does not need to be an oracle to see: if you have a finite resource, and you can no longer find it in quantities sufficient to meet demand, you will sooner or later run out of it. Before that happens, though, the rate of extraction will reach a peak, and as oil fields give up the ghost one by one, a ‘long descent’ will begin. Not that any of this should be big news: there were plenty of warnings from M. King Hubbert to Charles A.S. Hall and Jean Laherrère, or from the Limits to Growth study to the countless updates published to it, all saying that sooner or later the party will end. Our corporate overlords (hand-in-hand with the many autocrats of the world) however, spared no effort to deny that we could ever run low on oil—and we believed them, because we wanted to. The fireworks must never end, so we could keep driving our (ever bigger and heavier) cars to (ever larger) supermarkets and buy just about any product we wished to have.
Physics and geology rarely gives a hoot what we wish to have, though. Energy is everything, and now we have to face the prospect of a net decline in what’s available (that is: left to be used for purposes other than drilling more wells to replace depleted ones, or mining minerals to build power plants, dams or “renewables”). And while oil production is still growing in terms of the amount of crude oil and condensate brought to the surface, it still hasn’t reached its all time high in 2018. This production growth after the pandemic slump is even more staggering if you add in all sorts of other “liquids” like LNG, ethane, biofuels and what not—resulting in a number beyond a hundred million barrels per day—but what about net energy? (1)
When it comes to expanding drilling and pumping ever further the gains does not come from opening the taps a little wider. Each and every barrel added to the mix comes from ever deeper wells, from ever deeper seas, ever further away from the shore. As more and more CO2 and water must be pumped underground to force just a little more juice onto the surface, or ever more wells need to be fracked open (then re-fracked), the energy cost per barrel increases exponentially. To illustrate this process, compare a “simple” pumpjack consuming a modest amount of fuel to lift oil out of the ground, with a floating platform 24 stories high, weighing over 17000 tons, and with a deck the size of 15 basketball courts. Or how about Chevron’s latest invention of a new high-pressure extraction technology (announced last week) deployed at a deepwater well in the Gulf of Mexico? I wonder what the energy demand of these beasts are…
“Solving” this net energy dilemma is not a question of ingenuity, mind you. Adding more complexity and more technology always comes with an increased energy demand. As low cost drilling techniques fail to keep up with the depletion of easy to get oil, and will have to be replaced by ever more energy intensive methods, the situation can be expected to worsen even if we just try to maintain a steady supply of fuel. The question, whether we surpass the peak of November 2018 or not, will thus become moot. The aggregate net energy from oil (available for other uses) will most likely start to shrink after 2025 —no matter what we do. This is going to be one major event, a true turning point not only for western nations, but to the human enterprise as a whole. Combined with a looming peak in aggregate crude oil and condensate production (scheduled to arrive by 2030) it will be no longer possible to pretend that we have enough fuel to do everything we want. Actually, we would have to contend with less and less fuel production year after year.
. . . .
Mysterious Oropouche virus is spreading: what you should know
Mariana Lenharo August 26, 2024
Update: The US Centers for Disease Control and Prevention said 27 August that more than 20 US travelers returning from Cuba have been identified as having Oropouche virus.
Once confined to the Amazon region, the mysterious insect-borne virus that causes Oropouche fever has been expanding its range since late 2023, raising international concern. There have been more than 8,000 confirmed human infections in the Americas so far this year, most of them in Brazil, but Peru, Bolivia, Colombia and Cuba have also been affected.
In July, authorities in Brazil reported the deaths of two adults from the disease — the first fatalities recorded since the virus was identified almost 70 years ago. Brazilian officials are also investigating cases of fetal deaths and malformations that might have been caused by the virus, which investigations have shown can spread from a pregnant person to the fetus1. There are no vaccines or treatments for the disease.
Earlier this month, the Pan American Health Organization upgraded its risk level for Oropouche from moderate to high, citing the virus’s geographical spread and the occurrence of fatal cases, which are notable for a disease that has historically been known to cause mild to moderate symptoms. On 23 August, the World Health Organization published a note stating that the public-health risk posed by the virus is high at the regional level and low at the global level. The US Centers for Disease Control and Prevention has advised close surveillance of people returning from affected areas. Cases of Oropouche infection have been identified in people who have travelled to the United States, Spain, Italy and Germany from Brazil and Cuba.
What is ‘model collapse’? An expert explains the rumours about an impending AI doom
Aaron J. Snoswell August 18, 2024
Artificial intelligence (AI) prophets and newsmongers are forecasting the end of the generative AI hype, with talk of an impending catastrophic “model collapse”.
But how realistic are these predictions? And what is model collapse anyway?
Discussed in 2023, but popularised more recently, “model collapse” refers to a hypothetical scenario where future AI systems get progressively dumber due to the increase of AI-generated data on the internet.
The need for data
Modern AI systems are built using machine learning. Programmers set up the underlying mathematical structure, but the actual “intelligence” comes from training the system to mimic patterns in data.
But not just any data. The current crop of generative AI systems needs high quality data, and lots of it.
To source this data, big tech companies such as OpenAI, Google, Meta and Nvidia continually scour the internet, scooping up terabytes of content to feed the machines. But since the advent of widely available and useful generative AI systems in 2022, people are increasingly uploading and sharing content that is made, in part or whole, by AI.
In 2023, researchers started wondering if they could get away with only relying on AI-created data for training, instead of human-generated data.
There are huge incentives to make this work. In addition to proliferating on the internet, AI-made content is much cheaper than human data to source. It also isn’t ethically and legally questionable to collect en masse.
However, researchers found that without high-quality human data, AI systems trained on AI-made data get dumber and dumber as each model learns from the previous one. It’s like a digital version of the problem of inbreeding.
This “regurgitive training” seems to lead to a reduction in the quality and diversity of model behaviour. Quality here roughly means some combination of being helpful, harmless and honest. Diversity refers to the variation in responses, and which people’s cultural and social perspectives are represented in the AI outputs.
In short: by using AI systems so much, we could be polluting the very data source we need to make them useful in the first place.
Avoiding collapse
Can’t big tech just filter out AI-generated content? Not really. Tech companies already spend a lot of time and money cleaning and filtering the data they scrape, with one industry insider recently sharing they sometimes discard as much as 90% of the data they initially collect for training models.
These efforts might get more demanding as the need to specifically remove AI-generated content increases. But more importantly, in the long term it will actually get harder and harder to distinguish AI content. This will make the filtering and removal of synthetic data a game of diminishing (financial) returns.
Ultimately, the research so far shows we just can’t completely do away with human data. After all, it’s where the “I” in AI is coming from.
Are we headed for a catastrophe?
There are hints developers are already having to work harder to source high-quality data. For instance, the documentation accompanying the GPT-4 release credited an unprecedented number of staff involved in the data-related parts of the project.
We may also be running out of new human data. Some estimates say the pool of human-generated text data might be tapped out as soon as 2026.
It’s likely why OpenAI and others are racing to shore up exclusive partnerships with industry behemoths such as Shutterstock, Associated Press and NewsCorp. They own large proprietary collections of human data that aren’t readily available on the public internet.
However, the prospects of catastrophic model collapse might be overstated.
Until next week . . .