Group items tagged

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.”

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.)

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.”

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.”

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.

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

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?”

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.

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.”

if this were any normal product we’d assume that people love it and are grateful to the companies that provide it to them — without charge, no less.

Turning to their own lives, 52 percent of the total sample say social media has benefited their lives, and 29 percent say it has hurt them personally.

Is it more like walkie-talkies, where hardly anyone wished they had never been invented? Or is it more like cigarettes, where smokers often say they enjoy smoking, but more than 71 percent of smokers (in one 2014 survey) regret ever starting?

So what does Gen Z really think about social media?

We recently collaborated on a nationally representative survey of 1,006 Gen Z adults (ages 18-27). We asked them online about their own social media use, about their views on the effects of social media on themselves and on society and about what kinds of reforms they’d support.

First, the number of hours spent on social media each day is astonishing. Over 60 percent of our respondents said they spend at least four hours a day, with 23 percent saying they spend seven or more hours each day using social media.

Second, our respondents recognize the harm that social media causes society, with 60 percent saying it has a negative impact (versus 32 percent who say it has a positive impact).

This is especially painful for adolescents, whose social networks have migrated, since the early 2010s, onto a few giant platforms. Nearly all American teenagers use social media regularly, and they spend an average of nearly five hours a day just on these platforms.

Although the percentage citing specific personal benefits was usually higher than those citing harms, this was less true for women and L.G.B.T.Q. respondents

For example, 37 percent of respondents said social media had a negative impact on their emotional health, with significantly more women (44 percent) than men (31 percent), and with more L.G.B.T.Q. (47 percent) than non-L.G.B.T.Q. respondents (35 percent) saying so. We have found this pattern — that social media disproportionately hurts young people from historically disadvantaged groups — in a wide array of surveys.

And even when more respondents cite more benefits than harms, that does not justify the unregulated distribution of a consumer product that is hurting — damaging, really — millions of children and young adults

Our survey shows that many Gen Z-ers see substantial dangers and costs from social media. A majority of them want better and safer platforms, and many don’t think these platforms are suitable for children

If any other consumer product was causing serious harm to more than one out of every 10 of its young users, there would be a tidal wave of state and federal legislation to ban or regulate it.

Turning to the ultimate test of regret versus gratitude: We asked respondents to tell us, for various platforms and products, if they wished that it “was never invented.”

Five items produced relatively low levels of regret: YouTube (15 percent), Netflix (17 percent), the internet itself (17 percent), messaging apps (19 percent) and the smartphone (21 percent).

We interpret these low numbers as indicating that Gen Z does not heavily regret the basic communication, storytelling and information-seeking functions of the internet.

responses were different for the main social media platforms that parents and Gen Z itself worry about most. Many more respondents wished these products had never been invented: Instagram (34 percent), Facebook (37 percent), Snapchat (43 percent), and the most regretted platforms of all: TikTok (47 percent) and X/Twitter (50 percent).

We’re not just talking about sad feelings from FOMO or social comparison. We’re talking about a range of documented risks that affect heavy users, including sleep deprivation, body image distortion, depression, anxiety, exposure to content promoting suicide and eating disorders, sexual predation and sextortion, and “problematic use,” which is the term psychologists use to describe compulsive overuse that interferes with success in other areas of life.

Forty-five percent of Gen Z-ers report that they “would not or will not allow my child to have a smartphone before reaching high school age (i.e. about 14 years old)”

57 percent support the idea that parents should restrict their child’s access to smartphones before that age

Although only 36 percent support social media bans for those under the age of 16, 69 percent support a law requiring social media companies to develop a child-safe option for users under 18.

This high level of support is true across race, gender, social class and sexual orientatio

it has important implications for the House of Representatives, which is considering just such a bill, the Kids Online Safety Act. The bill would, among other things, disable addictive product features, require tech companies to offer young users the option to use non-personalized algorithmic feeds and mandate that platforms default to the safest settings possible for accounts believed to be held by minors.

imagine if walkie-talkies were harming millions of young people. Imagine if more than a third of young people wished that walkie-talkies didn’t exist, yet still felt compelled to use them for five hours every day.

We’d insist that the manufacturers make their products safer and less addictive for kids. Social media companies must be held to the same standard: Either fix their products to ensure the safety of young users or stop providing them to children altogether.

These come, as far as I can tell, in two broad types, corresponding to the two fundamental complaints that people voice about their undergraduate experience

The first complaint is that college did not prepare them for the real world: that the whole exercise—papers, busywork, pointless requirements; siloed disciplines and abstract theory—seemed remote from anything that they actually might want to do with their lives. 

Above all, they are student-centered. Participants are enabled (and expected) to direct their education by constructing bespoke curricula out of the resources the program gives them access to. In a word, these endeavors emphasize “engagement.”

A student will identify a problem (a human need, an injustice, an instance of underrepresentation), then devise and implement a response (a physical system, a community-facing program, an art project). 

Professors were often preoccupied, with little patience for mentorship, the open-ended office-hours exploration. Classes, even in fields like philosophy, felt lifeless, impersonal, like engineering but with words instead of numbers. Worst of all were their fellow undergraduates, those climbers and careerists. “It’s hard to build your soul,” as one of my students once put it to me, “when everyone around you is trying to sell theirs.”

Not everything in the world is a problem, and to see the world as a series of problems is to limit the potential of both world and self. What problem does a song address? What problem will reading Voltaire help you solve, in any predictable way? The “problem” approach—the “engagement” approach, the save-the-world approach—leaves out, finally, what I’d call learning.

that is the second complaint that graduates tend to express: that they finished college without the feeling that they had learned anything, in this essential sense.

That there is a treasure out there—call it the Great Books or just great books, the wisdom of the ages or the best that has been thought and said—that its purpose is to activate the treasure inside them, that they had come to one of these splendid institutions (whose architecture speaks of culture, whose age gives earnest of depth) to be initiated into it, but that they had been denied, deprived. For unclear reasons, cheated.

I had students like this at Columbia and Yale. There were never a lot of them, and to judge from what’s been happening to humanities enrollments, there are fewer and fewer. (From 2013 to 2022, the number of people graduating with bachelors degrees in English fell by 36%. As a share of all degrees, it fell by 42%, to less than 1 in 60.)

They would tell me—these pilgrims, these intellectuals in embryo, these kindled souls—how hard they were finding it to get the kind of education they had come to college for.

what bothers me about this educational approach—the “problem” approach, the “STEAM” (STEM + arts) approach—is what it leaves out. It leaves out the humanities. It leaves out books. It leaves out literature and philosophy, history and art history and the history of religion. It leaves out any mode of inquiry—reflection, speculation, conversation with the past—that cannot be turned to immediate practical ends

The Catherine Project sees itself as being in the business of creating “communities of learning”; its principles include “conversation and hospitality, “simplicity [and] transparency.” Classes (called tutorials, in keeping with the practice at St. John’s) are free (BISR’s cost $335), are capped at four to six students (at BISR, the limit is 23), run for two hours a week for twelve weeks, and skew towards the canon: the Greeks and Romans, Pascal and Kierkegaard, Dante and Cervantes (the project also hosts a large number of reading groups, which address a wider range of texts). If BISR aspires to create a fairer market for academic labor—instructors keep the lion’s share of fees—the Catherine Project functions as a gift economy (though plans are to begin to offer tutors modest honoraria).

As Russell Jacoby has noted, the migration of intellectuals into universities in the decades after World War II, which he documented in The Last Intellectuals, has more recently reversed itself. The rise, or re-rise, of little magazines (Dissent, Commentary, Partisan Review then; n+1, The New Inquiry, The Point, The Drift, et al. now) is part of the same story. 

a fourth factor. If there are students who despair at the condition of the humanities on campus, there are professors who do so as well. Many of her teachers, Hitz told me, have regular ladder appointments: “We draw academics—who attend our groups as well as leading them—because the life of the mind is dying or dead in conventional institutions.” Undergraduate teaching, she added, “is a particularly hard pull,” and the Catherine Project offers faculty the chance to teach people “who actually want to learn.

I’d add, who can. Nine years ago, Stephen Greenblatt wrote: “Even the highly gifted students in my Shakespeare classes at Harvard are less likely to be touched by the subtle magic of his words than I was so many years ago or than my students were in the 1980s in Berkeley. … The problem is that their engagement with language … often seems surprisingly shallow or tepid.” By now, of course, the picture is far worse.

The response to the announcement of our pilot programs confirmed for me the existence of a large, unmet desire for text-based exploration, touching on the deepest questions, outside the confines of higher education

Applicants ranged from graduating college seniors to people in their 70s. They included teachers, artists, scientists, and doctoral students from across the disciplines; a submarine officer, a rabbinical student, an accountant, and a venture capitalist; retirees, parents of small children, and twentysomethings at the crossroads. Forms came in from India, Jordan, Brazil, and nine other foreign countries. The applicants were, as a group, tremendously impressive. If it had been possible, we would have taken many more than fifteen.

When asked why they wanted to participate, a number of them spoke about the pathologies of formal education. “We have a really damaged relationship to learning,” said one. “It should be fun, not scary”—as in, you feel that you’re supposed to know the answer, which as a student, as she noted, makes no sense

“We need opportunities for reading and exploration that lie outside the credentialing system of the modern university,” he went on, because there’s so much in the latter that cuts against “the slow way that kind of learning unfolds.”

“How one might choose to live.” For many of our applicants—and this, of course, is what the program is about, what the humanities are about—learning has, or ought to have, an existential weight.

I detected a desire to be free of forces and agendas: the university’s agenda of “relevance,” the professoriate’s agenda of political mobilization, the market’s agenda of productivity, the internet’s agenda of surveillance and addiction. In short, the whole capitalistic algorithmic ideological hairball of coerced homogeneity

The desire is to not be recruited, to not be instrumentalized, to remain (or become) an individual, to resist regression toward the mean, or meme.

That is why it’s crucial that the Matthew Strother Center has no goal—and this is true of the Catherine Project and other off-campus humanities programs, as well—beyond the pursuit of learning for its own sake.

This is freedom. When education isn’t pointed in particular directions, its possibilities are endless

The term “deep state” comes from countries like Egypt and Turkey where the security services acted for many years as a shadow government. The United States has never had a deep state in this sense, except in the fevered imaginations of the MAGA right. It does have a permanent civil service that operates at federal, state, and local levels, and it is these that have become a regular conservative punching bag.

The Loper Bright decision invalidated a rule issued by the National Marine Fisheries Service requiring Atlantic fishing boats to carry, at their own expense, inspectors judging compliance with rules against overfishing. In ruling in favor of the fishing companies, SCOTUS invalidated the Chevron precedent entirely. This decision built on the same narrative feeding the Project 2025 plan: the administrative state had grown into a monster that made decisions harming the well-being of citizens without any fundamental democratic accountability.

The second initiative was the Supreme Court’s Loper Bright v. Raimondo decision issued in late June that abolished the 1984 Chevron Deference precedent. Chevron Deference provided a rule under which the courts would defer to the expert opinions of executive branch agencies in situations where a Congressional mandate was ambiguous or unclear, and the agency position seemed reasonable.

At the heart of the conservative critique of the administrative state lies a vision of democratic government “of the people, by the people, and for the people,” in which citizens would deliberate together on policies, and would themselves be responsible for carrying them out much as one imagines occurred in the proverbial New England town hall.

The problem, however, is the extreme complexity of the tasks that modern government is expected to accomplish.

None of these functions can be performed by ordinary citizens; they must be delegated to experts whose life work centers around the complex tasks they perform.

While some local issues could be settled on a local level, modern government does things like manage the money supply, regulate giant international banks, certify the safety and efficacy of drugs, forecast weather, control air traffic, intercept and decrypt the communications of adversaries, perform employment surveys, and monitor fraud in the payment of hundreds of billions of dollars in the Social Security and Medicare programs

Substantial delegation is therefore necessary. Some conservatives believe in a Constitutional “non-delegation doctrine,” but Congress has been delegating responsibility for complex tasks ever since Treasury Secretary Alexander Hamilton was given the job of cleaning up Revolutionary War debt by the first Congress of the United States.

Nor is it the case that the people’s elected representatives have no means of monitoring and holding accountable the bureaucracy they have created. There are both ex ante and ex post methods for doing this

There are, in other words, a huge number of mechanisms by which the political layer can control the administrative layer

The problem in these cases was not, however, an out-of-control bureaucracy exerting unaccountable power over citizens. The problem was a failure by plaintiffs to make use of the specific powers—the checks and balances—that the system made available to them. The failures of the early Trump administration to get its way cited in Project 2025 were largely due to the inexperience of that administration’s political appointees.

Removal of the property qualification for voting by most U.S. states in the 1820s vastly expanded the franchise to all white men. Politicians soon discovered, as they subsequently did in other new democracies, that the easiest way to get people to the polls was to bribe them—perhaps with a bottle of bourbon, a Christmas turkey, or a job in the post office. Thus began what was known as the patronage or spoils system, under which virtually every job in the civil service was given out by a politician in return for political support

The American patronage system was hugely corrupt, and provided opportunities for state capture by big business interests like the railroads that were spreading across the country. Congress did not want to give up its patronage powers, but eventually passed the Pendleton Act in 1883 that created a U.S. Civil Service Commission and established the principle of merit as a condition for hiring and promoting bureaucrats.

it was not until the time of the First World War that a majority of federal bureaucrats were appointed under the merit system.

The fundamental problem with a new Schedule F, as noted in my previous post, is that it will return the country to the period before the Pendleton Act, when political loyalty rather than merit, skill, or knowledge will be the primary criterion for government service

It took President Trump nearly four years (and 44 cabinet secretaries) to rid his administration of seasoned professionals and replace them with loyalists like Kash Patel at Defense or Jeffrey Clark at the Justice Department. This gives us a taste for the quality of officials who are likely to come in under a revived Schedule F. The doors to patronage, incompetence, and corruption will be thrown wide open.