The Single Most Important Thing

Is It Time for a Pause?

Many of the people building powerful AI systems think they’ll stumble on an AI system that forever changes our world fairly soon — three years, five years. I think they’re reasonably likely to be wrong about that, but I’m not sure they’re wrong about that. If we give them fifteen or twenty years, I start to suspect that they are entirely right.

And while I think that the enormous, terrifying challenges of making AI go well are very much solvable, it feels very possible, to me, that we won’t solve them in time.

It’s hard to overstate how much we have to gain from getting this right. It’s also hard to overstate how much we have to lose from getting it wrong. When I’m feeling optimistic about having grandchildren, I imagine that our grandchildren will look back in horror at how recklessly we endangered everyone in the world. And I’m much much more optimistic that humanity will figure this whole situation out in the end if we have twenty years than I am if we have five.

There’s all kinds of AI research being done — at labs, in academia, at nonprofits, and in a distributed fashion all across the internet — that’s so diffuse and varied that it would be hard to ‘slow down’ by fiat. But there’s one kind of AI research — training much larger, much more powerful language models — that it might make sense to try to slow down. If we could agree to hold off on training ever more powerful new models, we might buy more time to do AI alignment research on the models we have. This extra research could make it less likely that misaligned AI eventually seizes control from humans.

An open letter released on Wednesday, with signatures from Elon Musk, Apple co-founder Steve Wozniak, leading AI researcher Yoshua Bengio, and many other prominent figures, called for a six-month moratorium on training bigger, more dangerous ML models:

We call on all AI labs to immediately pause for at least 6 months the training of AI systems more powerful than GPT-4. This pause should be public and verifiable, and include all key actors. If such a pause cannot be enacted quickly, governments should step in and institute a moratorium.

I tend to think that we are developing and releasing AI systems much faster and much more carelessly than is in our interests. And from talking to people in Silicon Valley and policymakers in DC, I think efforts to change that are rapidly gaining traction. “We should slow down AI capabilities progress” is a much more mainstream view than it was six months ago, and to me that seems like great news.

In my ideal world, we absolutely would be pausing after the release of GPT-4. People have been speculating about the alignment problem for decades, but this moment is an obvious golden age for alignment work. We finally have models powerful enough to do useful empirical work on understanding them, changing their behavior, evaluating their capabilities, noticing when they’re being deceptive or manipulative, and so on. There are so many open questions in alignment that I expect we can make a lot of progress on in five years, with the benefit of what we’ve learned from existing models. We’d be in a much better position if we could collectively slow down to give ourselves more time to do this work, and I hope we find a way to do that intelligently and effectively. As I’ve said above, I think the stakes are unfathomable, and it’s exciting to see the early stages of coordination to change our current, unwise trajectory.

On the letter itself, though, I have a bunch of uncertainties around whether a six month pause right now would actually help. (I suspect many of the letter-signatories share these uncertainties, and I don’t have strong opinions about the wisdom of signing it). Here are some of my worries:

• Is it better to ask for evaluations rather than a pause? Personally, I think labs should sign on to ongoing commitments to subject each new generation of model to a third-party dangerous capabilities audit. I’m much more excited about requiring audits and oversight before training dangerous models than about asking for ‘pauses’, which are hard to enforce.

• Is the ask too small? I think I and the letter signers would generally agree that the ideal thing for society to do right now is something more continuous and iterative (and ultimately more ambitious) than a one-time six month pause at this stage. That means one big question is whether this opens the door to those larger efforts, or muddies the waters. Do steps that are in the right direction, but not sufficient, help us collectively produce common knowledge of the problem and build towards the right longer-term solutions, or do they mostly leave people misled about what it’s going to take to solve the problem? I’m not sure.

• What will we use the pause to do? An open letter like this one could be a step towards cooperative agreements on evaluations, standards, and governance, in which case it’s great. It could also go badly, if in six months labs go right back to developing powerful models and people walk away with the impression the pause was performative or meaningless. By itself, taking a few months off doesn’t gain us much (especially if a pause is entirely voluntary, so the least cooperative actors can simply ignore it). If we use that time well, to set up binding standards, good evaluations of whether our models are dangerous, and a much larger national conversation about what’s at stake here, then that could change everything.

• Does this ask impact companies unevenly? This specific call — to not train models larger than GPT-4 — is inapplicable to almost every AI lab today, because most of them can’t train models larger than GPT-4 in the next six months anyway. OpenAI may well be the only AI lab in a position to act on, or not act on, this demand.

That doesn’t delight me. Obviously, when regulations are being considered, one of the things companies inevitably do is try to design the regulations to advantage them and disadvantage their competitors. If proposed AI regulations appear to be an obvious grab at commercial rivals, I expect they’ll get less traction.

Moreover, I’m worried that an unevenly applied moratorium might backfire. If OpenAI can’t train GPT-5 for 6 months, other AI labs may use that time to rush to train GPT-4-sized models. That could mean that when the moratorium is lifted, OpenAI feels more pressure to get ahead again and may push for an even larger training run than they were planning originally. This moratorium could end up accomplishing very little except for making competitive dynamics even fiercer.

Overall, I’d prefer a policy that creates costs for all players and is careful to avoid creating potential perverse incentives.

Predicting the details of how future AI development will play out isn’t easy. But my best guess is that we’re facing a marathon, not a sprint. The next generation of language models will be even more powerful and scary than GPT-4, and the generation after that will be even scarier still. In my ideal world, we would pause and reflect and do a lot of safety evaluations, make models slightly bigger, and then pause again and do more reflecting and testing. We would do that over and over again as we inch toward transformative AI.

But we’re not living in an ideal world. The single most important thing we can do is to pause when the next model we train would be powerful enough to obsolete humans entirely, and then take as long as we need to work on AI alignment with the help of our existing models. That means that pausing now is mostly valuable insofar as it helps us build towards the harder, more complicated task of identifying when we might be at the brink and pausing for as long as we need to then. I’m not sure what the impact of this letter will be — it might help, or it might hurt.

by Kelsey Piper, Planned Obsolescence |  Read more:

[ed. See also: AI Risk is like Terminator; Stop Saying it's Not (Infinite Clout); and, The case for Terminator analogies (Slow Boring).]

The Great Forgetting - The Earth is Losing Its Memory

As a paleoclimatologist, my work revolves around the tenet that the past provides context and constraints for better understanding the future. Knowing how much the planet warmed when atmospheric carbon dioxide was as high or higher than it is today provides insight into the possible future trajectories of climate under rising greenhouse gases.

Earth records provide us with this information: Ice cores, tree rings, ocean sediments, stalactites and stalagmites in caves, growth rings in corals, tusks, and mollusks. These archives accrete memories on time periods varying from months to millions of years, allowing us to see a spectrum of Earth changes on various temporal and spatial scales—how biology, ocean, and ice respond to climate change in signature patterns, and the points at which those systems are pushed past thresholds.

This is one of the most important insights that paleoclimate archives provide: They show us how the real world breaks. How resilience folds into catastrophic failure. They show us the edges and asymmetries of the climate system: the thresholds of tolerance in ecological networks; the slow steady slog of diversification and the quick ax of extinction; the long timescales it takes for ice sheets to grow—accumulating million-year memories—and how fast they can melt, puddling history into storm surges that erode the banks of our futures. (...)

In geology, an “unconformity” represents an aberration in the normal accumulation of sediment, a glitch in the record-keeping of Earth’s history. “A stratum of amnesia in the geological record, where overlying rock, significantly younger than what lies below, represents some break in an otherwise continuous story of formation,” is how writer and poet Kim Stafford defined it.2

The longest lacuna in Earth’s history is known as the Great Unconformity. It represents a temporal gap ranging from a hundred million years to over a billion years, depending on the location. It’s visible in the Grand Canyon as the boundary between the Precambrian Vishnu Schist and the Cambrian Tapeats Sandstone, between which there is a billion years of missing time between about 1,600 and 600 million years ago. Looking at this line in the strata, it is hard to fathom all that would have conspired across that vast gulf of time, for which there is simply nothing. If it were instead to have been the last billion years that was erased, it would obliterate the entire history of complex life. No trace of a single animal having ever walked the land. No dinosaurs, no whales, no humans, no pyramids.

How does a billion years go missing? The Great Unconformity has long been a geological mystery, in no small part because it is a challenge to reconstruct history when records of history are missing.

It turns out, ice sheets are good shredders. Recent research suggests that the Great Unconformity may be a result of Snowball Earth—when the planet descended into deep cold (about 700 million years ago), and glaciers covered most of the land. A billion years of history was ground down by ice and bulldozed into the seafloor, where it was subducted into the Earth’s mantle and recycled into magma, ready to be remade into new history—albeit with a few hidden remnants of the past stored safely away in subterranean crystals.

While the erosive action of ice sheets may have been responsible for the largest unconformity in the Earth’s lithosphere, ice sheets themselves are some of the best memory banks on our planet. Greenland stores over 100,000 years of history. Antarctica stores over a million. These ice sheets are written by the daily weather, each snowstorm condensed into the jagged rhythms of ice age cycles that steadily build into mile-high mountains—the great brains of our planet, perched on the poles. (...)

Antarctica has been the slowest beast to awaken, but the icy tentacles that reach out to moor the giant are starting to slip. These floating ice shelves extend out from where the ice sheet is grounded to the bedrock, helping to stabilize the interior, but now they are starting to weaken from the forces of ocean warming and rising seas. As the ice shelves disintegrate into the ocean, the ice upstream accelerates its descent, increasing sea level.

In 2022, double heat waves hit the Arctic and Antarctic, temperatures soaring close to 40 degrees Celsius higher than usual. The Conger ice shelf in East Antarctica said its final farewell following this heat wave. West Antarctica has long been considered the more vulnerable to near-term ice shelf loss, but now, even the East is starting to show its fray. Heat makes easy work of forgetting. (...)

Before a tipping point in a complex system, there are early warning signals that may be detected. The most widely applicable of these early warning signals is “critical slowing down”—the phenomenon we are all familiar with before our computer crashes, and rather than heed the implications of this slower processing power, we jam at the keys in frustration, doubling down on our demands until the computer blacks out. These are the times information is most likely to be lost if it hasn’t been secured in long-term storage.

Critical slowing down indicates the system is losing its ability to attain its previous equilibrium and is instead becoming attracted or pulled into an alternate state. It is a loss of resilience, a loss of the negative feedbacks that help keep a system rooted in stability. Various subsystems that are sensitive to thresholds—such as the Amazon rainforest—are already showing signs of critical slowing down.

Given the complexity of the Earth System, it is hard to fathom the extent of information loss currently underway. There are, however, attempts to quantify the memory loss in the Earth System.

In one model, where anthropogenic CO2 emissions are the stressor, and the strain on the system is the ability of the land and ocean to sequester carbon, researchers show the latter is inherently slower than the former. They estimate that 60 percent of Earth’s memory had already been degraded by 1959, and that the ability for Earth to build-up memory has been impaired, reducing its capacity to respond to stresses within its natural stress-strain regime.16 Estimates of persistence in this model—akin to critical slowing down—are increasing, signaling a departure from the bounds of Earth’s natural regime well before 2050, if the stressors of rising atmospheric carbon dioxide continue their current trajectory. The ocean is undergoing memory loss too, increasing variability and reducing predictability of future temperature patterns.

The intractable problem we face is the asymmetry of timescales: It takes time to build memory, but it can be erased in a geological instant. Like so many things we take for granted, it is difficult to see these stabilizing forces until they are gone. As we untether the anchors of the past, the future becomes unmoored.

by Summer Praetorius, Nautilus |  Read more:

Image: Katherine Streeter

Why Asa Hutchinson’s View of the World Isn’t Working for Republicans

Ahead of his presidential announcement, Hutchinson, a Republican, spent several days in the first-in-the nation caucus state of Iowa, stirring speculation that he intended to enter into what he acknowledged is a tense national political landscape.

"I have made a decision, and my decision is I'm going to run for president of the United States," Hutchinson told Karl. "While the formal announcement will be later in April, in Bentonville [Arkansas], I want to make it clear to you, Jonathan, I am going to be running. And the reason is, I've traveled the country for six months, I hear people talk about the leadership of our country. I'm convinced that people want leaders that appeal to the best of America, and not simply appeal to our worst instincts."

The former governor told Karl he is inspired by his travels around the country over the last six months and acknowledged it would take "a lot of hard work and good messaging" to raise his national profile and break through a crowded primary field.

"It’s still about retail politics in many of these states, and also, this is one of the most unpredictable political environments that I've seen in my lifetime. So my message of experience, of consistent conservatism and hope for our future in solving problems that face Americans, I think that that resonates," Hutchinson said. (...)

"I think it's a sad day for America that we have a former president that's indicted, and so it's a great distraction, but at the same time, we can't set aside what our Constitution requires -- which is electing a new leader for our country -- just because we have this side controversy and criminal charges that are pending. And so we've got to press on, and the American people are gonna have to separate what the ideas are for our future," he said.

Karl pressed Hutchinson on whether he believes Trump should drop out of the race now that he’s been indicted.

"I do," Hutchinson said, standing by the position he took before Trump was charged. "I mean, first of all, the office is more important than any individual person. And so for the sake of the office of the presidency, I do think that's too much of a sideshow and distraction and he needs to be able to concentrate on his due process and there is a presumption of innocence."

"I've always said that people don't have to step aside from public office if they're under investigation, but if it reaches the point of criminal charges that have to be answered, the office is always more important than a person. And so, there's some consistency there. And I do believe if we're looking at the presidency and the future of our country, then we don't need that distraction," he added.

In contrast to GOP presidential candidates and potential candidates, Hutchinson did not blast the indictment as purely political, noting "the grand jury found probable cause and that's the standard for any criminal charges in our society." (...)

Despite his critique of Trump, the former Arkansas governor drew a distinction that his political position is in the “non-Trump lane” rather than the “anti-Trump lane.”

"When I say ‘non-Trump’, I want to be able to speak to the Trump voters. I want to be able to speak to all of the party and say, ‘This is the leadership that I want to provide, and I think that we need to have border security. I think we need to have a strong America; we need to spend less at the federal level.’ These are the values that I represent," Hutchinson said.

by Alisa Wiersema, ABC News | Read more:

Image: ABC News
[ed. He's still a Republican. But hey, anyone who promotes comity and normality and actual governing over batshit crazy and performative politics is definitely going to be an improvement. Granted, it's a pretty low bar, but at least he's giving the party an option going foward. We'll see what they do with it. See also: Why Asa Hutchinson’s view of the world isn’t working for Republicans (Politico).]

No master, daily work - The 9th Small Theater Festival BOOK PLAYS 沒有大師﹅日常用功　第九屆足跡小劇場演書節 Step Out 足跡 Nov 2018
via:

Life Long Pursuit. Feng zi watercolor exhibition《半生緣：一世情》鳯子畫展 / exhibition design / yunyi / Jun 2013
via:

Catherine DeQuattro Nolin, The Night Kitchen 2, 2022
via:

Executive Speak

“Our goal within this media evolution is to meet audiences wherever and however they consume content, by working with great worldwide storytellers to develop and execute their vision,” said Paul Buccieri, president and Chairman of A+E Networks Group. “Over the last four years, we have been on a journey to expand our production capabilities in both scripted and factual, and we’ve established key relationships in the talent management space as well as continued creating compelling content across our brands to further meet the needs of our valued partners and viewers.”

History Channel Greenlights New Projects From Kevin Costner, Bradley Cooper, Morgan Freeman (Variety)

Image: Yellowstone/Paramount

[ed. There are actually people on this planet that speak and understand this language. Imagine what that does to a person's brain. See also: How Kevin Costner Took Over Fox News Headquarters (Variety):]

The bulk of Fox Corp.’s portfolio is comprised of ad-supported media that consumers can access without a subscription. Even so, the company has dipped its toe into premium streaming waters with Fox Nation, which launched as a sort of “Netflix for conservatives,” but has expanded into what Klarman called “news adjacent” content. (...)

“I think the outdoors is something you can imagine is something our audience is very interested in, and Yellowstone in particular is an American icon, fronted by a guy who is starring in a show by the same name who is also an icon,” says Klarman. “It’s very meta.”

How Many New Songs Are Released Each Day?

Can you have too many songs?

That seems like a nice problem to have, until you start looking at the implications. They’re scary—both for musicians and fans.

Songs are like tribbles. One is cute, and you love it. Even five or ten are fine. But when you encounter a thousand of them, you run away as fast as you can.

I fear that’s happening in music right now. It’s become tribble-ized—and hence trivialized.

But let’s analyze the situation, and try to reach some conclusions. I’ll start with one tiny genre (jazz)—where it’s easier for me to wrap my mind around the numbers—and then try to assess the larger picture.

I warn you: It’s ugly.

Many jazz fans consider 1959 as the high point of the genre. That year saw the release of so many beloved recordings, including:

• Kind of Blue by Miles Davis

• Time Out by Dave Brubeck

• Mingus Ah Um by Charles Mingus

• The Shape of Jazz to Come by Ornette Coleman

• Giant Steps by John Coltrane

But that short list just scratches the surface. I could easily add another 10 or 20 jazz albums from 1959 that are undisputed classics—and probably another 50 to 100 recordings that are still cherished by many jazz fans today.

But here’s the surprising twist to all this. Somebody once told me that Downbeat, the leading jazz magazine, only received around 500 jazz records to review during the entire year of 1959.

It’s hard to imagine somebody releasing a jazz record in 1959, and not submitting it for review. So this means that around 20% or more of all the jazz records released that year are still heard and admired today—after 60 years!

In the fickle world of music, that’s an extraordinary success rate. Just releasing a jazz record that year gave you a 1-in-5 chance of ongoing success into the next millennium.

Now let’s compare this with the current day. It’s a depressing story, but somebody ought to tell it straight. I guess that means me.

Nobody really knows how many jazz albums are are getting released in the year 2023. But I probably have a better sense of this than most—because I constantly hear from musicians, labels, and publicists about their new music. Sometimes I will receive 50 or more pitches on new records in a single day.

Here’s my best guess: I’d estimate that somewhere between five and ten thousand new jazz recordings will be issued this year.

Clearly if you judge an art form by supply—and ignore that pesky little matter of demand—we are living in a golden age. We have far exceeded anything dreamt of by jazz lovers back in 1959.

In those distant days, a music fan could actually listen to every new jazz album, and still have time to spare. Nowadays that’s impossible. There’s not a single person on the planet who can even scratch the surface of all the accumulated music.

But here’s an even more extreme comparison. Let’s ask how many of these recordings will still be heard and loved in 60 years time. I’m afraid to answer that. But we are a long, long way from 1959 when a recording artist had a twenty percent chance, more or less, of making a lasting impact.

Here’s the sad truth for a musician in the current day. They can’t even begin to compete in their field, because they are almost always lost in the noise before the competition even begins.

This can’t be healthy. You flourish as a creative person when you actively compete at the highest levels of your vocation. If the actual situation is that you’re just lost in a crowd, you’re demoralized from the start. Instead of demonstrating your skills, you’re shouting out for attention—and ineffectively in almost every instance.

It’s like the difference between fighting for the heavyweight title, when you can demonstrate all your subtle boxing moves, and trying to prevail in a street riot. And that’s exactly what the music scene feels like in the current moment—a riot, with no rules or boundaries. You never win; at best, you survive.

Now let’s look for an even scarier number. How many recordings are released each year in all music genres?

In 2019, Spotify claimed that 40,000 tracks were added each day to its platform. And by 2021, the number increased to 60,000. But last September, the CEOs of two major labels made the staggering claim that 100,000 songs were now getting released each day.

Some argued that these figures were inflated. This couldn’t really be happening.

But a few weeks ago, Billboard reported that SoundCloud added 45 million tracks over a 12 month period. That works out to around 123,000 new songs every day.

The mind reels at this concept.

Here’s another odd twist: AI is now composing music at a fast clip. As I’ve written elsewhere, you can now get customized AI songs for just a few dollars—and the music is delivered almost instantaneously.

But do we really need AI songs, if human musicians are creating 123,000 per day? I’m not sure what the saturation point is, but we must have reached it long ago. And I’m not even going to try to guess how many of these songs resemble other songs—because, after all, there aren’t an infinite number of melodic phrases in a 12-tone system.

Is there a single unsung melody left to sing? The scope of copyright infringement must be off the charts.

Here’s another twist. The numbers are still staggering even if we focus solely on the three major labels.

These three companies alone issue 3,900 tracks per day.

Now that’s a truly humbling number. Not long ago, getting a record contract with a major label put you in a small elite group. Now you’re just another face in the crowd—and that crowd gets larger every day.

Let me try to put all of this into perspective, at least as best I can. Below are my conclusions—along with some suggestions on how we ought to respond to this deluge of music.

by Ted Gioia, The Honest Broker |  Read more:

Images: The Trouble With Tribbles/Star Trek; Jimmy Taurrell for Variety via:

Should We Ban the Purchase of Cigarettes for Life? A US Town is Trying

Silbaugh is one of 255 town meeting members in Brookline, Massachusetts, an urban-suburban “island inside Boston” – its neighboring boroughs have long been swallowed into the city. The town’s 63,000 residents are 70% white, with a median household income of $122,000.

Two years ago, Silbaugh and her neighbor Anthony Ishak passed an ordinance banning anyone born after 1 January 2000 from ever buying cigarettes in their town. The measure took effect in September 2021. The idea was to curb youth smoking rates without yanking anything away from people already addicted, essentially grandfathering out tobacco. Every year, there’d be a smaller slice of the population that could buy cigarettes, until one day no one would be left. At least, that was the vision.

In tobacco’s heyday in the mid-20th century, 45% of US adults smoked. Fast-forward to 2020, after decades of aggressive anti-smoking campaigns, and the rate was down to 12.5%. It’s progress, to be sure, but cigarettes still kill roughly half a million people in the US every year – more than car accidents, alcohol, murders, suicides and illegal drugs combined. If current trajectories persist, tobacco will kill 1bn people in the 21st century, or one person every three seconds.

So it’s hard to imagine that a world without cigarettes would be a bad thing. Prohibition might fast-track it and help avoid needless suffering, as public health officials will remind you. But at what cost? There’s obviously no enumerated right to cigarettes, but there is a right to live our lives as we see fit, so long as we don’t infringe on others’ ability to do the same.

While the tobacco endgame – smoking rates below 5% – seems ultimately inevitable, getting the timeline right is the $1.85tn question. Should cigarettes die on their own, or at the hands of the state? (...)

Over time, though, nicotinic receptors become less responsive to the drug, so the brain starts to create more of them. And these multiplying receptors are like petulant baby birds, incessantly crying to be fed. Go too long without a cigarette, and withdrawal’s sure to follow – irritability, depression and cravings galore. Of course, there’s no gun to the head, but smoking another cigarette takes the pain away and offers an immediate sense of pleasure, if only for a short while.

The biology is interesting and all, says Carl Hart, a neuroscientist and professor of psychology at Columbia University, but it alone can’t tell us what the boundaries of public health should be – whether prohibition is proper in the name of the collective good.

Known for his controversial views on drug use (he advocates for the legalization of all drugs and wrote a book in which he admits to having used heroin regularly for the past five years), Hart wears his hair in long dreadlocks, and his hands move prophetically, a beat or two ahead of his words. He gesticulates with indignation as he tells me that any vision of a tobacco endgame is that of sick or naive zealots.

“Our declaration of independence, the first founding principles of the country, says that we are free, and we have the right to these three birth rights: life, liberty and the pursuit of happiness. Now they’re trying to overturn those basic principles,” Hart says. “Public health is all around us to enhance safety, but not to take away the activity.”

He’s not advocating for a world without regulation – he believes in age restrictions, banning cigarettes indoors and eliminating harmful additives – but one with free choice. “I don’t smoke tobacco cigarettes, but it’s not up to me to decide what benefit other people get from that product,” Hart says. “I am not the ruler and lord of their domain.”

He adds that history is teeming with failures of similar policies. Alcohol prohibition was meant to reduce domestic violence and poverty but instead fueled unimaginable violence and organized crime, with homicide rates increasing by 78%. The “war on drugs” was similarly disastrous, incarcerating millions of Black and Hispanic Americans without reducing the availability of illegal drugs. “Now if we think about tobacco in the same way,” Hart says, “the people who are gonna pay the price are the same people who pay the price – poor people.”

So one of birthdate bans’ selling points is equity, reducing high smoking rates in vulnerable communities, according to Silbaugh and Ishak, the Brookline town meeting members.

Hart dismisses that as a cruel joke. “This is what we do: we pretend to care about those communities and not really deal with the issues they face.” They need gainful employment, health insurance, and pensions, but public health exploits their deprivation to justify paternalistic bans instead.

“What the fuck,” Hart scowls. “It just blows my mind.”

Take any set of drugs – say marijuana, heroin, amphetamines and hallucinogens – and ask yourself which ones should be legal versus illegal. There are certainly places at the extremes: Portugal, for instance, has decriminalized all drug use, while Singapore has attached the death penalty to consumption. But what exactly does the US do?

Mark Gottlieb, a public health lawyer at Northeastern University, describes how federal drug policy seems almost random, carved into stone by whoever’s in power, oddities of their own times.

Hart puts it more directly: “Whatever the white majority says is illegal, is illegal.”

The ideal, of course, would be to have society plot all drugs on a spectrum and legalize those whose benefits outweigh the harms. But it’s probably ridiculous to claim that harm reduction alone drives US drug policy, given that alcohol and cigarettes are not only legal but remarkably accessible. “I find it ironic that these two drugs collectively cause way more destruction, loss of life, loss of productivity than every other drug of abuse all combined,” says Lukas.

It might even be overly generous to lump cigarettes with alcohol. After all, when used exactly as its manufacturers intended, tobacco kills up to half its users, making cigarettes the deadliest object in the history of human civilization. And while 5% of individuals who drink are alcoholics, 90% of those who smoke are nicotinics, according to Robert Proctor, a professor of history at Stanford University. “But that’s not even a word, right? We don’t even have a word for nicotinics because almost everyone who smokes is addicted.” 

by Simar Bajaj, The Guardian |  Read more:
Image: Lucas Jackson/Reuters

Beware of The Blurb

Blurb is a funny sounding word. It’s phonetically unappealing, beginning and ending with unattractive voiced bilabial stops, and its definition—an advertisement or announcement, especially a laudatory one—carries some of the same meaning as another unattractive word, blubber, which evokes excess in its dual definition as both an expostulation of unrestrained emotion as well as excess fat. For these reasons alone, any sensible person should beware of blurbs.

First, an origin story. The year is 1907, and author and humorist Gelett Burgess has been invited to the annual American Booksellers Association dinner to present copies of his new book, Are You a Bromide? Burgess presented a mock cover for the book, featuring a made-up “spokesperson” named Miss Belinda Blurb, whose image was purportedly lifted from a dental advertisement “in the act of blurbing.” She shouts, hand cupped around her mouth, that the book has “gush and go to it,” and a “certain something which makes you want to crawl through thirty miles of dense tropical jungle and bite somebody in the neck.” This was the first use of the word blurb as we know it today. As a noun, Burgess himself defined a blurb as “a flamboyant advertisement; an inspired testimonial”; as a verb, “to flatter from interested motives; to compliment oneself.”

Blurbs had been used in publishing long before they had a name. One of the earliest examples of a book blurb in the United States was penned by none other than Ralph Waldo Emerson. It appeared on the jacket of Walt Whitman’s Leaves of Grass in 1855. The blurb was taken from a letter that Emerson had sent to Whitman, and which Whitman included on the spine of his book; it salutes his promise as a poet, and reads: “I greet you at the beginning of a great career, RW Emerson.” (...)

Few writers decline to blurb a book since, more often than not, they have been personally appealed to by the author, or the author’s editor or agent (both of whom they are likely to know). More importantly, the blurber’s name will appear on the book in conjunction with the author and other blurbers, so the blurb is as much an advertisement for the blurber as it is an endorsement of the book.

In a recent essay on the controversial publication of American Dirt, critic Christian Lorentzen questioned the validity of the glowing blurbs that the book received from such literary luminaries as Stephen King, Sandra Cisneros, and John Grisham. “The blurb system is corrupt on its face,” Lorentzen writes. “Blurbs may be earnest and true, but they are always the product of favors being called in: from authors’ friends, from agents’ other clients, from publishers’ other authors. Everyone knows this.”

During her tenure, former London Review of Books editor Mary-Kay Wilmers instituted a policy that any sentence in a review that could be used as a quote on a book was to be cut. When asked why, she answered: “Those are never good sentences.” David Foster Wallace shared Wilmers’s aversion to the language of blurbing. At a public reading in 2004, when questioned about the blurbs that adorn the jackets of his own novels, Wallace coined the term “blurbspeak,” which he defined as “a very special subdialect of English that’s partly hyperbole, but it’s also phrases that sound really good and are very compelling in an advertorial sense, but if you think about them, they’re literally meaningless.” This did not, however, stop him from providing blurbs for many books including Dave Eggers’s A Heartbreaking Work of Staggering Genius and his friend Jonathan Franzen’s The Corrections, which he called “a testament to the range and depth of pleasures great fiction affords.”

Even our resident blurber Shteyngart agrees with Wallace’s claim that hyperbole is inherent to the form. Making an appearance as his charming, irrepressible self in the documentary short “Shteyngart Blurbs,” the author defends such exaggeration: “No hyperbole can be hyperbolic enough because very few people want to read this stuff.” (...)

As Shteyngart notes in the documentary, when a galley arrives, many blurbers read no more than the publisher’s plot summary which is written by the editor or publicity department or both. It is then quite easy for a blurber to riff off of what they’ve been supplied. Blurbs generally share a common format across all genres of books: Author praise: “A talented writer who…”; “Her intelligence is such that….” One-word gushing: “electrifying”; “gripping.” Two-word slobbering: “wickedly smart”; “hauntingly beautiful.” Dubious equivalences: “as satisfying as it is unsettling”; “as sharply conceived as it is brilliantly written.”

Aside from the irony inherent in Shteyngart’s quip that “no hyperbole can be hyperbolic enough,” his assertion is insightful because language so easily crosses the threshold from hyperbole to hysteria. When the goal is to pump up the volume “high enough” so it can be heard far and wide, blurbers tend to say practically anything, confounding the boundaries between true and false in the hopes of grabbing attention. This frequently leads to situations where there is but the most tenuous connection between the work at hand and the blurb.

by GD Dess, The Millions |  Read more:

Image: uncredited

Antoni Tapies, nocturn matinal (1970)

via:

Dear Mar-a-Lago Club Members

My Indictment is Going to Cost You

Dear Mar-a-Lago Club Members:

This is Donald Trump, your president. As you might have heard, I have been charged by the totally corrupt Manhattan district attorney Alvin Bragg, for a $130,000 so-called hush money payment I made in the closing weeks of the 2016 presidential campaign to porn star Stormy Daniels, a person who is not only an untrusty witness but also a very poor judge of the surprising girth of mushrooms.

I want to assure you as Mar-a-Lago Club members that these vicious and totally unfair charges will be defended by my legal team, who will be some of the brightest legal minds in the country that return my phone calls. And Rudy Giuliani.

Rudy will have more to say about this at an upcoming news briefing outside of Amelia’s Smarty Plants in Lake Worth Beach later this week.

In the meantime, many of you have come up to me and said, “Sir, Sir, what can we do? Sir, — you all call me ‘Sir’ — how can we help you defend against these complete and total made-up charges, Sir?”

Finding a little extra money to help Trump

Normally, I would say there’s absolutely nothing you can do. As you know, I am like the richest person you’ve ever known. I got more money than anybody. Right? Forget about it. I mean, I'm super, super rich.

But lately, it’s been hard getting through to my longtime chief financial officer Allen Weisselberg. The phone service inside the Rikers Island jail is a national disgrace.

So, I’m asking you to do a little something. It’s the least you can do for your country, which I love so much, and prove every time I hug American flags on stage.

All I want each of you to do is to find $130,000 and send it to me for legal expenses. There’s nothing wrong with us saying that we’ve recalculated your dues.

Being able to applaud me at Mar-a-Lago as I take my meals with Melania at a roped-off table on the lanai is worth hundreds of thousands of dollars to you.

So what’s another $130,000 adjustment? Nothing. And people will love you for it.

It will make you a hero to the 74 million Americans who voted for me. On the other hand, if you don’t give me that $130,000 I’m asking for, well those same people will be very angry with you.

by Frank Cerebino, Palm Beach Post |  Read more:

Image: uncredited

[ed. Finally, the pussy grabs back. Best wishes and sincere hopes for many more to come. See also: Hush money to a porn star: of course this was how Trump was indicted (The Guardian):]

"Daniels, for years a successful porn performer, had met Donald Trump at a celebrity golf tournament in 2006. According to her, he invited her to his hotel room, offered her work on his TV show and then had sex with her. The two remained friendly afterwards; Trump invited Daniels to the launch of his Trump Vodka brand the following year. It’s the kind of thing you suspect that these two people would have written off as a funny story. Instead, it’s the impetus for one of the most politically volatile prosecutions in the nation’s history: the first criminal indictment of a former president, which was issued on Thursday by a federal grand jury in New York.

Stormy Daniels and the illegal, fraudulent machinations that the Trump campaign allegedly undertook to pay her off during the height of the presidential campaign in 2016 have always struck me as the most quintessential of Trump’s many scandals. Trump denies Daniels’ allegations, but in retrospect, with the hindsight of what we’ve come to learn of him, the scene she recounts is almost unbearably true to his character: the gathering of low-rent celebrities, the paltry quid pro quo offer, the golf and the sad, adolescent fantasy of sex with a porn star. The whole story drips with Trump’s defining attribute: the desperate and insatiable need to have his ego gratified."

and, of course: 

"The bigger the alleged crime, the louder he airs grievances and the more he plays the victim... This pattern came into a focus earlier this month when Trump falsely predicted his own arrest. Republicans leaped to his defence and he reportedly raised $1.5m in three days; on Thursday night he quickly sent out another fundraising email."

AI Risk ≠ AGI Risk

Is AI going to kill us all? I don’t know, and you don’t either.

But Geoff Hinton has started to worry, and so have I. I’d heard about Hinton’s concerns through the grapevine last week, and he acknowledged them publicly yesterday. (...)

My beliefs have not in fact changed. I still don’t think large language models have much to do with superintelligence or artificial general intelligence [AGI]; I still think, with Yann LeCun, that LLMs are an “off-ramp” on the road to AGI. And my scenarios for doom are perhaps not the same as Hinton’s or Musk’s; theirs (from what I can tell) seem to center mainly around what happens if computers rapidly and radically self-improve themselves, which I don’t see as an immediate possibility.

But here’s the thing: although a lot of the literature equates artificial intelligence risk with the risk of superintelligence or artificial general intelligence, you don’t have to be superintelligent to create serious problems. I am not worried, immediately, about “AGI risk” (the risk of superintelligent machines beyond our control), in the near term I am worried about what I will call “MAI risk”—Mediocre AI that is unreliable (a la Bing and GPT-4) but widely deployed—both in terms of the sheer number of people using it, and in terms of the access that the software has to the world. A company called Adept.AI just raised $350 million dollars to do just that, to allow large language models to access, well, pretty much everything (aiming to “supercharge your capabilities on any software tool or API in the world” with LLMs, despite their clear tendencies towards hallucination and unreliability).

Lots of ordinary humans, perhaps of above average intelligence but not necessarily genius-level, have created all kinds of problems throughout history; in many ways, the critical variable is not intelligence but power, which often caches out as access. In principle, a single idiot with the nuclear codes could destroy the world, with only a modest amount of intelligence and a surplus of ill-deserved access.

If an LLM can trick a single human into doing a Captcha, as OpenAI recently observed, it can, in the hands of a bad actor, create all kinds of mayhem. When LLMs were a lab curiosity, known only within the field, they didn’t pose much problem. But now that (a) they are widely known, and of interest to criminals, and (b) increasingly being given access to the external world (including humans), they can do more damage.

Although the AI community often focuses on long-term risk, I am not alone in worrying about serious, immediate implications. Europol came out yesterday with a report considering some of the criminal possibilities, and it’s sobering. (...)

Perhaps coupled with mass AI-generated propaganda, LLM-enhanced terrorism could in turn lead to nuclear war, or to the deliberate spread of pathogens worse than covid-19, etc. Many, many people could die; civilization could be utterly disrupted. Maybe humans would not literally be “wiped from the earth,” but things could get very bad indeed.

How likely is any of this? We have no earthly idea. My 1% number in the tweet was just a thought experiment. But it’s not 0%. (...)

We need to stop worrying (just) about Skynet and robots taking over the world, and think a lot more about what criminals, including terrorists, might do with LLMs, and what, if anything, we might do to stop them.

But we also need to treat LLMs as a dress rehearsal future synthetic intelligence, and ask ourselves hard questions about what on earth we are going to do with future technology, which might well be even more difficult to control. Hinton told CBS, “I think it's very reasonable for people to be worrying about these issues now, even though it's not going to happen in the next year or two”, and I agree.

by Gary Marcus, The Road to AI We Can Trust (Substack) |  Read more:

Image: uncredited

[ed. Hinton article here. Europol report here. See also: Nick Bostrom's paper Existential Risks - Analyzing Human Extinction Scenarios and Related Hazards (Bangs, Crunches, Shrieks, Whimpers); and, Global Catastrophic Risk (Wikipedia). Have a nice day.]

Massive Seaweed Belt Still on Track to Hit U.S.

A giant seaweed belt twice the width of the United States has its sights set on Florida, where it could wreak havoc as it washes ashore.

The 5,000-mile-wide bloom is a belt of sargassum, a floating brown algae that usually floats in the Atlantic. The problems start when it comes on land and starts to rot. As it decays, sargassum lets off hydrogen sulfide and smells like rotten eggs, explains the Florida Department of Health. It can irritate people’s eyes, nose and throat, and trigger breathing issues for people with asthma.

Massive seaweed belt still on track to hit US: When will it arrive? (The Hill).

Image: USF/NOAA

[ed. Didn't know it was this big.]

Michiko Itatani - Codebreaker, from Celestial Maze series 20-B-01, 2020
via:

Mount Fuji
via:

Prompt Engineering: The Art of Writing Text Prompts

Do you like AI tools but have trouble getting the desired results? Then it’s time to meet the concept of AI prompt engineering. Whether you use ChatGPT or DALL-E 2, i.e. text-to-text or text-to-image AI tools, you must learn to ask the right questions to get the results you want. In light of the potential of these tools, it is reasonable to assume that those who pose such inquiries will rise in prominence as they become more embedded in the business world. Given this foresight, perhaps we have already begun to experience the profession of the future: prompt engineers. (...)

AI prompt engineering is the process of designing and creating prompts, or input data, for AI models to train them to perform specific tasks. This includes selecting the appropriate data type and formatting it so the model can understand and use it to learn. AI prompt engineering aims to create high-quality training data that will enable the AI model to make accurate predictions and decisions. It is an essential step in the development and deployment of AI systems.

AI prompt engineering is the key to limitless worlds (Dataconomy)

Image: Brendan Lynch/Axios

[ed. That was fast. See also: AI's rise generates new job title: Prompt engineer (Axios); Best practices for prompt engineering with OpenAI API (OpenAI); Prompt Engineering (Wikipedia); and, Grantable (AI grant writing assistant.]

Seattle Lowered Its Standards All the Way Under a Bridge

Gov. Jay Inslee this past week dropped in on one of Seattle’s worst homeless encampments. Surrounded by heaps of trash along with dozens of stripped and burned-out cars under the First Avenue South bridge, he observed that it was all overwhelmingly awful.

“This is a scourge in our state,” Inslee said. “There is no excuse. … The people of the state of Washington do not accept this level of squalor.”

Well … we should talk about that.

It’s good the governor is finally bringing some outrage to this cause, even at this late date. Because it is outrageous that such a wealthy city has countenanced inhuman living conditions like this for years. And also that it sat idly by while some encampments, like this one, morphed from zones of social need into alleged criminal enterprises.

Do we really not accept this squalor, as the governor said? The record suggests that in Seattle at least, we do.

I’m not saying people are happy about it, or don’t care. Plenty of groups are working overtime to try to help.

But along the way, our city seems to have adapted. Whether it’s out of an abundance of tolerance, or concern about demonizing poverty, or simply feelings of powerlessness, we’ve become numb bit by bit to people living under bridges. Haven’t we?

Now you who are reading this may rightly say, “Well I don’t accept it, and I’ve wanted the city to do something about it for years.” I’m talking about the collective we, as carried out through the democratic actions and policies of the city. The city some years ago stopped acting as if people living in the most deplorable conditions imaginable was any urgent crisis.

Seven years ago, in fact, it started seeming like this might become Seattle’s Achilles’ heel.

Another politician, former King County Prosecutor Dan Satterberg, had just given a fiery call-to-action speech like Inslee’s. Satterberg pronounced the Jungle, an encampment on a greenbelt next to Beacon Hill, to be a humanitarian disaster. “Worse than Third World conditions,” he said.

At City Hall though, council members and nonprofit officials equivocated. They said it would be gentler to leave the 400 people in the Jungle than to push them into emergency shelter. They suggested equipping the Jungle like a campground, with lockers in the woods for storing belongings and bins under Interstate 5 for used hypodermic needles.

This was after five people had been shot, with two killed, in a dispute.

“They need to get people out of there,” I quoted one appalled, formerly homeless woman back then. “I can’t believe they were talking about leaving them there.”

I couldn’t either. They did leave them there, for nine more months, before finally clearing the encampment, over protests, in fall 2016. (...)

Inslee said there’s no excuse for it, but there are plenty. The main one is legit and well-intended: The city wanted to build real housing to solve this crisis, not more temporary shelter. This one thing, though, also effectively sanctioned these encampments. Housing takes years to build, so in the meantime, people under bridges it is.

Inslee, Seattle Mayor Bruce Harrell and others now are pushing to shift that. Encampments are being shut down after social outreach teams offer temporary shelter, which is what we should have been doing all along. Often the camps don’t get priority, though, until somebody gets shot or killed — more perverse fallout from our abundance of compassion, along with an inability to stand up shelter faster.

The police chief told me in January that if Seattle could get a handle on just the unauthorized encampments, crime would fall significantly. Not because most homeless people are committing crimes. But because crime-focused individuals have been using homeless camps as human shields.

“They’ll set up tents or RVs in the middle of an encampment and run drugs or other operations out of there,” Chief Adrian Diaz said. “It makes it very difficult for our officers to intervene. The other homeless persons who are there often end up as victims.”

I think a lot about the old phrase “defining deviancy down.” It’s when a society is beset by aberrant behaviors to the point that it responds by lowering its own standards.

by Danny Westneat, Seattle Times |  Read more:

Image: Ken Lambert / The Seattle Times

[ed. The U.S. government set up Japanese internment camps during the WWII that, while awful in law, intent, practice, and design at least provided basic forms of shelter, protection and hygene for thousands of people, even if they were policed like prisons.]

I Went on a Package Trip for Millennials Who Travel Alone. Help Me.

Imagine walking into a party where you know almost no one (pathetic) — a party at which I, a stranger to you (probably), have arrived well before you (sorry). Should this occur in real life, it is inevitable that shortly after your entrance, as you are tentatively probing the scene in search of safe ingress into social traffic, I will yank you, abruptly, into the middle of a conversation. I will turn to you and start talking as if you’d been involved in the discussion for an hour. I will lob questions at you that are tailored so that any answer you give can be right. Soon, you will forget I dragged you into this interaction; your easy popularity will seem, in retrospect, inevitable. You will most likely feel at least vaguely friendly toward me, because I so clearly want to be your friend. And the whole time I am doing this — because, despite your rewritten recollections, I am the one doing all of this — I will be thinking: Oh, my God, I’m doing it again. I hate this. I hate this. Why can’t I stop doing this to people?

Of all my bad habits, it is the ruthless desire to befriend that exerts the strongest pull on my behavior. Not that I want more friends — God, no. If anything, I’d love to drop about 80 percent of the ones I have, so I could stop remembering their birthdays. But because I can’t quit — because constantly pulling strangers into my orbit is what stabilizes my bearing in the universe — I have determined to double down. And so, in January, I booked a package vacation to Morocco through a company whose stated aim — beyond offering package vacations — is to help people in their 30s and 40s make new friends.

That millennials are the largest human adult cohort alive; in or about to enter their peak-earning years; less likely than earlier generations, at the same age, to live with a spouse and/or offspring; and highly susceptible to YOLO — a brain condition that makes a nine-day vacation to Croatia sound like a fun and affordable alternative to homeownership, which seems impossible anyway — would seemingly be enough to justify the existence of a travel company dedicated to serving them. Indeed, there is a nascent industry devoted to creating millennial-oriented travel package experiences of the type generally set aside for people much younger (e.g., Birthright Israel) or older (e.g., Rhine river cruises). In promotional copy, these companies’ sleek websites deploy the verb “curate” to describe the work of travel agents. Flash Pack, which aims to lure vacationers who would otherwise be traveling solo and marshal them into traveling bands of up to 14, is one such business.

What makes Flash Pack unusual is its “mission” — “to create one million meaningful friendships” — and a method of execution that it telegraphs with evangelistic zeal: “We obsess over the group dynamics,” its website explains on one page. “We absolutely obsess over the group dynamic,” it states on another. “We’re completely obsessed with it” (“it” being the group dynamic), Flash Pack’s 42-year-old chief executive, Radha Vyas, is quoted as saying on an F.A.Q. page intended to calm nervous vacationers. Another page, titled “How It Works,” opens with the promise that the company “obsesses over the group dynamic, doing everything in our power to ensure you’re comfortable and building friendships within the first 24 hours.”

With this intention, the agency stands in stark, even proud violation of a sociological paradox: to have many friends is a desirable condition; to plainly seek to make friends is unseemly and pitiful. Millennials’ broad acceptance of the taboo around extending oneself in friendship — perhaps an aversion to participation inherited from their direct predecessors, Generation X — is particularly irrational, given that millennials report feeling lonely “often” or “always” at much higher rates than members of previous generations.

Who, I wondered as I scrolled through the inviting images on the company’s home page, are the millennial adults drawn to a pricey international vacation for the purpose of befriending strangers? If I plunged into a trip chosen at random, would I surface to find myself flailing among social incompetents — phone-addled young people who yearn for real-life connections but are unable to forge them under normal conditions? Or would I be surrounded by the sociopathic winners of this great game — the Jeff Bezoses of friend-making? Obviously, my fellow vacationers would be natural freaks of some kind — but would they be so because they had overcome the intrinsic shame of seeking friends or because they were naturally immune to it?

The mystery started to resolve itself two weeks out from our trip, when every participant of the “Morocco Highlights” tour was added to a WhatsApp group and encouraged to introduce themselves — a suggestion we responded to with so much zeal you would think it were an assignment that constituted 60 percent of our grade; and we were determined to maintain our perfect grade-point average; and we had actually been secretly hired as “plants” by the school administration to sit in on this class, in the hope that we would contagiously motivate the real students to strive for comparable excellence, creating a domino effect that would boost the school’s rankings; and we would love to take advantage of extra-credit assignments (if they were available); and, actually, we had gone ahead and conceived and executed what we felt might be some edifying extra-credit assignments, in case none were available.

We sent portraits of our pets, announced which items on the itinerary we anticipated most eagerly and provided photos of what we loved most about the places where we lived (the mountains of North Carolina; sunlight gleaming off the Charles River; the solitary beauty of a Baltic beach, which, it was hoped even before meeting, some of us would “come visit one day!”). I wondered publicly in the chat if anyone in the group might be “superorganized” and willing to share a packing list. Within 60 seconds, I received in reply an image consisting of a tabular representation of our itinerary, each column head designating a day, underneath which was a cell listing the major activities of that day (extracted and paraphrased from the official itinerary), underneath which was a full-length photograph of the sender, wearing the exact outfit, including shoes and coat, she intended to wear on that day, for those activities.

When I asked Radha Vyas, who founded Flash Pack with her husband, Lee Thompson, to give me the profile of a typical patron, she described her clients as “decision makers or leaders” in their regular lives who “want somebody else to take control” of their vacations. “Lots of our customers are lawyers, doctors, and they’ve done really, really well in their careers,” she said over video chat from London — so well that they have developed “decision fatigue” from the litany of correct decisions they have been forced to make while scaling new professional heights. “They just want to turn up,” Vyas said. “Somebody tells you where to be, what time, what to do, what to wear, and you can just let go.”

by Caity Weaver, NY Times | Read more:
Image: Rosie Marks for NYT

Reaper Madness: What Really Happened Between Russian Fighter Jets and a US Drone Over the Black Sea

Reaper madness: What really happened between Russian fighter jets and a US drone over the Black Sea (Task and Purpose). How the incident really went down.
Image: A Sukhoi Su-27SKM fighter jet at MAKS-2005 airshow. (Dmitriy Pichugin/Wikimedia Commons)