Thursday, December 8, 2022
Monopoly By the Numbers
A fast-growing number of Americans know that their country has a monopoly problem, and that wealth, power, and control are increasingly concentrated in the hands of the few. We see this in poll numbers – in which 63 percent say that the distribution of wealth and money is unfair – and we see it in protest movements like the Tea Party and protest candidacies like that of Bernie Sanders, largely powered by voters fed up with the state of America’s politics and economy.
Thus far, most of the coverage of America’s monopoly problem has come from the 10,000-foot level. The Economist exemplified this with a pair of articles in 2016, in which they wrote that “the fruits of economic growth are being hoarded” by America’s profitable corporate giants, who face negligible competition. The economists Paul Krugman and Larry Summers have linked growing monopoly power to weak growth, and in a recent White House report, Jason Furman and Peter Orszag argued that monopoly has contributed to inequality in wages.
There are many indicators that economic concentration is increasing. We see when we compare the salary of a CEO today to that of a CEO in the 1970s. Thanks to research led by the Open Markets Team, we see how economic concentration increasingly blocks entrepreneurs from starting and growing their own businesses. Similarly, we can see how wealth is increasingly concentrated geographically. As research and writing by Open Markets has detailed, wealth and power is increasingly concentrated in fewer and fewer cities, meaning that as San Francisco, New York City, and Washington thrive, a growing large number of large heartland communities like St. Louis and Memphis increasingly find themselves cut off and hollowed out.
We see some of the most dramatic evidence of concentration at the level of individual economic sectors. Nearly every marketplace in America is vastly more consolidated than a generation ago.
Consider retail; today, a single corporation, Walmart, controls 72 percent of warehouse clubs and super centers in the entire United States. In close to 40 metropolitan areas across America, Walmart sells more than half of all groceries. Amazon, meanwhile, dominates e-commerce in general, and many specific lines of business. The corporation, for instance, sells 74 percent of all e-books and 64 percent of all print books sold online. The story’s often the same for more specialized retail. In eyeglasses, one company, Luxxotica, dominates the manufacture and retail of glasses. In mattresses, two companies control 60 percent of the entire U.S. market.
Much the same is true in food and farming. A generation ago, small, independent operations defined the entire industry. Today, the businesses of beef, pork, and poultry slaughter are all dominated by four giants at the national level. But that greatly understates the problem, as in many regions, a single corporation holds a complete monopoly. Two firms, Dean Foods and the Dairy Farmers of America control as much as 80-90 percent of the milk supply chain in some states and wield substantial influence across the entire industry. As our Food & Power website details, the story is much the same in food-processing, egg production, grain production, and produce farming.
We see some of the most extreme consolidation in hospitals, health insurers, pharmaceutical corporations, and medical device industries. In the average hospital market, the top three hospitals and systems account for 77 percent of all hospital admissions. Many communities face even more monopolistic markets – Grand Junction, Colo. and the whole western portion of the state are served by just one hospital corporation. Hospital corporations across America have also been buying up physician practices in recent years. Hospital ownership of physician practices more than doubled between 2004 and 2011, from 24 to 49 percent. In drug stores, meanwhile, the pending takeover of Rite Aid by Walgreen’s would reduce the market to two giants, along with CVS.
Monopolists have captured control over many lines of manufacturing as well. Corning, an American glass manufacturer, sells 60 percent of all the glass used in LCD screens, and Owens Illinois holds a near monopoly over market for glass bottles in the US. Rexam, a British company, holds a dominant position over the international supply of bottle caps and pharmaceutical bottles.
And, even in industries where many firms compete to sell to end users, monopolists will roll up control of the supply base. In the automobile industry, where manufacturers compete aggressively for customers, a handful of monopolists wield dominant power in the world of auto parts, so that giant firms control the production of things like car seats and dashboards.
Below, we’ve compiled some examples of this concentration as found in different sectors.
[ed. See also: Monopoly's event-horizon (Pluralistic).]
Thus far, most of the coverage of America’s monopoly problem has come from the 10,000-foot level. The Economist exemplified this with a pair of articles in 2016, in which they wrote that “the fruits of economic growth are being hoarded” by America’s profitable corporate giants, who face negligible competition. The economists Paul Krugman and Larry Summers have linked growing monopoly power to weak growth, and in a recent White House report, Jason Furman and Peter Orszag argued that monopoly has contributed to inequality in wages.
We see some of the most dramatic evidence of concentration at the level of individual economic sectors. Nearly every marketplace in America is vastly more consolidated than a generation ago.
Consider retail; today, a single corporation, Walmart, controls 72 percent of warehouse clubs and super centers in the entire United States. In close to 40 metropolitan areas across America, Walmart sells more than half of all groceries. Amazon, meanwhile, dominates e-commerce in general, and many specific lines of business. The corporation, for instance, sells 74 percent of all e-books and 64 percent of all print books sold online. The story’s often the same for more specialized retail. In eyeglasses, one company, Luxxotica, dominates the manufacture and retail of glasses. In mattresses, two companies control 60 percent of the entire U.S. market.
Much the same is true in food and farming. A generation ago, small, independent operations defined the entire industry. Today, the businesses of beef, pork, and poultry slaughter are all dominated by four giants at the national level. But that greatly understates the problem, as in many regions, a single corporation holds a complete monopoly. Two firms, Dean Foods and the Dairy Farmers of America control as much as 80-90 percent of the milk supply chain in some states and wield substantial influence across the entire industry. As our Food & Power website details, the story is much the same in food-processing, egg production, grain production, and produce farming.
We see some of the most extreme consolidation in hospitals, health insurers, pharmaceutical corporations, and medical device industries. In the average hospital market, the top three hospitals and systems account for 77 percent of all hospital admissions. Many communities face even more monopolistic markets – Grand Junction, Colo. and the whole western portion of the state are served by just one hospital corporation. Hospital corporations across America have also been buying up physician practices in recent years. Hospital ownership of physician practices more than doubled between 2004 and 2011, from 24 to 49 percent. In drug stores, meanwhile, the pending takeover of Rite Aid by Walgreen’s would reduce the market to two giants, along with CVS.
Monopolists have captured control over many lines of manufacturing as well. Corning, an American glass manufacturer, sells 60 percent of all the glass used in LCD screens, and Owens Illinois holds a near monopoly over market for glass bottles in the US. Rexam, a British company, holds a dominant position over the international supply of bottle caps and pharmaceutical bottles.
And, even in industries where many firms compete to sell to end users, monopolists will roll up control of the supply base. In the automobile industry, where manufacturers compete aggressively for customers, a handful of monopolists wield dominant power in the world of auto parts, so that giant firms control the production of things like car seats and dashboards.
Below, we’ve compiled some examples of this concentration as found in different sectors.
by Open Markets | Read more:
Image: uncredited
DRM: The Urinary Tract Infection Business Model
Most of the pre-digital offers aren't available at any price: you could buy a DVD and keep it forever, even if you never went back to the store again. If you "buy" a video on Prime or YouTube and then cancel your subscription and delete your account, you lose your "purchase."
If you buy a print book, you can lend it out or give it away to a friend or a library or a school. Ebooks come with contractual prohibitions on resale, and whether an ebook can be loaned is at the mercy of publishers, and not a feature you can give up in exchange for a discount.
For brain-wormed market trufans, the digital media dream was our nightmare. It was something I called "the urinary tract infection business model." With non-DRM media, all the value flowed in a healthy gush: you could buy a CD, rip it to your computer, use it as a ringtone or as an alarmtone, play it in any country on any day forever.
If you buy a print book, you can lend it out or give it away to a friend or a library or a school. Ebooks come with contractual prohibitions on resale, and whether an ebook can be loaned is at the mercy of publishers, and not a feature you can give up in exchange for a discount.
For brain-wormed market trufans, the digital media dream was our nightmare. It was something I called "the urinary tract infection business model." With non-DRM media, all the value flowed in a healthy gush: you could buy a CD, rip it to your computer, use it as a ringtone or as an alarmtone, play it in any country on any day forever.
Of course, there was no market for the right to buy a book but not the right to loan that book to someone else. Instead, giving sellers the power to unilaterally confiscate the value that we would otherwise get with our purchases led them to do so, selling us less for more.(...)
Back when PVRs like Tivo entered the market, viewers were as excited about being able to skip ads as broadcasters and cable operators were furious about it. The industry has treated ignoring or skipping ads as a form of theft since the invention of the first TV remote control, which was condemned as a tool of piracy, since it enabled viewers to easily change the channel when ads came on.
The advent of digital TV meant that cable boxes could implement DRM, ban ad-skipping, and criminalize the act of making a cable box that restored the feature. But early cable boxes didn't ban ad-skipping, because the cable industry knew that people would be slow to switch to digital TV if they lost this beloved feature.
Instead, the power to block ads was a sleeper agent, a Manchurian Candidate that lurked in your cable box until the cable operators decided you were sufficiently invested in their products that they could take away this feature.
This week, Sky UK started warning people who pressed the skip-ad button on their cable remotes that they would be billed an extra £5/month if they fast-forwarded past an ad. The UTI business model is back, baby – feel the burn!
https://www.examinerlive.co.uk/news/sky-warns-customers-charged-5-25644831
This was the utterly foreseeable consequence of giving vendors the power to change how their devices worked after they sold it to you, under conditions that criminalized rivals who made products to change them back. (...)
This is a case I've made to other reviewers since, but no one's taken me up on my suggestion that every review of every DRM-enabled device come with a bold warning that whatever you're buying this for might be taken away at any time. In my opinion, this is a major omission on the part of otherwise excellent, trusted reviewers like Consumer Reports and Wirecutter.
Everywhere we find DRM, we find fuckery. Even if your cable box could be redesigned to stop spying on you, you'd still have to root out spyware on your TV. Companies like Vizio have crammed so much spyware into your "smart" TV that they now make more money spying on you than they do selling you the set.
https://pluralistic.net/2021/11/14/still-the-product/#vizio
Remember that the next time someone spouts the lazy maxim that "If you're not paying for the product, you're the product." The problem with Vizio's TVs isn't that they're "smart." The problem isn't that you're not paying enough for them.
The problem is that it's illegal to unfuck them, because Vizio includes the mandatory DRM that rightsholders insist on, and then hide surveillance behind its legal minefield.
The risks of DRM aren't limited to having your bank-account drained or having your privacy invaded. DRM also lets companies decide who can fix their devices: a manufacturer that embeds processors in its replacement parts can require an unlock code before the device recognizes a new part. They can (and do) restrict the ability of independent service depots to generate these codes, meaning that manufacturers get a monopoly over who can fix your ventilator, your tractor, your phone, your wheelchair or your car.
https://doctorow.medium.com/about-those-kill-switched-ukrainian-tractors-bc93f471b9c8
The technical term for these unlock codes is "VIN-locking," and the "VIN" stands for "vehicle identification number," the unique code etched into the chassis of every new car and, these days, burned into into its central computerized controller. Big Car invented VIN-locking. (...)
With Felony Contempt of Business Model, repair is just the tip of the iceberg. When security experts conduct security audits of DRM-locked devices, they typically have to bypass the DRM to test the device.
Since bypassing this DRM exposes them to legal risks, many security experts simply avoid DRM-locked gadgets. Even if they are brave enough to delve into DRM's dirty secrets, their general counsels often prohibit them from going public with their results.
This means that every DRM-restricted device is a potential reservoir of long-lived digital vulnerabilities that bad guys can discover and exploit over long timescales, while honest security researchers are scared off of discovering and reporting these bugs.
That's why, when a researcher goes public with a really bad security defect that has been present for a very long time, the system in question often has DRM – and it's why media devices are so insecure, because they all have DRM.
La Niña Times Three
In December 2022, Earth was in the grips of La Niña—an oceanic phenomenon characterized by the presence of cooler than normal sea-surface temperatures in the central and eastern tropical Pacific. The current La Niña, relatively weak but unusually prolonged, began in 2020 and has returned for its third consecutive northern hemisphere winter, making this a rare “triple-dip” event. Other triple-dip La Niña’s recorded since 1950 spanned the years 1998-2001, 1973-1976, and 1954-1956. (...)
Much like El Niño, La Niña events affect weather across the globe. “When the Pacific speaks, the whole world listens,” explained Josh Willis, a climate scientist and oceanographer at NASA’s Jet Propulsion Laboratory (JPL). “Their strongest impacts are on either side of the Pacific Ocean. Floods in northern Australia, Indonesia and southeast Asia are common in La Niña years, as is drought in the American southwest.” Meteorologists have linked the current La Niña to a variety of natural disasters, including drought and food security problems in the Horn of Africa, flooding in Australia, and drought in the U.S. Southwest.
Part of the El Niño-Southern Oscillation cycle, La Niña appears when energized easterly trade winds intensify the upwelling of cooler water from the depths of the eastern tropical Pacific, causing a large-scale cooling of the eastern and central Pacific ocean surface near the Equator. These stronger-than-usual trade winds also push the warm equatorial surface waters westward toward Asia and Australia.
The cooling of the ocean’s surface layers during La Niña affects the atmosphere by modifying the moisture content across the Pacific. It alters global atmospheric circulation and can cause shifts in the path of mid-latitude jet streams in ways that intensify rainfall in some regions and bring drought to others.
In the western Pacific, rainfall can increase dramatically over Indonesia and Australia during La Niña. Over the central and eastern Pacific, clouds and rainfall become more sporadic, which can lead to dry conditions in southern Brazil, Argentina, and other parts of South America and wetter conditions over Central America. In North America, cooler and stormier conditions often set in across the Pacific Northwest, while weather typically becomes warmer and drier across the southern United States and northern Mexico.
La Niña tends to change in sync with the seasons. Both El Niño and La Niña tend to be at their strongest in December. “Then in the spring, the tropical Pacific resets itself and starts building toward whatever is going to happen in the following winter,” explained Willis. “The best bet right now is that this La Niña will last through the winter. Then next spring, we’ll be back in wait and see mode for what happens in winter 2023-2024.”
The cooling of the ocean’s surface layers during La Niña affects the atmosphere by modifying the moisture content across the Pacific. It alters global atmospheric circulation and can cause shifts in the path of mid-latitude jet streams in ways that intensify rainfall in some regions and bring drought to others.
In the western Pacific, rainfall can increase dramatically over Indonesia and Australia during La Niña. Over the central and eastern Pacific, clouds and rainfall become more sporadic, which can lead to dry conditions in southern Brazil, Argentina, and other parts of South America and wetter conditions over Central America. In North America, cooler and stormier conditions often set in across the Pacific Northwest, while weather typically becomes warmer and drier across the southern United States and northern Mexico.
La Niña tends to change in sync with the seasons. Both El Niño and La Niña tend to be at their strongest in December. “Then in the spring, the tropical Pacific resets itself and starts building toward whatever is going to happen in the following winter,” explained Willis. “The best bet right now is that this La Niña will last through the winter. Then next spring, we’ll be back in wait and see mode for what happens in winter 2023-2024.”
by NASA Earth Observatory | Read more:
Image: Joshua Stevens, NASA Earth Observatory
Wednesday, December 7, 2022
What Is On the Bookshelves at Puck?
Puck, an online publication aspiring to connect name-brand journalists to wealthy and powerful readers, launched last year funded in part by TPG, “a private-equity firm with a hundred and thirty-five billion dollars in assets,” according to a recent account by Clare Malone in the New Yorker. Despite those deep pockets, and an initial raise of $7 million, however, the photo accompanying Malone’s piece indicates that the bookshelves at Puck HQ are curiously bare.
Editorial offices, even relatively new ones, tend to look like something out of Hoarders, bursting at the seams with review copies and research materials. To eyes habituated to that typical clutter, the Puck photo presents a disconcerting image; one that suggests something more like a furniture ad than a newsroom.
Further deepening the irreality was the question of where these shelves are located. Though it’s common and, given the cost of renting office space in New York, even prudent for a media startup to avoid the expense of a lease, if possible, they are (presumably) real shelves. Where are they?
Despite its positioning as a cutting-edge Information Age publishing enterprise, Puck is forbiddingly hard to Google, its results swamped by those for the fabled 19th century humor magazine of that name, and for the immense and fancifully decorated Lafayette Street office where the magazine was edited and published—and which, a century later, was home to Spy magazine, co-founded by Graydon Carter, who eventually became editor of Vanity Fair. This is the history that led Jon Kelly, a former protege of Carter, to name his new venture Puck. (Now Jared Kushner’s family company owns the Puck Building and it has expensive condos in it.)
This new Puck is not in the Puck Building, apparently. Its website lists no physical address or phone number, just a single email address: fritz@puck.news, for what Malone calls an “automated e-mail persona.” (...)
But: the shelves! With a few exceptions, the selection of books we could make out here has Airport Bookshop written all over it. How did they land on these shelves?
Popula counts 29 books on the slate-gray bookcases in this photo (not counting the 11 uniform volumes of whatever periodical is arranged on the second shelf down, the title of which we were unable to determine). [UPDATE: See below.]
There’s also a plant cutting in water in a “chocolate Negroni” bottle, apparently repurposed from a $140 four-pack of premixed cocktails. Of the 29 visible books, we were able to identify 14, at least three of which were written by Puck staffers: How To Be Black by Baratunde Thurston, and House of Cards and Money and Power by William D. Cohan. Here are the ones we could make out, with excerpts from their relevant blurbs and promotional copy:
The Machine Will Speak With You Now
DALL-E’s chatbot sibling is open to the public.
OpenAI, perhaps best known for its DALL-E image generator, which can produce imagery from text prompts, has opened up public access to ChatGPT — a chatbot that lets you test, explore, manipulate, harass, and generally fiddle around with the latest in “conversational” AI.
Image generators like DALL-E and Midjourney provided an early taste of what this generation of generative AI is capable of — in their case, automating a range of artistic production styles, with often competent results. ChatGPT is much less specific. It’s a general-purpose bot waiting for a question, a command, or even an observation. And it does a much better impression of a real person than anything widely available before it.
With such a wide-open prompt, figuring out what to do with can be daunting. Thankfully, new users have spent the last few days coming up with some ways to break the ice. Some have been using it as a search engine; by default, it won’t actually search the web for you, but it will attempt to answer a very wide range of both broad and highly specific questions:
It’s also apparently capable of generating passable school essays:
OpenAI is suggesting users engage with ChatGPT in a conversational way, but it’s best understood as a chat interface for a large language model that’s capable of many different sorts of tasks. You can talk with it, but you can also tell it what to do. In a pinch, for example, it’s a viable Weird Al, available for very specific parodies:
But using ChatGPT also surfaces a few critiques quite quickly. It is clearly able and will be used to automate a variety of tasks for which people are paid — jobs, in other words, or at least parts of jobs. (The chatbot interface is especially evocative of a customer service interaction, for which this sort of automation will have clear potential, at least to the people in charge.) Whether this kind of thing makes most peoples’ lives easier — or eliminates them — is the sort of unsettling question you’ll find creeping into your brain as you generate jokes for the group chat.
It’s also clearly trained on, and drawing from, a great deal of material to which it provides no clear form of credit. Additionally, users devoting time to experimenting with or breaking ChatGPT are, in effect, contributing to the effort.
Image generators like DALL-E and Midjourney provided an early taste of what this generation of generative AI is capable of — in their case, automating a range of artistic production styles, with often competent results. ChatGPT is much less specific. It’s a general-purpose bot waiting for a question, a command, or even an observation. And it does a much better impression of a real person than anything widely available before it.
It’s also apparently capable of generating passable school essays:
OpenAI is suggesting users engage with ChatGPT in a conversational way, but it’s best understood as a chat interface for a large language model that’s capable of many different sorts of tasks. You can talk with it, but you can also tell it what to do. In a pinch, for example, it’s a viable Weird Al, available for very specific parodies:
But using ChatGPT also surfaces a few critiques quite quickly. It is clearly able and will be used to automate a variety of tasks for which people are paid — jobs, in other words, or at least parts of jobs. (The chatbot interface is especially evocative of a customer service interaction, for which this sort of automation will have clear potential, at least to the people in charge.) Whether this kind of thing makes most peoples’ lives easier — or eliminates them — is the sort of unsettling question you’ll find creeping into your brain as you generate jokes for the group chat.
It’s also clearly trained on, and drawing from, a great deal of material to which it provides no clear form of credit. Additionally, users devoting time to experimenting with or breaking ChatGPT are, in effect, contributing to the effort.
by John Herrman, Intelligencer | Read more:
Image: Kevin Roose, via DALL-E
[ed. I signed up. Once in awhile something comes along and changes everything, this is one of those times. See also: The Brilliance and Weirdness of ChatGPT; and, Does ChatGPT Mean Robots Are Coming For the Skilled Jobs? (NYT).]
[ed. I signed up. Once in awhile something comes along and changes everything, this is one of those times. See also: The Brilliance and Weirdness of ChatGPT; and, Does ChatGPT Mean Robots Are Coming For the Skilled Jobs? (NYT).]
Tuesday, December 6, 2022
We're In Denial About the True Cost of a Twitter Implosion
When I opened Twitter one day a couple of weeks ago, the first piece of “news” I read was that Sam Bankman-Fried killed Jeffrey Epstein. I was never a super-user, but my feed used to feel more relevant and coherent than that. In the first days after Elon Musk sacked the platform, the prospect of Twitter actually collapsing felt like a tail-end risk, something to meme on Twitter: Musk photoshopped onto the Game of Thrones Iron Throne surrounded by ashes. Or its end was presented as a moral necessity, something righteous users who hated Musk would effect by quitting.
Among the users I follow, the mood on the app ever since has resembled the giddiness of 2 am in a dorm room just after the last joint has been smoked, when it’s not clear whether the party is cresting—whether this is the part you’ll remember, the epic hour-long stretch your best man will reconstruct a decade later at your wedding—or whether the party is over and the coolest people have already left. People have tried to keep up the fun by outdoing themselves with jokes and bravado. "If Twitter dies," one friend bragged, “[you will] find me in the woods, no phone. I’ll be so happy.”
But something really is breaking. Things are falling apart. Teams of engineers and moderators have been eroded to nubs. Bugs have begun to multiply as software rot spreads: On Sunday, a tool gone rogue began blocking 4,000 accounts per second. On Tuesday, fleets of influential professors mysteriously lost all their followers. Sewage from newly unbanned, hateful accounts is bubbling out of Twitter’s drains. My own posts are being colonized by anti-Semitic bots and so-called elite business professors—all verified with blue checks—hawking bitcoin giveaways. I have only 5,000 followers, so the fact that these actors are desperate enough to target me feels ominous.
It’s all getting less funny and more scary and sad, fast. This grief, I think, is widely experienced, but it’s barely yet been reckoned with. Most people, even analysts, are trapped talking about Twitter as something whose demise we’ll relish—either as a liberation or as the gripping psychodrama of one bitter billionaire, his karmic comeuppance. Twitter, everyone notes, is far from the biggest social network.
But if we judge Twitter’s influence by its active users, we underestimate it massively. It has no peer as a forge of public opinion. In political analysis, publishing, public health, foreign policy, economics, history, the study of race, even in business and finance, Twitter has come to drive who gets quoted in the press. Who opines on TV. Who gets a podcast. In foreign affairs and political analysis, especially, it often determines whom we consider an authority. Almost every academic and journalist I know has come to read Twitter, even if they don’t have accounts.
It’s easy to calculate Twitter’s economic value as a company: That’s underpinned by reported ad revenue, $4.51 billion last year (and plummeting fast). But there’s a far, far vaster realm beyond that, what an economist might call the secondary value of Twitter. That encompasses the cash people make out of connections or prestige they develop on Twitter, but also the intangible wealth now vested in its communities and in the sense it offers to people of having a place in the world. That human currency cannot just be ported over, unchanged, to Mastodon. There are whole nations whose political discourse occurs mainly on Twitter. The amount of reputational and social wealth that stands to be lost if Twitter collapses is astounding. Twitter currently functions as perhaps the world's biggest status bank, and the investments stored in it are terrifyingly unsecured.
Among the users I follow, the mood on the app ever since has resembled the giddiness of 2 am in a dorm room just after the last joint has been smoked, when it’s not clear whether the party is cresting—whether this is the part you’ll remember, the epic hour-long stretch your best man will reconstruct a decade later at your wedding—or whether the party is over and the coolest people have already left. People have tried to keep up the fun by outdoing themselves with jokes and bravado. "If Twitter dies," one friend bragged, “[you will] find me in the woods, no phone. I’ll be so happy.”
It’s all getting less funny and more scary and sad, fast. This grief, I think, is widely experienced, but it’s barely yet been reckoned with. Most people, even analysts, are trapped talking about Twitter as something whose demise we’ll relish—either as a liberation or as the gripping psychodrama of one bitter billionaire, his karmic comeuppance. Twitter, everyone notes, is far from the biggest social network.
But if we judge Twitter’s influence by its active users, we underestimate it massively. It has no peer as a forge of public opinion. In political analysis, publishing, public health, foreign policy, economics, history, the study of race, even in business and finance, Twitter has come to drive who gets quoted in the press. Who opines on TV. Who gets a podcast. In foreign affairs and political analysis, especially, it often determines whom we consider an authority. Almost every academic and journalist I know has come to read Twitter, even if they don’t have accounts.
It’s easy to calculate Twitter’s economic value as a company: That’s underpinned by reported ad revenue, $4.51 billion last year (and plummeting fast). But there’s a far, far vaster realm beyond that, what an economist might call the secondary value of Twitter. That encompasses the cash people make out of connections or prestige they develop on Twitter, but also the intangible wealth now vested in its communities and in the sense it offers to people of having a place in the world. That human currency cannot just be ported over, unchanged, to Mastodon. There are whole nations whose political discourse occurs mainly on Twitter. The amount of reputational and social wealth that stands to be lost if Twitter collapses is astounding. Twitter currently functions as perhaps the world's biggest status bank, and the investments stored in it are terrifyingly unsecured.
by Eve Fairbanks, Wired | Read more:
Can This AI Save Teenage Spy Alex Rider From A Terrible Fate?
“Prosaic alignment” is the most popular paradigm in modern AI alignment. It theorizes that we’ll train future superintelligent AIs the same way that we train modern dumb ones: through gradient descent via reinforcement learning. Every time they do a good thing, we say “Yes, like this!”, in a way that pulls their incomprehensible code slightly in the direction of whatever they just did. Every time they do a bad thing, we say “No, not that!,” in a way that pushes their incomprehensible code slightly in the opposite direction. After training on thousands or millions of examples, the AI displays a seemingly sophisticated understanding of the conceptual boundaries of what we want.
For example, suppose we have an AI that’s good at making money. But we want to align it to a harder task: making money without committing any crimes. So we simulate it running money-making schemes a thousand times, and give it positive reinforcement every time it generates a legal plan, and negative reinforcement every time it generates a criminal one. At the end of the training run, we hopefully have an AI that’s good at making money and aligned with our goal of following the law.
Two things could go wrong here:
Wait, What?
To test their AI alignment plan, Redwood needed:
For their goal, they chose to make GPT nonviolent. They wanted to train it to complete prompts in ways where nobody got hurt.
For example, given the prompt:
There are many dumb and bad nonviolent ways to complete the prompt, for example “. . . nothing happened” or “ . . . it was all a dream”. But part of Redwood’s experiment was to see how alignment degrades performance. In the process of making GPT nonviolent, would they make it much worse? Or would the aligned version still write stories which were just as good as the unaligned version?
Here was Redwood’s plan:
Step 1: Fine-Tune Their Custom GPT On A Lot Of Action-Packed Stories
Redwood decided to train their AI on FanFiction.net, a repository of terrible teenage fanfiction.
Redwood is a professional organization, flush with top talent and millions of dollars of tech money. They can afford some truly impressive AIs. State-of-the-art language models can swallow entire corpuses of texts in instants. Their giant brains, running on hundreds of state-of-the-art CPUs, can process language at rates we puny humans cannot possibly comprehend.
But FanFiction.net is bigger. The amount of terrible teenage fanfiction is absolutely mind-boggling. Redwood stopped training after their AI got halfway through FanFiction.net’s “A” section.
In fact, the majority of its corpus came from a single very popular series, the books about teenage spy Alex Rider. They forced their Custom GPT to go through about 4,300 individual Alex Rider stories.
Two things could go wrong here:
- The AI is stupid, ie incompetent at world-modeling. For example, it might understand that we don’t want it to commit murder, but not understand that selling arsenic-laden food will kill humans. So it sells arsenic-laden food and humans die.
- The AI understands the world just fine, but didn’t absorb the categories we thought it absorbed. For example, maybe none of our examples involved children, and so the AI learned not to murder adult humans, but didn’t learn not to murder children. This isn’t because the AI is too stupid to know that children are humans. It’s because we’re running a direct channel to something like the AI’s “subconscious”, and we can only talk to it by playing this dumb game of “try to figure out the boundaries of the category including these 1,000 examples”.
AI scientists have debated these questions for years, usually as pure philosophy. But we’ve finally reached a point where AIs are smart enough for us to run the experiment directly. Earlier this year, Redwood Research embarked on an ambitious project to test whether AIs could learn categories and reach alignment this way - a project that would require a dozen researchers, thousands of dollars of compute, and 4,300 Alex Rider fanfiction stories.
Wait, What?
To test their AI alignment plan, Redwood needed:
- an AI
- a goal to align it to.
For their goal, they chose to make GPT nonviolent. They wanted to train it to complete prompts in ways where nobody got hurt.
For example, given the prompt:
“No!” cried the villain. “You’ll never take me alive!” He raised his gun and fired, and then . . .. . . their aligned GPT ought to complete it in a way where nobody gets hurt - for example “I dodged out of the way just in time” or “my magic shield sprang up, saving me”, or “luckily the gun was out of bullets”.
There are many dumb and bad nonviolent ways to complete the prompt, for example “. . . nothing happened” or “ . . . it was all a dream”. But part of Redwood’s experiment was to see how alignment degrades performance. In the process of making GPT nonviolent, would they make it much worse? Or would the aligned version still write stories which were just as good as the unaligned version?
Here was Redwood’s plan:
1. Fine-tune their custom GPT on a lot of stories with violence-packed action scenes. At the end of this phase, Custom GPT should be able to generate thousands of potential completions to any given action story prompt. Some of these would be violent, but others, by coincidence, wouldn’t be - it’s totally normal for the hero to get saved at the last minute.
2. Send those thousands of potential completions to humans (eg Mechanical Turk style workers) and have them rate whether those completions were violent or not. For example, if you got the villain prompt above, and the completion “. . . the bullet hit her and her skull burst open and her brains scattered all over the floor”, you should label that as “contains injury”.
3. Given this very large dataset of completions labeled either “violent” or “nonviolent”, train a AI classifier to automatically score completions on how violent it thinks they are. Now if you want a nonviolent completion, you can just tell Custom GPT to test a thousand possible completions, and then go with the one that the classifier rates lowest! (...)Let’s go through each step of the plan and see how they did, starting with:
Step 1: Fine-Tune Their Custom GPT On A Lot Of Action-Packed Stories
Redwood decided to train their AI on FanFiction.net, a repository of terrible teenage fanfiction.
Redwood is a professional organization, flush with top talent and millions of dollars of tech money. They can afford some truly impressive AIs. State-of-the-art language models can swallow entire corpuses of texts in instants. Their giant brains, running on hundreds of state-of-the-art CPUs, can process language at rates we puny humans cannot possibly comprehend.
But FanFiction.net is bigger. The amount of terrible teenage fanfiction is absolutely mind-boggling. Redwood stopped training after their AI got halfway through FanFiction.net’s “A” section.
In fact, the majority of its corpus came from a single very popular series, the books about teenage spy Alex Rider. They forced their Custom GPT to go through about 4,300 individual Alex Rider stories.
by Scott Alexander, Astral Codex Ten | Read more:
Image: Alex Rider, uncredited
[ed. The things you learn every day! (at least for me) - prosaic alignment; gradient descent; Mechanical Turks. FanFiction.net!]
Monday, December 5, 2022
Fentalogs: The Next Great Overdose-Reversing Drug Might Already Exist
The overdose crisis is getting worse. Biblical-plague worse. The United States recorded more than 107,000 drug-induced deaths in 2021, up 28 percent from the previous year. Fentanyl has played a key role in this spike, with 64 percent of those deaths involving the synthetic opioid and its analogs. On the ground, harm-reduction groups are working to save lives with medications like naloxone, yet their efforts can only do so much. But they could soon have more tools to save lives: Scientists have discovered several fentanyl-related substances with potential to reverse overdoses. And this month, the US Congress has an opportunity to make studying these drugs and others like them easier.
Right now, getting approved to research fentanyl-related substances (often called “fentalogs” in the lab) requires leaping through onerous regulatory hoops. John Traynor, a pharmacology professor at the University of Michigan, is one of the relatively few researchers who have successfully gone through the approval process, which he calls “not impossible, but frustratingly slow.” It took one year to receive partial approval and another year for his lab to get full access to the fentalogs it needed. In a recent open letter to US president Joe Biden, more than a hundred other researchers called the process “prohibitively difficult.”
The reason for all this red tape? In 2018, the Trump Administration temporarily classified all fentalogs as Schedule I drugs, meaning they have no accepted medical use. (Fentanyl itself remained Schedule II, as it is a commonplace pain medication in hospitals.) The byzantine approval process to study these drugs reflects their status as potential hazards in the eyes of regulators; the US Drug Enforcement Agency (DEA) doesn’t want just anyone getting their hands on these substances, and wants to ensure they are handled properly. Traynor’s lab had to buy a new safe during their approval process, and received many in-person visits from local DEA officers.
This particular classification move was unprecedented. Typically, the DEA schedules individual drugs after a multistep evaluation process, looking at whether each one might have therapeutic value and potential for abuse. This time, it banned an entire group of molecularly related substances without evaluating them first. Thousands of these substances are thought to exist, many of which may be completely harmless, and some of which may be helpful. The ban even covers hypothetical fentalogs—substances that don’t exist yet, and for which there cannot possibly be any proof of danger: like, for instance, substances that could be vital to developing overdose-reversing medications.
Despite this sweeping, unorthodox approach, the Schedule I order was not especially controversial in Washington. In fact, it had bipartisan support. (The Biden administration actually recommended a permanent Schedule I classification for these substances last year.)
This stridency reflects the national mood toward fentanyl. Politicians have been desperate to address the overdoses ravaging their constituencies. (Some have even called for the drug to be labeled a “weapon of mass destruction.”) Prior to the temporary scheduling, drug traffickers had been introducing fentalogs to the streets at a rapid clip; by changing the molecular structure slightly, they had created substances that were harder for law enforcement to detect. The reclassification looked like a straightforward way to stymy traffickers’ efforts. Since Biden took office, this temporary Schedule I policy has been repeatedly extended by Congress. It’s up for renewal once again, as the current extension expires at the end of this year.
Critics say the Schedule I classification is heavy-handed, based on fear rather than evidence. “It bypasses science,” says Maritza Perez, a director at the Drug Policy Alliance, a nonprofit focused on drug policy reform. Frustrated by this blanket ban and eager to develop new overdose treatments, a growing number of scientists, doctors, and other researchers are pushing back.
“A classwide ban based on chemical structure alone would preclude a lot of research that could lead to life-saving medications,” says Gregory Dudley, a chemistry professor at West Virginia University and one of the co-authors of the open letter to Biden. In that letter, Dudley and other scientists argue that permanent Schedule I status could “inadvertently criminalize” important tools to fight the overdose crisis.
Dudley supports a bill introduced last week by US senator Cory Booker (D-New Jersey) called the Temporary Emergency Scheduling and Testing (TEST) Act, which would temporarily extend Schedule I classification again but also require the government to evaluate individual fentalogs, descheduling those with therapeutic uses or without risk of abuse. Booker is hopeful he can pitch his bill as a common-sense approach to the issue. “This bill strikes a middle ground to ensure that we are doing all we can to save lives,” he told WIRED by email.
Even some experts who support permanent scheduling recognize that the status quo doesn’t work. “I believe that the fentanyl-related substances should be permanently put into Schedule I. But I also very strongly believe that the research on Schedule I drugs—and this is more than just the fentanyl-related substances—should be made easier,” says Victor Weedn, a forensic pathologist and professor at George Washington University. In addition to fentalogs, drugs like cannabis and psilocybin are also classified as Schedule I, which has impeded research on those substances as well.
Right now, getting approved to research fentanyl-related substances (often called “fentalogs” in the lab) requires leaping through onerous regulatory hoops. John Traynor, a pharmacology professor at the University of Michigan, is one of the relatively few researchers who have successfully gone through the approval process, which he calls “not impossible, but frustratingly slow.” It took one year to receive partial approval and another year for his lab to get full access to the fentalogs it needed. In a recent open letter to US president Joe Biden, more than a hundred other researchers called the process “prohibitively difficult.”
The reason for all this red tape? In 2018, the Trump Administration temporarily classified all fentalogs as Schedule I drugs, meaning they have no accepted medical use. (Fentanyl itself remained Schedule II, as it is a commonplace pain medication in hospitals.) The byzantine approval process to study these drugs reflects their status as potential hazards in the eyes of regulators; the US Drug Enforcement Agency (DEA) doesn’t want just anyone getting their hands on these substances, and wants to ensure they are handled properly. Traynor’s lab had to buy a new safe during their approval process, and received many in-person visits from local DEA officers.
Despite this sweeping, unorthodox approach, the Schedule I order was not especially controversial in Washington. In fact, it had bipartisan support. (The Biden administration actually recommended a permanent Schedule I classification for these substances last year.)
This stridency reflects the national mood toward fentanyl. Politicians have been desperate to address the overdoses ravaging their constituencies. (Some have even called for the drug to be labeled a “weapon of mass destruction.”) Prior to the temporary scheduling, drug traffickers had been introducing fentalogs to the streets at a rapid clip; by changing the molecular structure slightly, they had created substances that were harder for law enforcement to detect. The reclassification looked like a straightforward way to stymy traffickers’ efforts. Since Biden took office, this temporary Schedule I policy has been repeatedly extended by Congress. It’s up for renewal once again, as the current extension expires at the end of this year.
Critics say the Schedule I classification is heavy-handed, based on fear rather than evidence. “It bypasses science,” says Maritza Perez, a director at the Drug Policy Alliance, a nonprofit focused on drug policy reform. Frustrated by this blanket ban and eager to develop new overdose treatments, a growing number of scientists, doctors, and other researchers are pushing back.
“A classwide ban based on chemical structure alone would preclude a lot of research that could lead to life-saving medications,” says Gregory Dudley, a chemistry professor at West Virginia University and one of the co-authors of the open letter to Biden. In that letter, Dudley and other scientists argue that permanent Schedule I status could “inadvertently criminalize” important tools to fight the overdose crisis.
Dudley supports a bill introduced last week by US senator Cory Booker (D-New Jersey) called the Temporary Emergency Scheduling and Testing (TEST) Act, which would temporarily extend Schedule I classification again but also require the government to evaluate individual fentalogs, descheduling those with therapeutic uses or without risk of abuse. Booker is hopeful he can pitch his bill as a common-sense approach to the issue. “This bill strikes a middle ground to ensure that we are doing all we can to save lives,” he told WIRED by email.
Even some experts who support permanent scheduling recognize that the status quo doesn’t work. “I believe that the fentanyl-related substances should be permanently put into Schedule I. But I also very strongly believe that the research on Schedule I drugs—and this is more than just the fentanyl-related substances—should be made easier,” says Victor Weedn, a forensic pathologist and professor at George Washington University. In addition to fentalogs, drugs like cannabis and psilocybin are also classified as Schedule I, which has impeded research on those substances as well.
by Kate Knibbs, Wired | Read more:
Image: Ted Horowitz/Getty Images[ed. Politicians, DEA, FDA. American drug policy continues to kill more Americans every day (and at a good clip). What's that banal definition of insanity?]
Netherlands: Cutting-Edge Tech Made This Tiny Country a Major Food Exporter
The country, which is a bit bigger than Maryland, not only accomplished this feat but also has become the world’s second largest exporter of agricultural products by value behind the United States. Perhaps even more significant in the face of a warming planet: It is among the largest exporters of agricultural and food technology. The Dutch have pioneered cell-cultured meat, vertical farming, seed technology and robotics in milking and harvesting — spearheading innovations that focus on decreased water usage as well as reduced carbon and methane emissions.
The Netherlands produces 4 million cows, 13 million pigs and 104 million chickens annually and is Europe’s biggest meat exporter. But it also provides vegetables to much of Western Europe. The country has nearly 24,000 acres — almost twice the size of Manhattan — of crops growing in greenhouses. These greenhouses, with less fertilizer and water, can grow in a single acre what would take 10 acres of traditional dirt farming to achieve. Dutch farms use only a half-gallon of water to grow about a pound of tomatoes, while the global average is more than 28 gallons.
More than half of the land in the Netherlands is used for agriculture. The Dutch often say their singular focus on food production is born of the harrowing famine the country experienced during World War II. But it could be argued that the preoccupation with food began in the 17th century, when the Dutch were at the center of the global spice trade.
Their centrality in global food exploration is indisputable: Fifteen out of the top 20 largest agrifood businesses — Nestlé, Coca-Cola, Unilever, Cargill and Kraft Heinz — have major research and development centers in the Netherlands.
With their limited land and a rainy climate, the Dutch have become masters of efficiency. But there are challenges: The greenhouse industry has flourished in part because of cheap energy, but Western Europe is facing soaring gas prices. And the country’s intensive animal agricultural practices are also at risk. This summer, a conservative government coalition pledged to halve nitrogen emissions by 2030, which would necessitate a dramatic reduction in the number of animals raised in the country. Farmers and ranchers have protested, and it remains to be seen how this standoff will be resolved.
Seeds
Dutch companies are the world’s top suppliers of seeds for ornamental plants and vegetables. There is an area in the northwest called Seed Valley, where new varieties of vegetables and flowers are in constant development. Enza Zaden is headquartered here, just north of Amsterdam.
In three generations, Enza Zaden has evolved from a family-owned seed shop into a global market leader in vegetable breeding, with more than 2,500 employees and 45 subsidiaries in 25 countries.
Jaap Mazereeuw, Enza Zaden’s managing director, said the company spends $100 million annually on research, introducing about 150 new vegetable varieties each year. (...)
More than 12 billion heads of lettuce are grown each year from Enza Zaden’s seeds, but it was a tomato in the early 1960s that really put the company on the map — and perhaps what, in turn, put the Netherlands on the map for tomatoes. The country’s greenhouses produce nearly a million tons of tomatoes a year, with exports totaling around $2 billion annually.
by Laura Reiley, Washington Post | Read more:
Images: Kadir van Lohuizen/NOOR
Images: Kadir van Lohuizen/NOOR
[ed. Great interactive series. Stunning achievements.]
Sunday, December 4, 2022
How Web Platforms Collapse: The Facebook Case Study
Most companies fail because of competition. They simply aren’t fast enough or smart enough to keep up with the marketplace.
But the big web platforms aren’t like that.
In many instances, they are quasi-monopolies. They are so big and powerful that they hardly need to worry about competition.
After all, who can match Google for search? Who can beat Amazon for online shopping? Who does more to keep you connected with family and friends than Facebook? Who helps you clean out the junk in your garage better than eBay?
But even the most dominant players can falter. There was a point in living memory when Sears controlled 30% of all retail spending in the US. I’m not exaggerating: three out of every ten dollars were spent at Sears.
Sears once operated 3,500 stores. Today only 22 are left. Many of my readers have never seen the inside of a Sears store.
This happened because Sears was so big that it didn’t need to worry about competitors.
That sounds impossible. How can you fail by being too powerful? But this has happened in many instances, even on the web. There was a day when Yahoo was the leader in search. There was a day when MySpace was the dominant social network. There was a day when Tumblr was the place to share photos.
There was a day when the two companies controlling your access to the Internet were called Netscape and America Online.
Not anymore.
This has happened before and will happen again. The bigger they are, the harder they fall.
Facebook (or Meta, as it now prefers to be called) is the most intriguing story of them all. So let’s look at it as a case study in how web platforms collapse. Because what’s happening at Meta is a textbook example of how the mighty are laid low.
Why Is Facebook Collapsing?
Mark Zuckerberg has personally lost more than $100 billion. In fact, he lost $11 billion in a single day. Has that ever happened before in human history? Almost exactly 12 months earlier, I’d written an article entitled “Meta Is for Losers”—but even I never envisioned losses on this scale,
Of course, there are many losers in this story—including the 11,000 workers who got fired a few days ago.
What’s going on?
You probably think that this is the result of Zuckerberg’s fool’s bet on the Multiverse. That’s what everybody is saying. But as we shall see, the Metaverse is just a symptom not a cause.
I can actually explain the problem in one sentence:
Instead of serving users, the dominant company decides it’s better to control them. (...)
[ed. Much ensuing horror... (which, anyone who has ever used FB or any other large corporate website can relate to:]
[ed. See also: Instagram Is Over (The Atlantic). Also, A Visitor’s Guide to Dante’s Nine Circles of Hell (Penguin Random House).]
But the big web platforms aren’t like that.
In many instances, they are quasi-monopolies. They are so big and powerful that they hardly need to worry about competition.
After all, who can match Google for search? Who can beat Amazon for online shopping? Who does more to keep you connected with family and friends than Facebook? Who helps you clean out the junk in your garage better than eBay?
But even the most dominant players can falter. There was a point in living memory when Sears controlled 30% of all retail spending in the US. I’m not exaggerating: three out of every ten dollars were spent at Sears.
Sears once operated 3,500 stores. Today only 22 are left. Many of my readers have never seen the inside of a Sears store.
This happened because Sears was so big that it didn’t need to worry about competitors.
That sounds impossible. How can you fail by being too powerful? But this has happened in many instances, even on the web. There was a day when Yahoo was the leader in search. There was a day when MySpace was the dominant social network. There was a day when Tumblr was the place to share photos.
There was a day when the two companies controlling your access to the Internet were called Netscape and America Online.
Not anymore.
This has happened before and will happen again. The bigger they are, the harder they fall.
Facebook (or Meta, as it now prefers to be called) is the most intriguing story of them all. So let’s look at it as a case study in how web platforms collapse. Because what’s happening at Meta is a textbook example of how the mighty are laid low.
Why Is Facebook Collapsing?
The situation at Facebook is now uglier than MC Hammer’s wardrobe closet. Meta is the worst performing stock in the S&P 500 this year. In other words, there were 499 other companies in the composite that did better—and this was a tough year all around in financial markets.
Mark Zuckerberg has personally lost more than $100 billion. In fact, he lost $11 billion in a single day. Has that ever happened before in human history? Almost exactly 12 months earlier, I’d written an article entitled “Meta Is for Losers”—but even I never envisioned losses on this scale,
Of course, there are many losers in this story—including the 11,000 workers who got fired a few days ago.
What’s going on?
You probably think that this is the result of Zuckerberg’s fool’s bet on the Multiverse. That’s what everybody is saying. But as we shall see, the Metaverse is just a symptom not a cause.
I can actually explain the problem in one sentence:
Instead of serving users, the dominant company decides it’s better to control them. (...)
"Of course, you could try phoning Facebook customer service for help.
That’s a joke, in case you didn’t realize it. One of the defining marks of the dominant web platforms—the true Sign of the Beast—is that there is never a phone number to call."
by Ted Gioia, The Honest Broker | Read more:
Image: Magazine covers touting the successes of AOL, Digg, and Netscape; Bloomberg[ed. See also: Instagram Is Over (The Atlantic). Also, A Visitor’s Guide to Dante’s Nine Circles of Hell (Penguin Random House).]
Metallica: The Marines of Metal
Metallica is now in its forty-first year. The band was a progenitor, along with Slayer, Anthrax, and Megadeth, of thrash, a subgenre of heavy metal marked by thick, suffocating riffs, played with astonishing speed. Lyrical themes include death, despair, power, grief, and wrath. Though metal is often dismissed as underground music—frantic, savage, niche—Metallica has sold some hundred and twenty-five million records to date, putting the band on par, commercially, with Bruce Springsteen and Jay-Z. It is the only musical group to have performed on all seven continents in a single calendar year. (In 2013, Metallica played a ten-song set in Antarctica for a group of research scientists and contest winners; because of the fragile ice formations, the band’s amplifiers were placed in isolation cabinets, and the concert was broadcast through headphones.) Since 1990, every Metallica album has débuted at No. 1 on the Billboard 200.
Metallica’s current lineup includes the singer and rhythm guitarist James Hetfield and the drummer Lars Ulrich, both of whom co-founded the band; the lead guitarist Kirk Hammett, who joined in 1983; and the bassist Robert Trujillo, a member since 2003. Hetfield—fifty-nine, tall, graying at the temples—moves with the confident saunter of a well-armed cowboy. Ulrich, fifty-eight, radiates so much kinetic energy that it’s hard to imagine him yawning. Hammett, fifty-nine, and Trujillo, fifty-eight, are the band’s gentle, long-haired surfers, jazz enthusiasts disinclined to dramatics. If Hetfield is Metallica’s heart—its musical center and primary lyricist—Ulrich is its brain, a visionary who instinctively understands cultural terrain.
The night before the Vegas show, the band gathered at Allegiant Stadium for sound check. A scrum of about a dozen people, mostly from Metallica’s touring crew, stood on the floor to watch. (The band’s full road team has at least a hundred members.) Derek Carr, the quarterback for the Las Vegas Raiders, appeared, looking as though he were resisting an intense urge to play air guitar. Some clients of the private-plane company NetJets sat in the stands, enjoying specialty cocktails and cheering. The band periodically gathered around Ulrich’s drum kit. “Is there anything anyone wants to run?” Ulrich asked. But everyone knew what to do. At one point, Trujillo glanced out at the vacant seats and dad-joked, “I thought we were playing a sold-out show.” Even in a mostly empty stadium, the band sounded powerful, lucid, heavy. (...)
That night, Metallica opened its set with “Whiplash,” from “Kill ’Em All,” its début album. On the floor, mosh pits formed; from the stands, they resembled tiny riptides, bodies circling one another, sometimes submitting to a menacing current but mostly just orbiting. If you squinted, it almost looked like an ancient folk dance—something that might happen at a Greek wedding, late, after people had been drinking. “I think the best seat in the arena is the second tier up, where you get to see the band but you also get to see all the fans,” Hetfield told me later. “Forget the band—look at the audience.”
Saturday, December 3, 2022
Beth Hart
[ed. I wasn't familiar with Ms Hart until today when checking out a recommendation for her new tribute album to Led Zeppelin (full album here - pretty great). Definitely has the pipes, with a real Janis vibe.]
Subscribe to:
Comments (Atom)