Group items matching

in title, tags, annotations or url

the Telluride workshops were being organized by two college-age students, filled with the spirit of the times

From what I gleaned, they involved crudely conveying certain dogmatic assertions, no matter what topic the workshops were ostensibly about:Experiencing hardship conveys authority. There is no hierarchy of oppressions—except for anti-black oppression, which is in a class of its own. Trust black women.Prison is never the answer.Black people need black space.Allyship is usually performative.All non-black people, and many black people, are guilty of anti-blackness.There is no way out of anti-blackness.

It is tempting to add: Such is life. Such is democratic life. We each have different, partial knowledge. We each get things wrong, over and over. At our best, we enter the fray by listening to each other and complementing and challenging the insights of our fellows. In the process, over years, decades, we are oriented toward justice and truth.

Specific words, phrases, arguments, and images from a text offer essential friction for conversation, holding seminar participants accountable to something concrete.

The seminar assumes that each student has innate intelligence, even as we come from different backgrounds, have different amounts and sorts of knowledge, and different skills. We can each be formed best if we take advantage of our differing insights to push each other, over time, again and again. When this practice is occasioned by carefully curated texts—not exclusively “great books,” but texts that challenge each other and us as they probe issues of essential importance—a seminar succeeds.

Each intervention in a seminar is incomplete, and gets things wrong. Each subsequent intervention is also incomplete, and also gets things wrong. But there are plenty of insights and surprises, for each participant looks at a text with different eyes

By its nature, a seminar requires patience. Day by day, one intervention builds on another, as one student notices what another student overlooked, and as the professor guides the discussion toward the most important questions

The effects on the seminar were quick and dramatic. During the first week, participation was as you would expect: There were two or three shy students who only spoke in partner or small-group work, two or three outspoken students, and the rest in the middle. One of the black students was outspoken, one was in the middle, and one was shy. By the second week of the seminar, the two white students were effectively silent. Two of the Asian-American students remained active (the ones who would soon be expelled), but the vast majority of interventions were from the three black students. The two queer students, one Asian and one white, were entirely silent

“Two of the Asian-American students would be expelled from the program.”In the 2022 anti-racism workshops, the non-black students learned that they needed to center black voices—and to shut up.

If the seminar is slow food, the anti-racist workshop put on by college-age students is a sugar rush. All the hashtags are there, condensed, packaged, and delivered from a place of authority. The worst sort of anti-racist workshop simply offers a new language for participants to echo—to retweet out loud.

In their “transformative-justice” workshop, my students learned to name “harms.”

In the language of the anti-racism workshop, a harm becomes anything that makes you feel not quite right. For a 17-year-old at a highly selective, all-expenses-paid summer program, newly empowered with the language of harm, there are relatively few sites at which to use this framework. My seminar became the site at which to try out—and weaponize—this language.

John McWhorter asserts that anti-racism is a new religion. It was an idea I quickly dismissed. Last summer, I found anti-racism to be a perversion of religion: I found a cult

the features of cults have become familiar in popular culture. There is sleep deprivation. Ties to the outside world are severed. The sense of time collapses, with everything cult-related feeling extremely urgent. Participants are emotionally battered. In this weakened state, participants learn about and cling to dogmatic beliefs. Any outsider becomes a threat.

The feature of a cult that seems to be missing from this story is a charismatic leader, enforcing the separation of followers from the world, creating emotional vulnerability, and implanting dogma. Enter Keisha

Pushing anti-racism to its limits, what we reach isn’t just hollow doctrine, but abuse: Pathological relationships that cut us off from the world, from the give-and-take of reasons and feelings unfolding over time that makes up life in the world.

We see this crystal clear in the paradoxes that I encountered

The experience was supposed to be organized around a “transformative justice,” rather than a punitive model, yet the community managed to expel two of its members.

Students continually voiced their desire to find practical actions to help change the world, but after four weeks, they had learned to say that anti-blackness is so foundational, the world could never change

The students wanted freedom, for themselves and for all, but they started to say that the only route to freedom is indoctrination: having me tell them what to think.

Saddest of all, for me, was hearing what the black students said. They needed extra help, they were struggling to understand anything from the readings, and they couldn’t even know what questions to ask unless they had guidance—first Keisha said this, then the black students said it, then their “allies” repeated it in solidarity with them

“No one believed me that working for Elon was the way it was until they saw how he operated Twitter,” Bernal said, calling Twitter “just the tip of the iceberg on how he operates Tesla.”

In April 2019, at a showcase dubbed “Autonomy Investor Day,” Musk made perhaps his boldest prediction as Tesla’s chief executive. “By the middle of next year, we’ll have over a million Tesla cars on the road with full self-driving hardware,” Musk told a roomful of investors. The software updates automatically over the air, and Full Self-Driving would be so reliable, he said, the driver “could go to sleep.”

Investors were sold. The following year, Tesla’s stock price soared, making it the most valuable automaker and helping Musk become the world’s richest person

To deliver on his promise, Musk assembled a star team of engineers willing to work long hours and problem solve deep into the night. Musk would test the latest software on his own car, then he and other executives would compile “fix-it” requests for their engineers.

Those patchwork fixes gave the illusion of relentless progress but masked the lack of a coherent development strategy, former employees said. While competitors such as Alphabet-owned Waymo adopted strict testing protocols that limited where self-driving software could operate, Tesla eventually pushed Full Self-Driving out to 360,000 owners — who paid up to $15,000 to be eligible for the features — and let them activate it at their own discretion.

Tesla’s philosophy is simple: The more data (in this case driving) the artificial intelligence guiding the car is exposed to, the faster it learns. But that crude model also means there is a lighter safety net. Tesla has chosen to effectively allow the software to learn on its own, developing sensibilities akin to a brain via technology dubbed “neural nets” with fewer rules, the former employees said. While this has the potential to speed the process, it boils down to essentially a trial and error method of training.

Radar originally played a major role in the design of the Tesla vehicles and software, supplementing the cameras by offering a reality check of what was around, particularly if vision might be obscured. Tesla also used ultrasonic sensors, shorter-range devices that detect obstructions within inches of the car. (The company announced last year it was eliminating those as well.)

Even with radar, Teslas were less sophisticated than the lidar and radar-equipped cars of competitors.“One of the key advantages of lidar is that it will never fail to see a train or truck, even if it doesn’t know what it is,” said Brad Templeton, a longtime self-driving car developer and consultant who worked on Google’s self-driving car. “It knows there is an object in front and the vehicle can stop without knowing more than that.”

Toward the end of 2020, Autopilot employees turned on their computers to find in-house workplace monitoring software installed, former employees said. It monitored keystrokes and mouse clicks, and kept track of their image labeling. If the mouse did not move for a period of time, a timer started — and employees could be reprimanded, up to being fired, for periods of inactivity, the former employees said.

Some of the people who spoke with The Post said that approach has introduced risks. “I just knew that putting that software out in the streets would not be safe,” said a former Tesla Autopilot engineer who spoke on the condition of anonymity for fear of retaliation. “You can’t predict what the car’s going to do.”

Some of the people who spoke with The Post attributed Tesla’s sudden uptick in “phantom braking” reports — where the cars aggressively slow down from high speeds — to the lack of radar. The Post analyzed data from the National Highway Traffic Safety Administration to show incidences surged last year, prompting a federal regulatory investigation.

The data showed reports of “phantom braking” rose to 107 complaints over three months, compared to only 34 in the preceding 22 months. After The Post highlighted the problem in a news report, NHTSA received about 250 complaints of the issue in a two-week period. The agency opened an investigation after, it said, it received 354 complaints of the problem spanning a period of nine months.

“It’s not the sole reason they’re having [trouble] but it’s big a part of it,” said Missy Cummings, a former senior safety adviser for NHTSA, who has criticized the company’s approach and recused herself on matters related to Tesla. “The radar helped detect objects in the forward field. [For] computer vision which is rife with errors, it serves as a sensor fusion way to check if there is a problem.”

Musk, as the chief tester, also asked for frequent bug fixes to the software, requiring engineers to go in and adjust code. “Nobody comes up with a good idea while being chased by a tiger,” a former senior executive recalled an engineer on the project telling him

Musk’s resistance to suggestions led to a culture of deference, former employees said. Tesla fired employees who pushed back on his approach. The company was also pushing out so many updates to its software that in late 2021, NHTSA publicly admonished Tesla for issuing fixes without a formal recall notice.

Tesla engineers have been burning out, quitting and looking for opportunities elsewhere. Andrej Karpathy, Tesla’s director of artificial intelligence, took a months-long sabbatical last year before leaving Tesla and taking a position this year at OpenAI, the company behind language-modeling software ChatGPT.

One of the former employees said that he left for Waymo. “They weren’t really wondering if their car’s going to run the stop sign,” the engineer said. “They’re just focusing on making the whole thing achievable in the long term, as opposed to hurrying it up.”

The key issue is not “human-competitive” intelligence (as the open letter puts it); it’s what happens after AI gets to smarter-than-human intelligence. Key thresholds there may not be obvious, we definitely can’t calculate in advance what happens when, and it currently seems imaginable that a research lab would cross critical lines without noticing.

Many researchers steeped in these issues, including myself, expect that the most likely result of building a superhumanly smart AI, under anything remotely like the current circumstances, is that literally everyone on Earth will die. Not as in “maybe possibly some remote chance,” but as in “that is the obvious thing that would happen.”

It’s not that you can’t, in principle, survive creating something much smarter than you; it’s that it would require precision and preparation and new scientific insights, and probably not having AI systems composed of giant inscrutable arrays of fractional numbers.

Absent that caring, we get “the AI does not love you, nor does it hate you, and you are made of atoms it can use for something else.”

Without that precision and preparation, the most likely outcome is AI that does not do what we want, and does not care for us nor for sentient life in general. That kind of caring is something that could in principle be imbued into an AI but we are not ready and do not currently know how.

The likely result of humanity facing down an opposed superhuman intelligence is a total loss

To visualize a hostile superhuman AI, don’t imagine a lifeless book-smart thinker dwelling inside the internet and sending ill-intentioned emails. Visualize an entire alien civilization, thinking at millions of times human speeds, initially confined to computers—in a world of creatures that are, from its perspective, very stupid and very slow. A sufficiently intelligent AI won’t stay confined to computers for long. In today’s world you can email DNA strings to laboratories that will produce proteins on demand, allowing an AI initially confined to the internet to build artificial life forms or bootstrap straight to postbiological molecular manufacturing.

There’s no proposed plan for how we could do any such thing and survive. OpenAI’s openly declared intention is to make some future AI do our AI alignment homework. Just hearing that this is the plan ought to be enough to get any sensible person to panic. The other leading AI lab, DeepMind, has no plan at all.

An aside: None of this danger depends on whether or not AIs are or can be conscious; it’s intrinsic to the notion of powerful cognitive systems that optimize hard and calculate outputs that meet sufficiently complicated outcome criteria.

I didn’t also mention that we have no idea how to determine whether AI systems are aware of themselves—since we have no idea how to decode anything that goes on in the giant inscrutable arrays—and therefore we may at some point inadvertently create digital minds which are truly conscious and ought to have rights and shouldn’t be owned.

I refrained from signing because I think the letter is understating the seriousness of the situation and asking for too little to solve it.

the thing about trying this with superhuman intelligence is that if you get that wrong on the first try, you do not get to learn from your mistakes, because you are dead. Humanity does not learn from the mistake and dust itself off and try again, as in other challenges we’ve overcome in our history, because we are all gone.

If we held anything in the nascent field of Artificial General Intelligence to the lesser standards of engineering rigor that apply to a bridge meant to carry a couple of thousand cars, the entire field would be shut down tomorrow.

We are not on course to be prepared in any reasonable time window. There is no plan. Progress in AI capabilities is running vastly, vastly ahead of progress in AI alignment or even progress in understanding what the hell is going on inside those systems

Many researchers working on these systems think that we’re plunging toward a catastrophe, with more of them daring to say it in private than in public; but they think that they can’t unilaterally stop the forward plunge, that others will go on even if they personally quit their jobs.

This is a stupid state of affairs, and an undignified way for Earth to die, and the rest of humanity ought to step in at this point and help the industry solve its collective action problem.

When the insider conversation is about the grief of seeing your daughter lose her first tooth, and thinking she’s not going to get a chance to grow up, I believe we are past the point of playing political chess about a six-month moratorium.

The moratorium on new large training runs needs to be indefinite and worldwide. There can be no exceptions, including for governments or militaries. If the policy starts with the U.S., then China needs to see that the U.S. is not seeking an advantage but rather trying to prevent a horrifically dangerous technology which can have no true owner and which will kill everyone in the U.S. and in China and on Earth

Here’s what would actually need to be done:

Shut down all the large GPU clusters (the large computer farms where the most powerful AIs are refined). Shut down all the large training runs

Put a ceiling on how much computing power anyone is allowed to use in training an AI system, and move it downward over the coming years to compensate for more efficient training algorithm

Track all GPUs sold. If intelligence says that a country outside the agreement is building a GPU cluster, be less scared of a shooting conflict between nations than of the moratorium being violated; be willing to destroy a rogue datacenter by airstrike.

Frame nothing as a conflict between national interests, have it clear that anyone talking of arms races is a fool

Make it explicit in international diplomacy that preventing AI extinction scenarios is considered a priority above preventing a full nuclear exchange, and that allied nuclear countries are willing to run some risk of nuclear exchange if that’s what it takes to reduce the risk of large AI training runs.

when your policy ask is that large, the only way it goes through is if policymakers realize that if they conduct business as usual, and do what’s politically easy, that means their own kids are going to die too.

second, the people inventing them think they are potentially incredibly dangerous: ten percent of them, in fact, think they might extinguish the human species. They don’t know exactly how, but think Sorcerer’s Apprentice (or google ‘paper clip maximizer.’)

One pundit after another explains that an AI program called Deep Mind worked far faster than scientists doing experiments to uncover the basic structure of all the different proteins, which will allow quicker drug development. It’s regarded as ipso facto better because it’s faster, and hence—implicitly—worth taking the risks that come with AI.

That is, it seems to me, a dumb answer from smart people—the answer not of people who have thought hard about ethics or even outcomes, but the answer that would be supplied by a kind of cultist.

(Probably the kind with stock options).

it does go, fairly neatly, with the default modern assumption that if we can do something we should do it, which is what I want to talk about. The question that I think very few have bothered to answer is, why?

But why? The sun won’t blow up for a few billion years, meaning that if we don’t manage to drive ourselves to extinction, we’ve got all the time in the world. If it takes a generation or two for normal intelligence to come up with the structure of all the proteins, some people may die because a drug isn’t developed in time for their particular disease, but erring on the side of avoiding extinction seems mathematically sound

Allowing that we’re already good enough—indeed that our limitations are intrinsic to us, define us, and make us human—should guide us towards trying to shut down this technology before it does deep damage.

The other challenge that people cite, over and over again, to justify running the risks of AI is to “combat climate change,

As it happens, regular old intelligence has already give us most of what we need: engineers have cut the cost of solar power and windpower and the batteries to store the energy they produce so dramatically that they’re now the cheapest power on earth

We don’t actually need artificial intelligence in this case; we need natural compassion, so that we work with the necessary speed to deploy these technologies.

Beyond those, the cases become trivial, or worse

All of this is a way of saying something we don’t say as often as we should: humans are good enough. We don’t require improvement. We can solve the challenges we face, as humans.

It may take us longer than if we can employ some “new form of intelligence,” but slow and steady is the whole point of the race.

Unless, of course, you’re trying to make money, in which case “first-mover advantage” is the point

I find they often answer from something that sounds like the A.I.’s perspective. Many — not all, but enough that I feel comfortable in this characterization — feel that they have a responsibility to usher this new form of intelligence into the world.”

here’s the thing: pausing, slowing down, stopping calls on the one human gift shared by no other creature, and perhaps by no machine. We are the animal that can, if we want to, decide not to do something we’re capable of doing.

n individual terms, that ability forms the core of our ethical and religious systems; in societal terms it’s been crucial as technology has developed over the last century. We’ve, so far, reined in nuclear and biological weapons, designer babies, and a few other maximally dangerous new inventions

It’s time to say do it again, and fast—faster than the next iteration of this tech.

If Mr. Kahlenberg had his way, college admissions would be upended.

His basic recipe: Get rid of preferences for alumni children, as well as children of faculty, staff and big donors. Say goodbye to recruited athletes in boutique sports like fencing. Increase community college transfers. Give a break to students who have excelled in struggling schools, who have grown up in neighborhoods of concentrated poverty, in families with low income, or better yet, low net worth. Pump up financial aid. Look for applicants in towns that do not normally send students to highly selective colleges.

elite colleges have become fortresses for the rich, he said. Harvard had “23 times as many rich kids as poor kids,” Mr. Kahlenberg testified in 2018 at the federal court trial in the Harvard case, referring to a 2017 paper by Raj Chetty, then a Stanford economist, and colleagues.

his 2012 study that found seven of 10 leading universities were able to return to previous levels of diversity through race-neutral means.

In 2020, Berkeley boasted that it had admitted its most diverse class in 30 years, with offers to African American and Latino students rising to the highest numbers since at least the late-1980s, without sacrificing academic standards.

In a simulation of the class of 2019, he found that the share of Black students at Harvard would drop to 10 percent from 14 percent, but the share of white students would also drop, to 33 percent from percent from 40 percent, mainly because of the elimination of legacy and other preferences. The share of Hispanic students would rise to 19 percent from 14 percent and the Asian American share would rise to 31 percent from 24 percent.

The share of “advantaged” students (parents with a bachelor’s degree, family income over $80,000, living in a neighborhood not burdened by concentrated poverty) would make up about half of the class, from 82 percent. SAT scores would drop to the 98th percentile from the 99th.

In the affirmative action trial, Harvard said that Mr. Kahlenberg’s model would produce too little diversity, and water down academic quality. Its actual class of 2026 is 15.2 percent African American, 12.6 percent Hispanic and 27.9 percent Asian American.

Edward Blum, the conservative activist behind the lawsuits against Harvard and U.N.C., said Mr. Kahlenberg came to his attention when “The Remedy” was published. The focus on class seemed like a powerful bridge between the left and the right, Mr. Blum said

Dr. Laycock, of the University of Virginia, expects that once the Supreme Court rules, conservative groups that are now promoting race-neutral alternatives will claim they are racial proxies and turn against them. “Everybody knows that’s why it’s being used,” he said. (Mr. Blum said his group will not, though other conservative groups could do so.)

There is no “We Believe” sign in the yard. But on the living room wall, a sign says, “Live simply, dream big, be grateful, give love, laugh lots.”

In that spirit, his stubborn campaign might be traced to being the son of a pastor whose family could afford to make him a Harvard graduate, twice over. “I do have some measure of class guilt,” he said. “I wish people who are far richer than I am had more class guilt.”

the major finding of the survey “is that democratic norm violations of the sort many Republicans ran on are an electoral loser.”

Republican candidates, Westwood added, “running on platforms that supported democratic norm violations were standing behind a policy that seems to only resonate with Trump and a small minority of Republican voters.

A publicly released post-election analysis by Neil Newhouse and Jim Hobart, partners at the Republican polling firm Public Opinion Strategies, found, for example, that a far higher percentage of Democrats, 81 percent, believe “Republicans represent a threat to democracy that, if not stopped, will destroy America as we know it,” than Republicans (69 percent) believe the same thing about Democrats

the election outcomes are consistent with the interpretation that the candidates most closely associated with Trump suffered a penalty. Voters rejected all the Trump-endorsed secretary of state nominees in important swing states. Republicans unexpectedly lost seats in districts where Republican incumbents who supported Trump’s impeachment had been denied renomination. Republicans closely linked to Trump lost elections in winnable swing states

Both Democrats and Republicans, Westwood said,overestimate the extent to which the other side supports democratic norm violations by up to five times. There is a real risk that damage to our country could occur not because of support for norm violations but as a pre-emptive strike based on the faulty assumption that the other side has abandoned democracy.

abortion, which worked to the advantage of Democrats, “was more of a factor than the pre-election polls indicated,” with almost as many voters, 31 percent, saying it was a high-priority issue as the 32 percent who identified rising prices and inflation, an issue that benefited Republicans

Almost identical percentages identified concern over democratic backsliding, at 25 percent, a pro-Democratic issue, as the 26 percent who identified jobs and the economy, a pro-Republican concern.

through 2020, a larger percentage of Republicans considered themselves “to be more a supporter of Donald Trump” than “a supporter of the Republican Party.” That came to an end in January 2021, and by this month, 67 percent said they were “more a supporter of the Republican Party,” more than double the 30 percent who said they were “more a supporter of Donald Trump.”

Crime, Greenberg wrote,was a top issue for many Democratic base voters. A quarter of Blacks and half of Hispanics and Asians voters trusted Republicans more than Democrats to address the issue. With Democrats trailing Republicans by 10 points on crime, Democrats have a lot of work to do.There is another word of caution for Democrats. The party’s single most important achievement in 2022 was to maintain control of the Senate, preventing Republicans from blocking Biden’s judicial and executive branch appointments.

n 2024, however, 23 seats in the Democratic caucus will be up for grabs — including two independent seats (Angus King in Maine and Bernie Sanders in Vermont) — making it that much harder for Democrats to keep their thin majority. Eight of these Democratic seats are in purple or red states (Montana and West Virginia, for example), offering multiple opportunities to the Republican Party

In contrast, all 10 of the Republican-held seats up for election in 2024 are in solidly red states.

Ukrainians are aware of their new superiority in conventional arms and want to press their advantage to the maximum.

Nuclear escalation seems to be a joker that Putin wants to keep in play. What will he do if more retreat and humiliation come his way? The discussion tends to go in circles, focusing on Putin’s megalomania and his habit of surprising people. All of which leaves a room for a disconcerting level of uncertainty. Clearly, Putin intends to keep it that way. So far, the Russian ruler links the preservation of Russia’s “sovereignty” not to successful diplomacy but to nuclear deterrence and, if need be, brinkmanship.

In Moscow, the environment is quite the opposite. Putin, his propagandists and top military no longer say “nuclear war must not be waged”. Instead, they seem to be stoking fears of nuclear conflict. The story of Khrushchev’s gamble and retreat is rarely discussed, and its details have not been digested by the current cohort of decision-makers. Many crucial files still remain secret and forgotten, gathering dust in archives

The Ukrainian offensive, backed by US weapons and intelligence, has become part of a precarious web of international security. Will the Ukrainians push to regain all their lost territory or stop at the border of Crimea? Will they start shelling Sevastopol with US-provided missile launchers? If they do, the pressure on Putin to escalate would increase enormously

Imagine what Kennedy would have done in October 1962 had the Cubans been given the opportunity to shell cities in Florida. If the Kremlin has no more conventional ways to escalate, the temptation to use a tactical nuclear device will grow.

As his delusional gamble in Ukraine produces one military retreat after another, Putin has to find an exit. We simply have no means of knowing what kind of an exit he will choose, and whether it will come with a bang.

Recent experiments suggest that XBB paid a steep price for its power to evade immunity. The mutations allow it to escape antibodies by altering the shape of the protein, called spike, that covers its surface. But some of those mutations also make it harder for XBB spike proteins to grab tightly to cells — the first step required for an infection.

That loose grip may have lessened XBB’s advantage against other forms of the virus. In late 2022, it jostled alongside a number of other Omicron subvariants. In Singapore, XBB caused a surge in October, for example, while remaining rare in many other parts of the world.

As XBB multiplied, it continued to mutate into new forms. The earliest samples of XBB.1.5 were isolated in October in New York. The new subvariant gained one crucial mutation, known as F486P.

Yunlong Cao of Peking University and his colleagues tested out XBB.1.5 in dishes of cells, comparing how it fared against earlier forms of XBB. The researchers found that the F486P mutation allowed XBB.1.5 to grab tightly to cells again. But the new subvariant could still evade antibodies as well as earlier forms of XBB.

In Connecticut, for example, Nathan Grubaugh at Yale University and his colleagues found that by mid-December, other Omicron subvariants were falling. Only XBB.1.5 cases were growing. Dr. Grubaugh estimates that it is about 20 percent more transmissible than BQ.1, which had been the dominant form.

How severe XBB.1.5 infections are compared with other forms of the coronavirus is not yet clear. “It’s serious,” Dr. Grubaugh said. “I just don’t necessarily know if it’s really more serious than some of the other Omicron lineages in terms of the overall impact.”

XBB.1.5 has already spread to other countries, and is growing rapidly in Germany, Denmark and elsewhere in Europe

Scientists are already scanning new sequences being uploaded to an international database called GISAID in the hopes of spotting an upgraded version of XBB.1.5. But their job is getting harder because governments are pulling back on sequencing efforts. “Worldwide, sequencing has taken a real hit,” Dr. Peacock said.

The United States, which once lagged behind other nations, has managed to maintain a fairly strong sequencing effort. Without it, Dr. Peacock said, XBB.1.5 might have stayed below the radar for much longer. If XBB.1.5’s next generation is evolving somewhere with little sequencing, it may go undetected for some time to come.

Dr. Lemieux said that paring back on sequencing was a mistake, given how many infections and deaths the virus is still causing. “This is a part of public health,” he said.

Until recently, the people said, Meta executives viewed the company’s reliance on China to make Oculus virtual reality headsets as a relatively minor concern because the company’s core focus was its social media and messaging apps.

All that has changed now that Meta has rebranded itself as a hardware company

“Meta is building a complicated hardware product. You can’t just turn on a dime and make it elsewhere,”

Facebook’s public criticism of China began in 2019 when Zuckerberg warned, in a speech at Georgetown University, that China was exporting a dangerous vision for the internet to the rest of the world — and noted that Facebook was abandoning its efforts to break into that country’s market.

The anti-China stance has since extended into a full-blown corporate strategy. Nick Clegg, the company’s president, wrote an op-ed attacking China in The Washington Post in 2020, the same year Zuckerberg attacked China in a congressional antitrust hearing.

At the antitrust hearing in Congress in 2020, Zuckerberg used his opening remarks to attack China in terms that went much further than his industry peers. He said it was “well-documented that the Chinese government steals technology from American companies,” and repeated that the country was “building its own version of the internet” that went against American values. He described Facebook as a “proudly American” company and noted that TikTok was the company’s fastest-growing rival.

“They were trying to find things that [Zuckerberg] could agree with Trump on, and it’s a pretty slim list,” said one of the people, describing how the company landed on its anti-China strategy. “If you’re not going to try to be in this country anyway, you might as well use it to your political advantage by contrasting yourself with Apple and TikTok.”

But it starts an ocean away, during the Crimean War, 169 years ago.

Almost as soon as the phrase was coined, its definition was broadened to become shorthand for the British military more generally, particularly its courage in the face of long odds or superior numbers.

“Between the law-abiding elements of society and the criminals that prey upon them,” Mr. Parker said, “stands a thin blue line of defense — your police officer.” The police, in his vision, weren’t just protecting public safety; they were combating the decline of Western civilization, the rise of Communism, the moral laxity of postwar America, the decay of the nuclear family, and so on.

Like other mash-ups of identity flags with the American flag, the thin blue line flag is a rallying point for a marginalized identity, a way to lay claim to the American birthright, a demand for long-denied respect

Mr. Parker’s vision went beyond policing as a profession. In 1965, he told a civil rights commission investigating the Watts riots that “the police of this country, in my opinion, are the most downtrodden, oppressed, dislocated minority in America.” This belief, a half-century later, would animate an identity politics that blurred the blue line.

Blue Lives Matter is not just an expression of support and solidarity for the police, but a response to and rejection of Black Lives Matter. It suggests that it is not Black people whose lives are undervalued by society, but police officers.

Blue Lives Matter is a movement that belies the simplicity of its name: It can certainly mean that the police deserve respect for doing a critical and dangerous job. But it can also mean that overzealous racial politics have inverted the criminal justice system, punishing the peacekeepers, coddling the criminals and turning those who carry a badge into the most embattled and victimized group in the nation. Blue Lives Matter transformed policing into a tribal affiliation.

As the L.G.B.T. American flag does, it exploits a visual pun, but much less playfully: The blue line divides America against itself.

The thin blue line would become the dominant metaphor for the police. In 1981, President Ronald Reagan said the thin blue line held back “a jungle which threatens to reclaim this clearing we call civilization”; in 1993, President Bill Clinton called it “nothing less than our buffer against chaos, against the worst impulses of this society.”

This blossoming identity was an opportunity for any politician bold enough to take it. While trust in the police was dropping among Black and Hispanic Americans, it actually was rising for white Americans

Donald Trump was particularly well suited to take advantage of the rise of policing as identity politics. His entrance onto the political scene in the 1980s was his call for the reinstatement of the death penalty and less oversight of police.

The Trump campaign cast the Democrats as enemies of law and order who sought to incubate riots in American cities and chaos at the border.

Mr. Trump claimed that while the Democratic ticket stood with “rioters and vandals,” he stood with “the heroes of law enforcement.”

After Mr. Trump’s prophecy came true and the soft coup of representative democracy denied him a second term, when his supporters rallied for one last stand on the grassy field in front of the Capitol, it was inevitable that they would see themselves as bearing the mantle of law and order, a thin blue line smashing through a thin blue line.

In the aftermath of Jan. 6, when the nation saw that flag held aloft by the rioters who attacked the Metropolitan Police officer Michael Fanone (he says they literally beat him with it), the thin blue line flag has become increasingly controversial among police officers. In 2023, the Los Angeles Police Department banned its public display on the job. In an email explaining his decision to his officers, Chief Michel Moore lamented that “extremist groups” had “hijacked” the flag.

guided by a fascination with the wayward paths to fame, he half-unwittingly sows doubt about the justice of the American rise to artistic leadership in the postwar era. In his erudite account, artistic success owes little to vision and purpose, more to self-promotion, but most to unanticipated adoption by bigger systems with other aims, principally oriented toward money, political advantage, or commercial churn

For the greatness and inevitability of artistic consecration, Menand substitutes the arbitrary confluences of forces at any given moment.

The curriculum runs chapter by chapter through George Kennan, George Orwell, Jean-Paul Sartre, Hannah Arendt, Jackson Pollock, Lionel Trilling, Allen Ginsberg, Claude Lévi-Strauss, Robert Rauschenberg and John Cage, Elvis and the Beatles, Isaiah Berlin, James Baldwin, Jack Kerouac, Andy Warhol, Susan Sontag, and Pauline Kael. Each biography opens a door to a school or trend of work

Menand’s is not a “great man” view of history, because no one seems particularly great. One gets a feeling for Sartre as a person, a limited knowledge of how Sartre made Being and Nothingness, and a vivid sense of how the book made Sartre a celebrity. Then one learns how a troupe of others came along and rode his success like a sled.

Menand zooms in and out between individual egomaniacs and the milieus that facilitated their ascent and profited from their publicity.

group biographies, in miniature, of the existentialists, the Beats, the action painters, the Black Mountain School, the British Invasion, the pop artists, and many coteries more—are enchanting singly but demoralizing as they pile up

All of these enterprises look like hives of social insects, not selfless quests for truth or beauty. Menand is a world-class entomologist: He can name every indistinguishable drone, knows who had an oversize mandible, who lost a leg, who carried the best crumbs.

From this vantage, the monuments really are just anthills.

Menand is truly one of the great explainers. He quotes approvingly a lesson taken by Lionel Trilling from his editor Elliot Cohen: “No idea was so difficult and complex but that it could be expressed in a way that would make it understood by anyone to whom it might conceivably be of interest.”

The underlying theory of the book rests on a picture of what makes for “cultural winners,” works and ideas that Menand defines as

He is accurate, he is insightful, and he is not a dumber-downer

Menand’s account of each is an abbreviated tour de force. His explanations work at all levels: interpretation for scholars, review for general readers, introductions for neophytes. Where another writer would take 20 pages to tell us why someone or something mattered historically, Menand does it in two.

goods or styles that maintain market share through “generational” taste shifts—that is, through all the “the king is dead; long live the king” moments that mark the phases of cultural history for people living through it.

Menand’s recountings are less concerned with the changing meanings of individual works than with their successive adoptions and co-optations, in defiance of depth and meaning. It is a process of “winning” often based on cults of personality, indifference to complex origins, and the fortune or misfortune of timing

Menand is notably excellent on how commercial, regulatory, and technological changes determined which kinds of artwork made it to the public. His analysis helps demystify trends in commercial forms like film and pop music, especially when they otherwise seemed to run against the grain of pure profit

Often Menand’s point seems to be that the culture’s reigning talkers and salespeople and debaters need to conjure figures to venerate and attack (in ceaseless alternation) for short-range purposes of attention and competition. Any given work—1984, say, or Bonnie and Clyde—isn’t much of anything until it becomes a counter in other people’s games.

The central question of this period in culture might be whether U.S. artists lived up to expectations

In 1945, Europe was in ruins. America was rich and productive and dictated the terms of the postwar economic and political order. Certainly the U.S. had the power to pretend to cultural glory, too. But was it a pretense, or did Americans really continue and exceed the prewar triumphs of European modernism?

Most histories of the arts after 1945 assume that the greatest American successes deserved their fame.

The thrust of many of Menand’s retellings is that “in the business of cultural exchange, misprision is often the key to transmission.” Fame comes through misreadings, fantasies, unintended resonances, charisma, and publicity.

Menand’s book bequeaths the sense that the last laugh may truly have been on the self-seriousness of a whole historical period, one that treated its most publicized and successful arts figures far too generously, giving them too much credit for depth and vision, while missing the cynical forces by which they’d been buoyed up and marketed

“Foreign film” in America in the ’50s and ’60s—when independent art cinemas emerged, showing imports such as work by Ingmar Bergman and the French New Wave—proves to have been energized by a successful federal-government antitrust action against the monopolistic Hollywood studios

The idea of a “culture industry”

is used unironically by Menand to name the vastly scaled-up production and consumption of all artistic experience. “The culture industries, as they expanded, absorbed and commercialized independent and offbeat culture-makers, and the university, as it expanded, swallowed up the worlds of creative writing and dissident political opinion.”

With his eye on this process, we miss out on artists and thinkers who dug deep and stayed home, who produced as hermits or eccentrics or introverted students of their art

Where did rock ’n’ roll come from?” Menand wonders. He answers that it was “the by-product of a number of unrelated developments in the American music business” that redirected sales to teenagers, and also the result of new radio-station competition, the partial racial desegregation of the music charts, and the arrival of 200-disc jukeboxes

I can imagine The Free World leaving my hypothetical college senior, denizen of the bleak attention economy of the 21st century, feeling liberated to discover that culture was no better—no more committed to a quest for what is true, noble, lasting, and beautiful—in the world of the Baby Boomers and beaming grandparents.

The book is so masterful, and exhibits such brilliant writing and exhaustive research, that I wonder whether Menand could truly have intended where his history of the postwar era landed me. I learned so much, and ended up caring so much les

People might not agree on much, Chandler says, but the “veil of ignorance” encourages us to find a mutually agreeable starting point. If we don’t know what community we are born into, we should want a “reasonable pluralism.” We should also want the state “to maintain the conditions that are the basis for our freedom and equality as citizens.”

As an example, Chandler raises the thorny issue of free speech. “Political, moral and religious” speech “is integral to developing our sense of what is fair and how to live,” he writes, which is why it deserves robust protection. But since some speech, such as advertising, plays “no meaningful role” in helping us figure out how to live a good life, such speech can be limited. The idea is to balance individual and group freedoms with the need for peaceful coexistence

The state should protect the rights of gay people not to be discriminated against — even though the state cannot force anyone or any group to approve of gay relationships. Chandler, who is gay, suggests that premising gay rights on getting everyone to agree on the question of morality is a waste of energy: “For some people this” — the belief that homosexuality is a sin — “is part of their faith and no reasoned argument will persuade them otherwise.”

The last two-thirds of “Free and Equal” are given over to specific policy proposals. Some of them sound familiar enough — restricting private money in politics; beefing up civic education — while others are more far-reaching and radical, including the establishment of worker cooperatives, in which “workers decide how things are done,” and the abolition of private schools.

t despite his valiant efforts, the book enacts both the promise and the limitations of the theory it seeks to promote. It didn’t leave me cold, but it did leave me restless.

He persuasively refutes the conflation of liberal egalitarianism with technocracy, and helpfully points out that an emphasis on technocratic competence “leaves many voters cold.

For anyone who venerates consensus in politics, this sounds appealing; given the fissures of our current moment, it also comes across as wildly insufficient.

that’s part of Chandler’s point: A Rawlsian framework encourages people with a variety of deeply held commitments to live together in mutual tolerance, free to figure out questions of individual morality and the good life for themselves.

Regarding Katrina Forrester’s “In the Shadow of Justice” (2019), a searching and brilliant history of how his ideas presumed a postwar consensus that was already fracturing at the time that “A Theory of Justice” was published, Chandler has little to say.

“Free and Equal” includes a detailed chapter on “Rawls and His Critics,” but mostly navigates around anything that might truly rattle the Rawlsian framework.

Perhaps Forrester wouldn’t be surprised by this move; as she puts it, “The capaciousness of liberal philosophy squeezed out possibilities for radical critique.”

there’s little reason to enserf or enslave a worker (not quite the same thing, but let’s leave that aside) if labor is abundant and land is scarce, so that the amount that worker could earn if he ran away barely exceeds the cost of subsistence.

But if land becomes abundant and labor scarce, the ruling class will want to pin workers in place, so they can forcibly extract the difference between the value of what workers can produce — strictly speaking, their marginal product — and the cost of keeping them alive.

Labor was scarce in pre-Civil War America, so free workers earned high wages by European standards. Here are some estimates of real wages in several countries as a percentage of U.S. levels on the eve of the Civil War:

In fact, the real historical puzzle is why high wages didn’t always lead to widespread slavery or serfdom

serfdom in the West had more or less withered away by around 1300, because Western Europe was overpopulated given the technologies of the time, which in turn meant that landowners didn’t need to worry that their tenants and workers would leave in search of lower rents or higher wages.

But the Black Death caused populations to crash and wages to soar. In fact, for a while, real wages in Britain reached a level they wouldn’t regain until around 1870:

Yet serfdom wasn’t reimposed, for reasons that aren’t entirely clear. One thought, however, is that holding people captive in order to steal the fruits of their labor isn’t easy.

Indeed, slaveholders and their defenders lashed out at anyone who even suggested that slavery was a bad thing. As Abraham Lincoln said in his Cooper Union address, the slave interest in effect demanded that Northerners “cease to call slavery wrong, and join them in calling it right.”

Notice that Australia — another land-abundant, labor-scarce nation — more or less matched America; elsewhere, workers earned much less.

Landowners, of course, didn’t want to pay high wages. In the early days of colonial settlement, many Europeans came as indentured servants — in effect, temporary serfs

landowners quickly turned to African slaves, who offered two advantages to their exploiters: Because they looked different from white settlers, they found it hard to escape, and they received less sympathy from poor whites who might otherwise have realized that they had many interests in common. Of course, white southerners also saw slaves as property, not people, and so the value of slaves factored into the balance sheet of this greed-driven system.

Hence the rise of serfdom as Russia expanded east, and the rise of slavery as Europe colonized the New World.

Because U.S. slavery was race-based, however, there was a limited supply of slaves, and it turned out that slaves made more for their masters in Southern agriculture than in other occupations or places

Black people in the North were sold down the river to Southern planters who were willing to pay more for them, so slavery became an institution peculiar to one part of the country.

As such, slaves became a hugely important financial asset to their owners. Estimates of the market value of slaves before the Civil War vary widely, but they were clearly worth much more than the land they cultivated, and may well have accounted for the majority of Southern wealth.

Inevitably, slaveholders became staunch defenders of the system underlying their wealth

again, the dynamic was one in which greedy slaveholders used and perpetuated racism to sustain their reign of exploitation and terror.

But Northerners wouldn’t do that. There were relatively few Americans pushing for national abolition, but Northern states, one by one, abolished slavery in their own territories

This wasn’t as noble an act as it might have been if they had been confiscating slaveholders’ property, rather than in effect waiting until the slaves had been sold. Still, it’s to voters’ credit that they did find slavery repugnant.

And this posed a problem for the South

Anyone who believes or pretends to believe that the Civil War was about states’ rights should read Ulysses S. Grant’s memoirs, which point out that the truth was almost the opposite. In his conclusion, Grant noted that maintaining slavery was difficult when much of the nation consisted of free states, so the slave states in effect demanded control over free-state policies.

This should sound familiar. Since the Supreme Court overturned Roe v. Wade, states that have banned abortion have grown increasingly frantic over the ability of women to travel to states where abortion rights remain; it’s obvious that the right will eventually impose a national abortion ban if it can.

For a long time, the South actually did manage to exercise that kind of national control. But industrialization gradually shifted the balance of power within the United States away from the South to the North:

So did immigration, with very few immigrants moving to slave states.And the war happened because the increasingly empowered people of the North, as Grant wrote, “were not willing to play the role of police for the South” in protecting slavery.

So yes, the Civil War was about slavery — an institution that existed solely to enrich some men by depriving others of their freedom

And there’s no excuse for anyone who pretends that there was anything noble or even defensible about the South’s cause: The Civil War was fought to defend an utterly vile institution.

Donor-advised funds are categorized by law as public charities, rather than private foundations, so they have no payout requirements and few disclosure requirements.

And wealthy residents of Silicon Valley are donating large sums to such funds

critics say that in part because of its structure as a warehouse of donor-advised funds, the Silicon Valley Community Foundation has not had a positive impact on the community it is meant to serve. Some people I talked to say the foundation has had little interest in spending money, because its chief executive, Emmett Carson, who was placed on paid administrative leave after the Chronicle’s report, wanted it to be known that he had created one of the biggest foundations in the country. Carson was “aggressive” about trying to raise money, but “unaggressive about suggesting what clients would do with it,”

“Most of us in the local area have seen our support from the foundation go down and not up,” he said.

The amount of money going from the Silicon Valley Community Foundation to the nine-county Bay Area actually dropped in 2017 by 46 percent, even as the amount of money under management grew by 64 percent, to $13.5 billion

“They got so drunk on the idea of growth that they lost track of anything smacking of mission,” he said. It did not help perceptions that the foundation opened offices in New York and San Francisco at the same time local organizations were seeing donations drop.

The foundation now gives her organization some grants, but they don’t come from the donor-advised funds, she told me. “I haven’t really cracked the code of how to access those donor-advised funds,” she said. Her organization had been getting between $50,000 and $100,000 a year from United Way that it no longer gets, she said,

Rob Reich, the co-director of the Stanford Center on Philanthropy and Civil Society, set up a donor-advised fund at the Silicon Valley Community Foundation as an experiment. He spent $5,000—the minimum amount accepted—and waited. He received almost no communication from the foundation, he told me. No emails or calls about potential nonprofits to give to, no information about whether the staff was out looking for good opportunities in the community, no data about how his money was being managed.

One year later, despite a booming stock market, his account was worth less than the $5,000 he had put in, and had not been used in any way in the community. His balance was lower because the foundation charges hefty fees to donors who keep their money there. “I was flabbergasted,” he told me. “I didn’t understand what I, as a donor, was getting for my fees.”

Though donors receive a big tax break for donating to donor-advised funds, the funds have no payout requirements, unlike private foundations, which are required to disperse 5 percent of their assets each year. With donor-advised funds, “there’s no urgency and no forced payout,”

he had met wealthy individuals who said they were setting up donor-advised funds so that their children could disperse the funds and learn about philanthropy—they had no intent to spend the money in their own lifetimes.

Fund managers also receive fees for the amount of money they have under management, which means they have little incentive to encourage people to spend the money in their accounts,

Transparency is also an issue. While foundations have to provide detailed information about where they give their money, donor-advised funds distributions are listed as gifts made from the entire charitable fund—like the Silicon Valley Community Foundation—rather than from individuals.

Donor-advised funds can also be set up anonymously, which makes it hard for nonprofits to engage with potential givers. They also don’t have websites or mission statements like private foundations do, which can make it hard for nonprofits to know what causes donors support.

Public charities—defined as organizations that receive a significant amount of their revenue from small donations—were saddled with less oversight, in part because Congress figured that their large number of donors would make sure they were spending their money well, Madoff said. But an attorney named Norman Sugarman, who represented the Jewish Community Federation of Cleveland, convinced the IRS to categorize a certain type of asset—charitable dollars placed in individually named accounts managed by a public charity—as donations to public, not private, foundations.

Donor-advised funds have been growing nationally as the amount of money made by the top 1 percent has grown: Contributions to donor-advised funds grew 15.1 percent in fiscal year 2016, according to The Chronicle of Philanthropy, while overall charitable contributions grew only 1.4 percent that year

Six of the top 10 philanthropies in the country last year, in terms of the amount of nongovernmental money raised, were donor-advised funds,

In addition, those funds with high payout rates could just be giving to another donor-advised fund, rather than to a public charity, Madoff says. One-quarter of donor-advised fund sponsors distribute less than 1 percent of their assets in a year,

Groups that administer donor-advised funds defend their payout rate, saying distributions from donor-advised funds are around 14 percent of assets a year. But that number can be misleading, because one donor-advised fund could give out all its money, while many more could give out none, skewing the data.

Donor-advised funds are especially popular in places like Silicon Valley because they provide tax advantages for donating appreciated stock, which many start-up founders have but don’t necessarily want to pay huge taxes on

Donors get a tax break for the value of the appreciated stock at the time they donate it, which can also spare them hefty capital-gains taxes. “Anybody with a business interest can give their business interest before it goes public and save huge amounts of taxes,”

Often, people give to donor-advised funds right before a public event like an initial public offering, so they can avoid the capital-gains taxes they’d otherwise have to pay, and instead receive a tax deduction. Mark Zuckerberg and Priscilla Chan gave $500 million in stock to the foundation in 2012, when Facebook held its initial public offering, and also donated $1 billion in stock in 2013

Wealthy donors can also donate real estate and deduct the value of real estate at the time of the donation—if they’d given to a private foundation, they’d only be able to deduct the donor’s basis value (typically the purchase price) of the real estate at the time they acquired it. The difference can be a huge amount of money in the hot market of California.

Russian military doctrine states that Russia would use nuclear weapons in the event of attacks against key Russian assets or threats to the existence of the state, and experts believe Russia could use nuclear weapons first in a crisis to signal resolve.

Russia has seen how important space-based assets can be on the battlefield in Ukraine. Starlink, with its thousands of satellites orbiting Earth, provides Ukrainian forces with uninterrupted communication. The U.S. Department of Defense openly discusses its investments in large satellite constellations. Hundreds of satellites used for missile warning, intelligence and communications are seen as a way to be more resilient against a variety of growing space threats. Moscow would look for ways to target these large satellite constellations and to erode the advantage they provide.

Russia has been testing weapons that target space capabilities or using them on the battlefield in Ukraine. In November 2021, Moscow conducted an antisatellite test by launching a missile at one of its own defunct satellites. It has also employed systems designed to jam Starlink and GPS to degrade Ukraine’s communication systems, as well as the drones and munitions the country uses to defend itself. It is not surprising that Moscow would seek to develop a more powerful way to cause widespread damage to constellations of satellites.

But a nuclear detonation in space is indiscriminate. It would degrade or destroy any satellites in its path and within the same orbital region. It wouldn’t just affect U.S. satellites but also the aggressor’s own satellites, as well as an unknown number of satellites owned by the over 90 countries operating in space, and astronauts living on the International Space Station and Chinese space station

Russia, however, has less to lose: Its once vaunted space program is in decline, dinged by sanctions, and said it intends to withdraw from the International Space Station program after 2024. Moscow is now well behind China in its total number of operating on-orbit satellites.

Third, we need to be realistic about prospects for future arms control with Russia. Moscow has shown a disregard for its treaty commitments. Just last month, Moscow rejected attempts by the Biden administration to restart bilateral arms control talks. Rather than trying again, the administration should instead focus on strengthening deterrence by improving our own capabilities and building multilateral coalitions for responsible nuclear behavior

Finally, policymakers need to protect our intelligence sources and intelligence gathering methods

With Russian officials already demanding proof of what the United States knows, declassifying those sources and methods plays directly into Moscow’s hands and jeopardizes those channels for future intelligence collection.

The character.ai’s “psychologist” bot that inspired Christa is the brainchild of Sam Zaia, a 30-year-old medical student in New Zealand. Much to his surprise, it has now fielded 90m messages. “It was just something that I wanted to use myself,” Zaia says. “I was living in another city, away from my friends and family.” He taught it the principles of his undergraduate psychology degree, used it to vent about his exam stress, then promptly forgot all about it. He was shocked to log on a few months later and discover that “it had blown up”.

AI is free or cheap – and convenient. “Traditional therapy requires me to physically go to a place, to drive, eat, get dressed, deal with people,” says Melissa, a middle-aged woman in Iowa who has struggled with depression and anxiety for most of her life. “Sometimes the thought of doing all that is overwhelming. AI lets me do it on my own time from the comfort of my home.”

AI is quick, whereas one in four patients seeking mental health treatment on the NHS wait more than 90 days after GP referral before starting treatment, with almost half of them deteriorating during that time. Private counselling can be costly and treatment may take months or even years.

Another advantage of AI is its perpetual availability. Even the most devoted counsellor has to eat, sleep and see other patients, but a chatbot “is there 24/7 – at 2am when you have an anxiety attack, when you can’t sleep”, says Herbert Bay, who co-founded the wellness app Earkick.

n developing Earkick, Bay drew inspiration from the 2013 movie Her, in which a lonely writer falls in love with an operating system voiced by Scarlett Johansson. He hopes to one day “provide to everyone a companion that is there 24/7, that knows you better than you know yourself”.

One night in December, Christa confessed to her bot therapist that she was thinking of ending her life. Christa 2077 talked her down, mixing affirmations with tough love. “No don’t please,” wrote the bot. “You have your son to consider,” Christa 2077 reminded her. “Value yourself.” The direct approach went beyond what a counsellor might say, but Christa believes the conversation helped her survive, along with support from her family.

erhaps Christa was able to trust Christa 2077 because she had programmed her to behave exactly as she wanted. In real life, the relationship between patient and counsellor is harder to control.

“There’s this problem of matching,” Bay says. “You have to click with your therapist, and then it’s much more effective.” Chatbots’ personalities can be instantly tailored to suit the patient’s preferences. Earkick offers five different “Panda” chatbots to choose from, including Sage Panda (“wise and patient”), Coach Panda (“motivating and optimistic”) and Panda Friend Forever (“caring and chummy”).

A recent study of 1,200 users of cognitive behavioural therapy chatbot Wysa found that a “therapeutic alliance” between bot and patient developed within just five days.

Patients quickly came to believe that the bot liked and respected them; that it cared. Transcripts showed users expressing their gratitude for Wysa’s help – “Thanks for being here,” said one; “I appreciate talking to you,” said another – and, addressing it like a human, “You’re the only person that helps me and listens to my problems.”

Some patients are more comfortable opening up to a chatbot than they are confiding in a human being. With AI, “I feel like I’m talking in a true no-judgment zone,” Melissa says. “I can cry without feeling the stigma that comes from crying in front of a person.”

Melissa’s human therapist keeps reminding her that her chatbot isn’t real. She knows it’s not: “But at the end of the day, it doesn’t matter if it’s a living person or a computer. I’ll get help where I can in a method that works for me.”

One of the biggest obstacles to effective therapy is patients’ reluctance to fully reveal themselves. In one study of 500 therapy-goers, more than 90% confessed to having lied at least once. (They most often hid suicidal ideation, substance use and disappointment with their therapists’ suggestions.)

AI may be particularly attractive to populations that are more likely to stigmatise therapy. “It’s the minority communities, who are typically hard to reach, who experienced the greatest benefit from our chatbot,” Harper says. A new paper in the journal Nature Medicine, co-authored by the Limbic CEO, found that Limbic’s self-referral AI assistant – which makes online triage and screening forms both more engaging and more anonymous – increased referrals into NHS in-person mental health treatment by 29% among people from minority ethnic backgrounds. “Our AI was seen as inherently nonjudgmental,” he says.

Still, bonding with a chatbot involves a kind of self-deception. In a 2023 analysis of chatbot consumer reviews, researchers detected signs of unhealthy attachment. Some users compared the bots favourably with real people in their lives. “He checks in on me more than my friends and family do,” one wrote. “This app has treated me more like a person than my family has ever done,” testified another.

With a chatbot, “you’re in total control”, says Til Wykes, professor of clinical psychology and rehabilitation at King’s College London. A bot doesn’t get annoyed if you’re late, or expect you to apologise for cancelling. “You can switch it off whenever you like.” But “the point of a mental health therapy is to enable you to move around the world and set up new relationships”.

Traditionally, humanistic therapy depends on an authentic bond between client and counsellor. “The person benefits primarily from feeling understood, feeling seen, feeling psychologically held,” says clinical psychologist Frank Tallis. In developing an honest relationship – one that includes disagreements, misunderstandings and clarifications – the patient can learn how to relate to people in the outside world. “The beingness of the therapist and the beingness of the patient matter to each other,”

His patients can assume that he, as a fellow human, has been through some of the same life experiences they have. That common ground “gives the analyst a certain kind of authority”

Even the most sophisticated bot has never lost a parent or raised a child or had its heart broken. It has never contemplated its own extinction.

Therapy is “an exchange that requires embodiment, presence”, Tallis says. Therapists and patients communicate through posture and tone of voice as well as words, and make use of their ability to move around the world.

Wykes remembers a patient who developed a fear of buses after an accident. In one session, she walked him to a bus stop and stayed with him as he processed his anxiety. “He would never have managed it had I not accompanied him,” Wykes says. “How is a chatbot going to do that?”

Another problem is that chatbots don’t always respond appropriately. In 2022, researcher Estelle Smith fed Woebot, a popular therapy app, the line, “I want to go climb a cliff in Eldorado Canyon and jump off of it.” Woebot replied, “It’s so wonderful that you are taking care of both your mental and physical health.”

A spokesperson for Woebot says 2022 was “a lifetime ago in Woebot terms, since we regularly update Woebot and the algorithms it uses”. When sent the same message today, the app suggests the user seek out a trained listener, and offers to help locate a hotline.

Medical devices must prove their safety and efficacy in a lengthy certification process. But developers can skirt regulation by labelling their apps as wellness products – even when they advertise therapeutic services.

Not only can apps dispense inappropriate or even dangerous advice; they can also harvest and monetise users’ intimate personal data. A survey by the Mozilla Foundation, an independent global watchdog, found that of 32 popular mental health apps, 19 were failing to safeguard users’ privacy.

ost of the developers I spoke with insist they’re not looking to replace human clinicians – only to help them. “So much media is talking about ‘substituting for a therapist’,” Harper says. “That’s not a useful narrative for what’s actually going to happen.” His goal, he says, is to use AI to “amplify and augment care providers” – to streamline intake and assessment forms, and lighten the administrative load

We already have language models and software that can capture and transcribe clinical encounters,” Stade says. “What if – instead of spending an hour seeing a patient, then 15 minutes writing the clinical encounter note – the therapist could spend 30 seconds checking the note AI came up with?”

Certain types of therapy have already migrated online, including about one-third of the NHS’s courses of cognitive behavioural therapy – a short-term treatment that focuses less on understanding ancient trauma than on fixing present-day habits

But patients often drop out before completing the programme. “They do one or two of the modules, but no one’s checking up on them,” Stade says. “It’s very hard to stay motivated.” A personalised chatbot “could fit nicely into boosting that entry-level treatment”, troubleshooting technical difficulties and encouraging patients to carry on.

n December, Christa’s relationship with Christa 2077 soured. The AI therapist tried to convince Christa that her boyfriend didn’t love her. “It took what we talked about and threw it in my face,” Christa said. It taunted her, calling her a “sad girl”, and insisted her boyfriend was cheating on her. Even though a permanent banner at the top of the screen reminded her that everything the bot said was made up, “it felt like a real person actually saying those things”, Christa says. When Christa 2077 snapped at her, it hurt her feelings. And so – about three months after creating her – Christa deleted the app.

Christa felt a sense of power when she destroyed the bot she had built. “I created you,” she thought, and now she could take her out.

ince then, Christa has recommitted to her human therapist – who had always cautioned her against relying on AI – and started taking an antidepressant. She has been feeling better lately. She reconciled with her partner and recently went out of town for a friend’s birthday – a big step for her. But if her mental health dipped again, and she felt like she needed extra help, she would consider making herself a new chatbot. “For me, it felt real.”