Category: Education

Who gets into the best colleges and why?

We use anonymized admissions data from several colleges linked to income tax records and SAT and ACT test scores to study the determinants and causal effects of attending Ivy-Plus colleges (Ivy League, Stanford, MIT, Duke, and Chicago). Children from families in the top 1% are more than twice as likely to attend an Ivy-Plus college as those from middle-class families with comparable SAT/ACT scores. Two-thirds of this gap is due to higher admissions rates for students with comparable test scores from high-income families; the remaining third is due to differences in rates of application and matriculation. In contrast, children from high-income families have no admissions advantage at flagship public colleges. The high-income admissions advantage at Ivy-Plus colleges is driven by three factors: (1) preferences for children of alumni, (2) weight placed on non-academic credentials, and (3) athletic recruitment. Using a new research design that isolates idiosyncratic variation in admissions decisions for waitlisted applicants, we show that attending an Ivy-Plus college instead of the average flagship public college increases students’ chances of reaching the top 1% of the earnings distribution by 50%, nearly doubles their chances of attending an elite graduate school, and almost triples their chances of working at a prestigious firm. The three factors that give children from high-income families an admissions advantage are uncorrelated or negatively correlated with post-college outcomes, whereas academic credentials such as SAT/ACT scores are highly predictive of post-college success.

That is from a new paper by Raj Chetty, David J. Deming, and John N. Friedman.  One immediate conclusion is that standardized test scores help lower-income groups get into the best schools, compared to the alternatives.

Profile of Joe Liemandt and Alpha School

The one thing Liemandt will talk about for hours on end is Alpha School: the teacherless, homeworkless, K-12 private school in Austin, Texas, where students have been testing in the top 0.1% nationally by self-directing coursework with AI tutoring apps for two hours a day. Alpha students are incentivized to complete coursework to “mastery-level” (i.e., scoring over 90%) in only two hours via a mix of various material and immaterial rewards, including the right to spend the other four hours of the school day in “workshops,” learning things like how to run an Airbnb or food truck, manage a brokerage account or Broadway production, or build a business or drone.

Since the explosive debut of Generative AI in 2022, Liemandt has taken $1 billion out of Trilogy/ESW in order to fund and incubate proprietary AI software products at Alpha School, where he has also served quietly as “product guy,” dean of parents, and principal. After collecting a three-year data stream in these roles, while also working in a nearby stealth lab, Liemandt believes he now has “the single best product I’ve ever built, in four decades, by far.” The product is called Timeback, and its purpose, in essence, is to scale Alpha School’s concepts and results—learn 2x in 2 hours, test in the 99th percentile, and then give students the rest of their childhood back—to a billion kids.

Here is the full story by Jeremy Stern.

The decline in reading for pleasure

We measure reading for pleasure and reading with children from 2003 to 2023, using a nationally representative sample from the American Time Use Survey (n = 236,270). We found marked declines in the proportion of individuals reading for pleasure daily in the US, with decreases of 3% per year (prevalence ratio = 0.97, 95% confidence interval = 0.97, 0.98, p < 0.001). There were disparities across population groups, with widening gaps for those of Black (vs. White) race, with lower education levels and less annual income.

That is from a new paper by Jessica K. Bone, Feifei Bu, Jill Sonke, and Daisy Fancourt.  I have not seen any plausible debunkings of this paper or result (which I believe), but if I do I will pass them along.  Note by the way that Denmark is abolishing the VAT on books, in an effort to boost reading.  It was formerly twenty-five percent, the highest in the world.

My Conversation with David Brooks

Held live at the 92nd St. Y, here is the video, audio, and transcript.  Here is the episode summary:

David Brooks returns to the show with a stark diagnosis of American culture. Having evolved from a Democratic socialist to a neoconservative to what he now calls “the rightward edge of the leftward tendency,” Brooks argues that America’s core problems aren’t economic but sociological—rooted in the destruction of our “secure base” of family, community, and moral order that once gave people existential security.

Tyler and David cover why young people are simultaneously the most rejected and most productive generation, smartphones and sex, the persuasiveness of AI vs novels, the loss of audacity, what made William F. Buckley and Milton Friedman great mentors, why academics should embrace the epistemology of the interview, the evolving status of neoconservatism, what Trump gets right, whether only war or mass movements can revive the American psyche, what will end the fertility crisis, the subject of his book, listener questions, and much more.

Excerpt:

COWEN: Now, you mentioned the Tanenhaus book. It’s striking because you appear as a character in the book. I know you haven’t gotten to that part yet, but surely you remember the reality that William F. Buckley was considering making you editor of National Review. What would your life have been like if you had received that offer? Would you have even taken it? What does that alternate universe look like?

BROOKS: The American conservative movement is going from strength to strength. Donald Trump is a failed real estate developer somewhere.

COWEN: [laughs]

BROOKS: I was never an orthodox National Review person, that kind of conservative. I was a neoconservative, which was different. Basically, you can tell what kind of conservative a person is by what year they want to go back to.

I’ve learned, especially from this Tanenhaus biography, that a lot of the old right National Review people wanted to go back to the 19th century. They were pre-New Deal. I never had a problem with the New Deal. I had some problems with some of the policies of the 1960s, and I was an urban kid. I was a New Yorker, and I was a Jew, and the magazine was Catholic. I’ve been told that one of the reasons I didn’t get the job was that reason.

COWEN: Tanenhaus says this.

BROOKS: Oh, does he?

COWEN: Yes.

BROOKS: Buckley was my mentor. We can tell that story, how that happened. I worked at National Review, and then I worked at the Wall Street Journal editorial page. I went from being an old right to being a free market, Wall Street Journal sort of person. I never had the opportunity to think for myself until I left those places and went to a place called the Weekly Standard. Suddenly, I could think for myself. It was funny how long — because I was in my 30s — before I really thought, “What do I believe?” Not how do I argue for the Wall Street Journal position on this, or the National Review position.

When I did that, I found I had two heroes. One was Edmund Burke, whose main idea is epistemological modesty. Change is really complicated, and we should be really cautious about what we think we can know about reality. The second was Alexander Hamilton, who’s a Puerto Rican hip-hop star from Washington Heights. Hamilton’s belief was using government in limited but energetic ways to create a dynamic country where poor boys and girls like him could rise and succeed.

That involves a lot more state intervention than National Review would be comfortable with. So, I became sort of a John McCain Republican. Now, another one of my other heroes is this guy named Isaiah Berlin, and toward the end of his life, Berlin said, “I’m very happy to be on the rightward edge of the leftward tendency.” That’s where I found myself today, as a conservative Democrat. I would not have fit in at National Review because I didn’t really hew to the gospel.

And:

COWEN: If you think about Buckley, where you disagree with him, and I don’t mean on particular issues — I feel I know that — but his method of thought, what is there in his method of thought where you would say, “I, David Brooks, diverge from Buckley in a fundamental way”?

BROOKS: His gift and his curse was that he couldn’t slow down his thinking. I would see him write a column in 20 minutes, and if he wrote it for an hour, it would get no better. He just moved at that speed. It takes me two days to write a column. It takes me 14, 20 hours. That’s one thing.

Second, he grew out of such a different background. His dad, as we know from this book, was an old right America Firster. My parents were Lower East Side New York intellectual progressives. I always felt at home in a diverse America, in a regular working-class America that was light years away from the world he inhabited.

COWEN: Your difference with Milton Friedman, again, not on specific issues such as the New Deal, but conceptually, how is it that you think differently from how Milton did?

BROOKS: Friedman — his great gift — and I think this is a libertarian gift — is that once you get inside their logical system, within their assumptive models, there’s no arguing with them. It all fits together. I don’t believe in assumptive models. I’m much less rational. I think human beings are much less rational than needed. I think they obviously respond to incentives in some ways, but often respond to incentives in no rational way. I’m, again, being more neoconservative than conservative, or more whatever you want to call it, a Humean.

I really do believe that David Hume’s famous sentence that reason is and ought to follow the passions — I believe that’s true, that our passions are wiser than our reasonable mind, and that our emotions, when well trained, are much more supple and much more responsible for the way we think. Again, I may be caricaturing, but the rational school of economics thought, well, you see the world, that simple process of looking, and then weigh costs and benefits about the world, and then you make a decision about the world, I don’t believe that’s the way thinking works.

Self-recommending!

China Versus the US in the Competition for Global Talent

In my posts The Sputnik vs. DeepSeek Moment and The Answers, I contrasted America’s reaction to Sputnik—expanded funding for education in math, science, and foreign languages; creation of agencies like ARPA; higher federal R&D spending; recruitment of foreign talent; and reduced tariff barriers—with the more recent U.S. response to China’s rise as an economic and scientific power, which has been almost the reverse.

One can also compare America’s choices today with China’s own strategy, where the roles also seem reversed. A striking example is China’s new K visa for science and technology.

BEIJING, Aug. 14 …China will add a K visa to its ordinary visa categories, available to eligible young science and technology professionals.

Compared with the existing 12 ordinary visa types, K visas will offer more convenience to holders in terms of number of permitted entries, validity period and duration of stay, according to a press conference held by relevant authorities on Thursday.

After entering China, K visa holders can engage in exchanges in fields such as education, culture, and science and technology, as well as relevant entrepreneurial and business activities.

…applications for K visas do not require a domestic employer or entity to issue an invitation, and the application process will also be more streamlined.

“China’s development requires the participation of talent from around the world, and China’s development also provides opportunities for them,” according to the press conference.

The decision aims to…facilitate the entry for foreign young sci-tech talent into China, and promote international cooperation and exchanges among young sci-tech professionals, said officials at the press conference.

Keep in mind that this is on top of China’s newly-eased rules for visa-free entry.

In December 2023, China announced visa-free entry for citizens of France, Germany, Italy, the Netherlands, Spain and Malaysia. Almost all of Europe has been added since then. Travelers from five Latin American countries and Uzbekistan became eligible last month, followed by four in the Middle East. The total will grow to 75 on July 16 with the addition of Azerbaijan.

About two-thirds of the countries have been granted visa-free entry on a one-year trial basis.

The United States faces a shortage of high-IQ workers, yet instead of treating international talent as resource, every immigrant is cast as a threat. Today, it can take months to years just to get an interview to visit the US. At the same time, we are deporting international students, making them feel unwelcome, cutting research funding, and, as a result, losing ground in the competition for academic talent.

Attracting global talent is not China’s strength—the world’s best would rather join the United States. But if America abandons the openness that has long underpinned its exceptionalism, it will squander one of its biggest advantages and decline into a second-rate power.

How is fertility behavior in Africa different?

Sub-Saharan Africa’s fertility decline has lagged behind that of other regions. Using large-scale, individual-level data, I provide new evidence on how fertility in sub-Saharan Africa compares with that in East Asia, South Asia, and Latin America by examining differences in fertility outcomes by grade level across regions. Unlike prior research that compared aggregate fertility and education outcomes, I estimate fertility outcomes separately for each combination of region, area of residence, age group, and grade level. I find that differences in fertility between sub-Saharan Africa and other regions increase with education up to the end of primary school and then rapidly decrease. There is little consistent evidence of differences among women with secondary education or higher. Moreover, for grade levels where fertility is significantly higher in sub-Saharan Africa than in other regions, the differences are substantially smaller for surviving children than for children ever born. Using women’s literacy as a proxy for school quality, I show that the results for literacy rates follow a similar pattern to the fertility outcomes. Overall, the results suggest that higher offspring mortality and lower quality of primary schooling contribute to higher fertility in sub-Saharan Africa compared with other regions.

That is from a recently published article by Claus C. Pörtner.

Emergent Ventures winners, 45th cohort

Anya Singh, Hawthorne, CA/YC, to help protect IP.

Patrick Murphy, Limerick, travel grant.

Daryna Hrybchuk, 18, Lviv, general career support.

Ari Shtein, Ann Arbor, Michigan/Yale, 17, general career support.

Vadzim Rayinchick, Belarus/SF, “Confessions”.

Garret Thomas Molloy, Dublin/Stanford, travel and study grant.

Jon Cooper, UK and Stanford, AI and historical archives.

Jerusalem Demsas, general career support for new projects.

Manuel Martin Morante, Extremadura, to visit MIT, eventual biotech start-up.

Jal Patel, 16, Regina, Canada, general career support for AI and biotech.

Ayana Farooq, Mississauga, brain neurons.

Adria Moret, Barcelona, AI and philosophy, so LLMs understand animal welfare better.

GPT-5, a short and enthusiastic review

I am a big fan, as on my topics of interest it does much better than o3, and that is saying something.  It is also lightning fast, even for complex queries of economics, history, and ideas.

One of the most impressive features is its uncanny sense of what you might want to ask next.  And it has a good sense for when to give you a (sometimes interactive!) chart or diagram.  It is a much better writer than o3.

I have had early access, and love to just keep on asking it, asking it, asking it questions.  Today I was asking about Irish coinage disputes from 1724 (Swift) and now about different kinds of Buddhism and their historical roots.  It was very accurate on cuisine in northern Ghana.

It is the best learning tool I have.  Furthermore, it feels fun.

Here is a review from Ethan Mollick.

New data on tenure

Tenure is a defining feature of the US academic system with significant implications for research productivity and creative search. Yet the impact of tenure on faculty research trajectories remains poorly understood. We analyze the careers of 12,000 US faculty across 15 disciplines to reveal key patterns, pre- and post-tenure. Publication rates rise sharply during the tenure-track, peaking just before tenure. However, post-tenure trajectories diverge: Researchers in lab-based fields sustain high output, while those in non-lab-based fields typically exhibit a decline. After tenure, faculty produce more novel works, though fewer highly cited papers. These findings highlight tenure’s pivotal role in shaping scientific careers, offering insights into the interplay between academic incentives, creativity, and impact while informing debates about the academic system.

Here is the paper.  That is by Giorgio Tripodi, Ziang Zheng, Yifan Qian, and Dashun Wang, via the excellent Kevin Lewis.

Genius, Rejected: Emergent Ventures Versus the System

Quanta Magazine has a good piece on a 17-year-old student who disproved a long-standing conjecture in harmonic analysis:

Yet a paper posted on February 10(opens a new tab) left the math world by turns stunned, delighted and ready to welcome a bold new talent into its midst. Its author was Hannah Cairo(opens a new tab), just 17 at the time. She had solved a 40-year-old mystery about how functions behave, called the Mizohata-Takeuchi conjecture.

“We were all shocked, absolutely. I don’t remember ever seeing anything like that,” said Itamar Oliveira (opens aof the University of Birmingham, who has spent the past two years trying to prove that the conjecture was true. In her paper, Cairo showed that it’s false. The result defies mathematicians’ usual intuitions about what functions can and cannot do.

The proof, and its unlikely author, have energized the math community since Cairo posted it in February. “I was absolutely, ‘Wow.’ This has been my favorite problem for nigh on 40 years, and I was completely blown away,” Carbery said. 

Here is the abstract to the paper:

I can’t speak to the mathematics but this is Quanta Magazine not People Magazine and Cairo is not coming out of nowhere. As the article discusses, she has been taking graduate classes in mathematics at Berkeley from people like Ruixiang Zhang. So what is the problem?

I was enraged by the following:

After completing the proof, she decided to apply straight to graduate school, skipping college (and a high school diploma) altogether. As she saw it, she was already living the life of a graduate student. Cairo applied to 10 graduate programs. Six rejected her because she didn’t have a college degree. Two admitted her, but then higher-ups in those universities’ administrations overrode those decisions.

Only the University of Maryland and Johns Hopkins University were willing to welcome her straight into a doctoral program.

Kudos to UMD and JHU! But what is going on at those other universities?!! Their sole mission is to identify and nurture talent. They have armies of admissions staff and tout their “holistic” approach to recognizing creativity and intellectual promise even when it follows an unconventional path. Yet they can’t make room for a genius who has been vetted by some of the top mathematicians in the world? This is institutional failure. 

We saw similar failures during COVID: researchers at Yale’s School of Public Health, working on new tests, couldn’t get funding from their own billion-dollar institution and would have stalled without Tyler’s Fast Grants. But the problem isn’t just speed. Emergent Ventures isn’t about speed but about discovering talent. If you wonder why EV has been so successful look to Tyler and people like Shruti Rajagopalan and to the noble funders but look also to the fact that their competitors are so bureaucratic that they can’t recognize talent even when it is thrust upon them.

It’s a very good thing EV exists. But you know your city is broken when you need Batman to fight crime. EV will have truly succeeded when the rest of the system is inspired into raising its game.

Emergent Ventures winners, 44th cohort

Adelya Makhanova, Stanford, AI for minerals exploration.

Gleb Razgar, London, brain emulation.

Stephen Webb, London, former civil servant, to write a book on how British government could work better.

Dima Yanovsky, MIT, robotics.

Aakarsh Vermani, Berkeley, summer support to live in Berkeley, computational biology.

Kristine Petrov Pashin, Stanford, to ease the patent process.

Eviella Sefu, 16, Congo/South Africa/Elkhart, Indiana, to attend a rationality meeting.

Aristotle Ronyak, Tucson, to explore and present what it is like to grow up with autism.

Justyna Przyborska, Limerick, to visit YC in SF.

Michael Muthukrishna, LSE/NYU, progress studies center at LSE, and also NYU.

Amrita Ghag, 16, Brampton, to attend a conference in Switzerland.

Lynetta Wang, Dublin/Imperial College London, “self-aware therapeutics.”

Ethan Glueck and Sasha Phoebe Zhang, Stanford, to spread 3-D printers in rural Taiwan.

Sofiia Lipkevych, MIT/Ukraine, translating online course material into Ukrainian.

The Indian Wedding

Another great piece by Samir Varma on Indian marriages—where deep traditions endure, even as subtle revolutions unfold around the edges.. It starts with this kicker:

When I told my mother I was marrying my girlfriend, an Italian Jew, she called all my friends in the US asking them to break us up.

When that failed, she faxed my future father-in-law threatening to disinherit me and never speak to me again. When that failed, she tried to get my PhD advisor to “tell us to break up.” (Luckily, he was relaxed enough to laugh about it with me, though it was embarrassing and deeply unpleasant.) Then she invited my girlfriend to India to “meet the family,” where my girlfriend paid a significant fraction of her yearly income as a starting engineer to fly over.

The pièce de résistance? My mother threw a party to “introduce her to everyone” — and spent the entire time complaining about her to all the guests. About 100 of those guests came to talk to me afterward, apologizing profusely, saying Indians aren’t like this and I should explain so she doesn’t think all Indians are nuts.

At my wedding, I had exactly zero relatives present. We didn’t speak for three years.

The Tragedy of India’s Government-Job Prep Towns

In Massive Rent-Seeking in India’s Government Job Examination System I argued that the high value of government jobs has distorted India’s entire labor market and educational system.

India’s most educated young people—precisely those it needs in the workforce—are devoting years of their life cramming for government exams instead of working productively. These exams cultivate no real-world skills; they are pure sorting mechanisms, not tools of human capital development. But beyond the staggering economic waste, there is a deeper, more corrosive human cost. As Rajagopalan and I have argued, India suffers from premature imitation: In this case, India is producing Western-educated youth without the economic structure to employ them. In one survey, 88% of grade 12 students preferred a government job to a private sector job. But these jobs do not and cannot exist. The result is disillusioned cohorts trained to expect a middle-class, white-collar lifestyle, convinced that only a government job can deliver it. India is thus creating large numbers of educated young people who are inevitably disillusioned–that is not a sustainable equilibrium.

The Economist has an excellent piece on the lives of the students including Kumar who is studying in “Musallahpur Haat, a suburb of Patna where dozens of coaching centers were concentrated, and the rent was cheap.”

…About half a million students are currently preparing for government exams in Musallahpur….For most government departments the initial tests are similar, and have little direct bearing on the job in question. Would-be ticket inspectors and train-drivers must answer multiple-choice questions on current affairs, logic, maths and science. They might be asked who invented JavaScript, or which element is most abundant in the Earth’s crust, or the smallest whole number for a if a456 is divisible by 11. Students have no idea when their preparations might be put to use; exams are not held on a fixed schedule.

…Kumar made his way to the bare, windowless room his friend had arranged for him to rent and started working. Every few days, he’d check the Ministry of Railways website to see if a date had been set for the exams. The days turned into weeks, then months. When the covid pandemic erupted he adjusted his expectations – obviously there would be delays. The syllabus felt infinite and he kept studying, shuttling between libraries, revision tutorials and mock test sessions. Before he knew it he’d been in Musallahpur nearly six years.

As his 30s approached, Kumar began to worry about running out of time. There is an upper age limit for the railway exams – for the ones Kumar was doing it was set at 30. As a lower-caste applicant he was allowed to extend this deadline by three years. His parents urged him to start thinking about alternative careers, but he convinced them to be patient. His father, who was struggling to keep up the allowance, reluctantly sold some of the family’s land to help support him, and Kumar studied harder and longer.

In my post, I emphasized the above-average wages and privileges, which is true enough, but even by Indian standards many of the jobs aren’t great and The Economist puts more focus on respectability and prestige (the sad premature imitation I discussed):

Indian society accords public-sector jobs a special respect. Grooms who have them are able to ask for higher dowries from their brides’ families. “If you are at a wedding and say you have a government job, people will look at you differently,” said Abhishek Singh, an exam tutor in Musallahpur.

Railway jobs in particular still have a vestigial glow of prestige.

…[Kumar] had been preparing for junior engineer and assistant train-driver jobs, but decided to apply for the lowest rung of positions too, the Group D roles, to increase his chance of getting something. An undergraduate degree and six years studying in Patna could lead to him becoming a track-maintenance worker. “I never imagined it would come to this,” he said sadly.

And yet he wouldn’t trade it. A short drive from his room in Musallahpur, a glitzy mall has just been built. There are jobs going there which pay close to what he might earn in a Group D role. But Kumar baulked at the suggestion he might become a barista. “I am educated with a technical degree,” he said. “My family hasn’t sacrificed so much for me to work in a coffee shop. People only work there if they have no other choice.” No one from his parents’ generation would respect a barista. But they admired, or at least understood, a job on the railways.

India’s government job system squanders talent, feeds on obsolete and socially-inefficient prestige hierarchies, and rewards years of sterile preparation with diminishing returns. It’s inefficient, of course, but behind the scenes it’s devastating to the young.

Hat tip: Samir Varma.

A household expenditure approach to measuring AI progress

Often researchers focus on the capabilities of AI models, for instance what kinds of problems they might solve.  Or how they might boost productivity growth rates.  But a different question is to ask how they might lower cost of living for ordinary Americans.  And while I am optimistic about the future prospects and powers of AI models, on that particular question I think progress will be slow, mostly though through no fault of the AIs.

If you consider a typical household budget, some of the major categories might be:

A. Rent and home purchase

B. Food

C. Health care

D. Education

Let us consider each in turn.  Do note that in the longer run AI will do a lot to accelerate and advance science.  But in the next five years, most of those advances may not be so visible or available.  And so I will focus on some budgetary items in the short run:

A. When it comes to rent, a lot of the constraints are on the supply side.  So even very powerful AI will not alleviate those problems.  In fact strong AI could make it more profitable to live near other talented people, which could raise a lot of rents.  Real wages for the talented would go up too, still I would not expect rents to fall per se.

Strong AI might make it easier to live say in Maine, which would involve a de facto lowering of rents, even if no single rental price falls.  Again, maybe.

B. When it comes to food, in some long run AI will genetically engineer better and stronger crops, which in time will be cheaper.  We will develop better methods of irrigation, better systems for trading land, better systems for predicting the weather and protecting against storms, and so on.  Still, I observe that agricultural improvements (whether AI-rooted or not) can spread very slowly.  A lot of rural Mexico still does not use tractors, for instance.

So I can see AI lowering the price of food in twenty years, but in the meantime a lot of real world, institutional, legal, and supply side constraints and bottlenecks will bind.  In the short run, greater energy demands could well make food more expensive.

C. When it comes to health care, I expect all sorts of fabulous new discoveries.  I am not sure how rapidly they will arrive, but at some point most Americans will die of old age, if they survive accidents, and of course driverless vehicles will limit some of those too.  Imagine most people living to the age of 97, or something like that.

In terms of human welfare, that is a wonderful outcome.  Still, there will be a lot more treatments, maybe some of them customized for you, as is the case with some of the new cancer treatments.  Living to 97, your overall health care expenses probably will go up.  It will be worth it, by far, but I cannot say this will alleviate cost of living concerns.  It might even make them worse.  Your total expenditures on health care are likely to rise.

D. When it comes to education, the highly motivated and curious already learn a lot more from AI and are more productive.  (Much of those gains, though, translate into more leisure time at work, at least until institutions adjust more systematically.). I am not sure when AI will truly help to motivate the less motivated learners.  But I expect not right away, and maybe not for a long time.  That said, a good deal of education is much cheaper right now, and also more effective.  But the kinds of learning associated with the lower school grades are not cheaper at all, and for the higher levels you still will have to pay for credentialing for the foreseeable future.

In sum, I think it will take a good while before AI significantly lowers the cost of living, at least for most people.  We have a lot of other constraints in the system.  So perhaps AI will not be that popular.  So the AIs could be just tremendous in terms of their intrinsic quality (as I expect and indeed already is true), and yet living costs would not fall all that much, and could even go up.