My full reaction to sebjenseb's new article on academia and prestige:
THE INDUSTRIAL REVOLUTION HOAX:
Huebner's finding that innovation peaked between 1840 and 1880 is wild. Every modern convenience -- phones, computers, refrigeration, air conditioning, cars -- didn't become accessible to consumers until the 1950s. These conclusions are radically contradictory to "common sense,” but I’m open to them.
Murray's estimate that "rate of accomplishment" peaked in 1550 is even more astounding. It debunks the entire concept of the "industrial revolution." According to Murray, we are less innovative than Europeans in the 1400s. If peak invention occurred in 1550, and everything since then has been downhill, this challenges a few assumptions:
1. That the dysgenic selection of the 18th century is responsible for a decline in innovation (the decline began two centuries earlier);
2. That democracy or liberalism is at fault (again, two centuries late)
3. That warfare is bad for innovation (there’s no correlation, maybe even a positive correlation between military deaths per capita and innovation)
THE EVERYTHING CRISIS:
It’s not just academia that is becoming less prestigious. Everything is becoming less prestigious: corporations are less productive, businesses are less efficient, anyone can call themselves a CEO, anyone can be a college graduate, it's all worthless. It's an "everything crisis" of prestige. As a result, people are ghettoizing into political theologies where prestige is revived within cloistered spaces. E-celebs are prestigious within their internet circles. This is more volatile and less stable than the traditional university system, but so was the "prestige" of a 15th century inventor.
WHAT IS PRESTIGE?
Prestige is "social recognition and acclaim." In the 16th century, inventors didn't get much recognition and acclaim. Who even knows the name of the guy who invented the pocket watch? Anyone? (Henlein, 1510) But we know Edison and Tesla because the prestige of invention started to exponentially increase as it started to affect the lives of common people. Isaac Newton was one of the first “celebrity scientists,” and the Royal Society helped increase his prestige. Everyone knows who Bill Gates, Steve Jobs, and Elon Musk are. They are famous, but they are not necessarily prestigious. It seems that innovation prestige peaked around the time of Tesla's death in the 1940s, and since then, Americans have become increasingly cynical about inventors.
Part of this might have to do with sci-fi. On the one hand, sci-fi oversells science, so that real life seems disappointing. On the other hand, it fear-mongers (Frankenstein) about the negative effects of technology. Most Americans see social media and automation as the fruit of technology, and they hate it (even as they use it and benefit from it).
THE GRADING APOCALYPSE:
The problem of incentives in grading is huge. Universities are never punished for grade inflation. Grade inflation is just as problematic as real inflation, but from the reverse. Whereas monetary inflation destroys the wages of the poor and the savings of the middle class, grade inflation destroys our trust in institutions, because it levels the playing field and destroys competence. There is absolutely no will in academia to fight this -- not even from conservatives. Conservatives are too busy fighting "wokeness" and "gender studies." No one is willing to start firing professors or holding them accountable for handing out A's like breadsticks at the Olive Garden.
AI INTERVIEWS:
Companies have always had difficulty testing employees for competence. That's why degrees are valuable -- they tell companies something about employees, and help filter out low-value candidates. But if a degree becomes worthless, then they need another way to filter employees. However, interviews take time and are expensive. Anyone who goes through a corporate hiring process understands that it's not just one manager taking one hour to hire a person. It's four rounds of interviews with ten different people. Assuming a company is paying every $100 an hour ($200k salaries), that's $4,000 just to get a person through the interview process. If you have 100 applicants, that's $400,000 just to hire a single person for a single position.
Maybe this is where AI saves the day. AI spends hours and hours putting applicants through various hoops without any human oversight or interview necessary. If the applicant satisfies the AI, then they move on to a real human interview, which hopefully is a one-and-done with their direct manager, to control for personality defects and clear mental illness. Basically, the job of AI is to construct a virtual job environment for the applicant, and force them to waste enormous amounts of time succeeding in that environment. It's a bit like Ender's Game.
BIRTH-RATES:
The graph you show is amazing. Between 1972 and 2007, the birth-rate-acceleration "slammed the brakes." What I mean by this is that during those 35 years, the decline in the birth-rate slowed dramatically, to the point that in 2007 it seemed like birth-rates may have begun to turn positive again. EDIT: I originally misread this as USA data, but it's actually global data.
Seb says: "my own prediction is that fertility will stop declining around 2030, as the drop is largely driven by the increase in smartphones and internet use. In addition to this, I predict that culture in societies with consistently low fertility will shift slowly towards high fertility norms"
I am more cynical here, because I believe low fertility is being driven by biological factors, both in terms of psychology and biological fertility. As people become more mentally ill, LGBTQ+, neurotic, fearful, and risk averse, they want less kids. And then even if they want kids, they can't have them because their reproductive organs aren't functioning properly. Even the Amish are seeing a decline in fertility. This doesn't rely upon mutational load theory, because it can be a function of microplastics and forever chemicals. No cultural revolution can fight this, at least not without biologically changing our environment. And even if we did get rid of all the microplastics, the effects are epigenetic, and therefore heritable. You can't reverse these effects without inducing a massive genetic bottleneck, which takes time to bear fruit "naturally" (meaning, no eugenics).
Yes, eventually humans will evolve to cope with microplastics by overcompensating with ridiculous amounts of testosterone production and excessive sperm counts. But that type of evolution would take several generations, and I don't expect to see the selective effects becoming visible in this century. But that's an assertion that needs to be proven by mathematical models. The problem is that most of these models contain assertions that don't account for the heritability of fertility, and instead rely on ad hoc sociological models in an area where we have no historical data, because fertility rates have never been this low before.
>What this shows us is that, despite record immigration between 2007 and 2024, birth rates have declined at consistently faster rates. There was no "Trump baby boom." The economic recovery after 2014 didn't help. The South Korean fertility projections seem like wishful thinking.
This is a global chart, measuring fertility_year1 - fertility_year0
Deepleftanalysis what do you think is causing LGBT increases? People have fewer siblings now than ever before so birth order effect is out of the window. Maybe a lot of people were always LGBT and only now are expressing themselves, but if that was the case, wouldn't there have been way more pressure for decriminalisation historically.
My best guess would be that human sexuality is somewhat fluid/is a latent desire and now that people have their Maslow hierarchy of need taken care of, they can spend more time worrying about such stuff while their ancestors were thinking more about getting the next meal.
I find it a big enigma and I think the main question would be whether these tendencies have always been there (but just very dormant) or have sprung up at such a scale very recently.
> My best guess would be that human sexuality is somewhat fluid/is a latent desire
Related to this point, I think one thing we might have discovered in modernity is that men and women don't really like each other all that much. I bring this point/hypothesis up because my world model was always built off of an assumption that men and women are generally good friends with each other And that getting them to love each other is a just a matter of some pushes and nudges here and there.
Now I kind of more think that getting men and women to love each other is a matter of setting up infrastructure so that they can get along in spite of overall mutual dislike. That is a much more depressing and hostile world to live in.
How much of this hostility is naturally extant versus induced by lack of privacy from the internet where ee all see our worst qualities? Who knows? But it does bum me out...
I did a third-level degree in the late 20th century and not to have done so was literally unthinkable for someone of my social class and IQ. None of my grandparents and only one of my parents had gone to third level mainly for economic reasons and there was huge social pressure. All of my fourteen first cousins have or will complete their level.
I keep a casual eye on local academic output and it seems to have gone downhill over the last 25 years since I was engaged . Large parts of the academy are just doing grievance studies and over in STEM everyone is in a micro-specialism. No one has Big Ideas anymore.
For my own kids - who will leave school over the course of the next decade - I’m not so sure. I can see merit in them doing something technical/vocational like medicine or engineering. But for much of the humanities I really see very little point given that everyone (even the professors) will rely on LLM-generated content and the curriculum is actively getting worse.
I would trust a successful businessman over a random person with a PhD in gender studies.Even if a plumber has a slightly lower iq that a colonial oppression studies scholar,I would still favor him to better complete 99% of given tasks,considering lower neuroticism and narcissistic that the activist upper cllass is cultivated with.If universities grade people on leftists instead of skills/competence/test scores why should we assign value to their judgements
Seb I think that fertility will continue declining throughout this century. Here's why
1) Africa's TFR will continue decreasing. Currently the only continent with an above replacement TFR is Africa. It stands at around 4, while everywhere else is <2.1. African TFR falls slowly, but it is falling and it is fully reasonable to believe that it will reach ~2. Therefore, the fall in African TFR will result in lower TFR globally, even if other regions stagnate/have modest increases in fertility.
2) There appears to be no 'fertility floor'. South Korea et al show us that there is no TFR that humans cannot go below. There is not some hard-wired level at which people must reproduce at. While the West experienced a respectable TFR boost in the pre-2008 boom, there seems to be a sharp decline post-Covid. Fertility also appears to be cultural/mimetic and very fluid so people getting used to smartphone use will probably not have an effect.
3) I am not a scientist but I do not believe that fertility is (moderately) genetically heritable because every ethnic group and every religion demonstrates similar fertility under similar conditions. If it were highly genetically heritable, we might see the emergence of groups who are unaffected by liberalism/technology, but we do not see this.
For sure. I remember reading that the optimal TFR for developed countries is around 1.8, so sub 2.1 rates are not something to worry about atm, but of course a TFR of 0.7 would be.
I looked at your link. Well done on such diligence; although I would disagree with the idea that some genetically ultra high fertility population group would emerge.
TFR for developed countries I find very difficult to predict, however with Africa, I could bet a lot of money that its TFR will continue decreasing.
Does it matter? Government funded orgs and corporate need to see that price of paper to get their payroll. Foreign influx will remain steady because they take over the useful stuff and don't have to build their own system. Change has to come from the grassroots or this process will continue.
> To some extent I cannot fault them for thinking this; my own prediction is that fertility will stop declining around 2030, as the drop is largely driven by the increase in smartphones and internet use.
The Korean plot shows huge fertility drop between 1950 and 1990, a period without either smartphones or internet available...
> Admittedly, in the old days, people predicted that the internet would make higher education obsolete because anybody could look up anything. Which did not exactly pan out.
It kind of did. Although people do go to university and study computer science the reality is that most people in our industry are primarily self-taught. You can get a high paid prestigious job in the software industry with either no degree at all or a degree in an irrelevant subject, purely on the back of self teaching, which itself is 95% just looking things up on the internet.
This hasn't happened in other professions but this is largely thanks to occupational licensing requirements preventing it from happening, rather than any failure of the prediction or internet itself. The best scientific and medical analysis during COVID reliably came from people who were not public health or medical professionals, mostly because the former are fools and the latter had the sword of license revocation forcing them to agree with the fools. It wasn't hard for "amateurs" who just looked things up to beat their predictive accuracy.
LLMs will destroy academia, not because LLMs themselves are particularly special (though they are), but because the prestige of academia is already in free-fall and thus people's minds are now open to the possibility that there's no way to stop cheating except by forcing offline essay writing in a closely supervised room. Schools will do this (more), universities won't, and thus everyone will intuit that everyone else is probably cheating because why wouldn't you? Professors can't tell and employers don't care.
The decline in innovation outside of computing does feel real. It's hard to think of examples of big quality-of-life improvements that come from other fields in the past 20 years (don't say mRNA vaccines please, they made COVID worse not better). My suspicion is that this is related to the decline of institutions more generally. Computing is unusual in that it has low startup costs and ready access to capital, so it's easy to leave declining institutions and create fresh companies. This is so much the case that "startup" is essentially synonymous with "computing startup". Other industries have very large high barriers to entry, so innovation has to take place in the context of existing institutions. If they decay then innovation slows down or becomes outright impossible.
As to why we're seeing across-the-board institutional decay, I think it can be traced back to the left's strategic switch in the 1970s-era away from trying to change society through anti-capitalist revolution / social democratic takeover, towards the broader "social justice" movement, which necessarily entails capturing companies and NGOs as much as or more than governments. Because the left believes in equity over excellence, any institution so captured becomes ever less competent and ever more willing to punish risk taking innovators. Innovation slowly dies as people learn it's not worth the risks, whilst also finding there's nowhere else to go.
The engine of innovation therefore won't be restarted by AGI even if it were to be achieved. Captured institutions ignore their existing humans who innovate, why would they care about humans who innovative better using AI? The only thing that can restore the march of progress is a comprehensive defeat of leftist ideology, e.g. a rollback of DEI and occupational licensing regimes.
Microplastics! You were making sense then you fell off the rail with that deus ex machina. I fear that unexpected effects of the continuing RNA Jabs will prove more deleterious to humanity's future than microplastics. But nice try.
My two younger kids went to a major state university - the University of Washington, in the US. We never considered the ivy's even though my daughter certainly would have had no problem getting into them. But they graduated with their Masters by age 21 with no or minimal student debt. My daughter with a MS in Structural Engineering and my son with a MS - MIS - Data Security. The major state universities are the primary producers of engineering and STEM talent in the economy - and the employers know it. Compared to what I saw as a teaching assistant in Mechanical Engineering 40 years ago, I don't think the expectations in engineering are significanly less than they were half a century ago. Yes we have more tools now - but the problems assigned are correspondingly harder.
> The method by which students are graded is up to the professor, and professors don’t really have much of an incentive to make those systems effective.
The incentive is literally the title of this article.
The professors don't care about the average prestige of their school, or even their entire profession, more than they care about not appearing racist, for which they are treated the more the demographics desparities of their classrooms are apparent.
The only way to do this is to degrade the standards in some form or another.
AI is a good argument, because AI does have the potential to dramatically disrupt academia, and there is a possibility that AI will make all human cognitive work obsolete. Who knows what that kind of world will look like?
As for Republicans' distrust of institutions, it's unclear whether this matters in the long-term or by how much it matters, since conservatives almost always lose and Republicans are increasingly stupid. There is also the possibility that Republicans issues with low trust are just part of a transitional period, that populism will fade and that American politics will see a new equilibrium with the two parties coming closer together again.
As for innovation and productivity, it's unclear whether innovation has slowed down, some of the arguments used in favor of this idea are really unconvincing, but even if innovation has slowed down it's possible that we are about to experience very significant innovations in the fields of medicine and tech, see Ozempic, embryo selection research, gene editing research and anti-aging research for the former and AI and new products like AR glasses for the latter. You also have the fact that solar power and batteries are getting very cheap, and cheaper energy could be the key to significantly increasing productivity. Productivity in the US has increased in recent years by the way.
Demographic aging is a problem for all aspects of society, but the US has the ability to continue importing high IQ people in relatively large numbers, and it is very plausible that eugenic technologies coming in the next few decades could cause very significant IQ gains. As for academia becoming less selective, that's true, but people probably already understand that and take it into account. People understand that different fields and different universities have different selection processes, and that there can be differences in the quality of individuals' work even for individuals who obtain the same degrees from the same universities.
A.I. will never make all human cognitive work obsolete. The smartest humans will still have the advantage of being more clever in what they ask of the computer, and the computer still shows no sign of having a will of its own for which to drive its own cognitive labors.
My full reaction to sebjenseb's new article on academia and prestige:
THE INDUSTRIAL REVOLUTION HOAX:
Huebner's finding that innovation peaked between 1840 and 1880 is wild. Every modern convenience -- phones, computers, refrigeration, air conditioning, cars -- didn't become accessible to consumers until the 1950s. These conclusions are radically contradictory to "common sense,” but I’m open to them.
Murray's estimate that "rate of accomplishment" peaked in 1550 is even more astounding. It debunks the entire concept of the "industrial revolution." According to Murray, we are less innovative than Europeans in the 1400s. If peak invention occurred in 1550, and everything since then has been downhill, this challenges a few assumptions:
1. That the dysgenic selection of the 18th century is responsible for a decline in innovation (the decline began two centuries earlier);
2. That democracy or liberalism is at fault (again, two centuries late)
3. That warfare is bad for innovation (there’s no correlation, maybe even a positive correlation between military deaths per capita and innovation)
THE EVERYTHING CRISIS:
It’s not just academia that is becoming less prestigious. Everything is becoming less prestigious: corporations are less productive, businesses are less efficient, anyone can call themselves a CEO, anyone can be a college graduate, it's all worthless. It's an "everything crisis" of prestige. As a result, people are ghettoizing into political theologies where prestige is revived within cloistered spaces. E-celebs are prestigious within their internet circles. This is more volatile and less stable than the traditional university system, but so was the "prestige" of a 15th century inventor.
WHAT IS PRESTIGE?
Prestige is "social recognition and acclaim." In the 16th century, inventors didn't get much recognition and acclaim. Who even knows the name of the guy who invented the pocket watch? Anyone? (Henlein, 1510) But we know Edison and Tesla because the prestige of invention started to exponentially increase as it started to affect the lives of common people. Isaac Newton was one of the first “celebrity scientists,” and the Royal Society helped increase his prestige. Everyone knows who Bill Gates, Steve Jobs, and Elon Musk are. They are famous, but they are not necessarily prestigious. It seems that innovation prestige peaked around the time of Tesla's death in the 1940s, and since then, Americans have become increasingly cynical about inventors.
Part of this might have to do with sci-fi. On the one hand, sci-fi oversells science, so that real life seems disappointing. On the other hand, it fear-mongers (Frankenstein) about the negative effects of technology. Most Americans see social media and automation as the fruit of technology, and they hate it (even as they use it and benefit from it).
THE GRADING APOCALYPSE:
The problem of incentives in grading is huge. Universities are never punished for grade inflation. Grade inflation is just as problematic as real inflation, but from the reverse. Whereas monetary inflation destroys the wages of the poor and the savings of the middle class, grade inflation destroys our trust in institutions, because it levels the playing field and destroys competence. There is absolutely no will in academia to fight this -- not even from conservatives. Conservatives are too busy fighting "wokeness" and "gender studies." No one is willing to start firing professors or holding them accountable for handing out A's like breadsticks at the Olive Garden.
AI INTERVIEWS:
Companies have always had difficulty testing employees for competence. That's why degrees are valuable -- they tell companies something about employees, and help filter out low-value candidates. But if a degree becomes worthless, then they need another way to filter employees. However, interviews take time and are expensive. Anyone who goes through a corporate hiring process understands that it's not just one manager taking one hour to hire a person. It's four rounds of interviews with ten different people. Assuming a company is paying every $100 an hour ($200k salaries), that's $4,000 just to get a person through the interview process. If you have 100 applicants, that's $400,000 just to hire a single person for a single position.
Maybe this is where AI saves the day. AI spends hours and hours putting applicants through various hoops without any human oversight or interview necessary. If the applicant satisfies the AI, then they move on to a real human interview, which hopefully is a one-and-done with their direct manager, to control for personality defects and clear mental illness. Basically, the job of AI is to construct a virtual job environment for the applicant, and force them to waste enormous amounts of time succeeding in that environment. It's a bit like Ender's Game.
BIRTH-RATES:
The graph you show is amazing. Between 1972 and 2007, the birth-rate-acceleration "slammed the brakes." What I mean by this is that during those 35 years, the decline in the birth-rate slowed dramatically, to the point that in 2007 it seemed like birth-rates may have begun to turn positive again. EDIT: I originally misread this as USA data, but it's actually global data.
Seb says: "my own prediction is that fertility will stop declining around 2030, as the drop is largely driven by the increase in smartphones and internet use. In addition to this, I predict that culture in societies with consistently low fertility will shift slowly towards high fertility norms"
I am more cynical here, because I believe low fertility is being driven by biological factors, both in terms of psychology and biological fertility. As people become more mentally ill, LGBTQ+, neurotic, fearful, and risk averse, they want less kids. And then even if they want kids, they can't have them because their reproductive organs aren't functioning properly. Even the Amish are seeing a decline in fertility. This doesn't rely upon mutational load theory, because it can be a function of microplastics and forever chemicals. No cultural revolution can fight this, at least not without biologically changing our environment. And even if we did get rid of all the microplastics, the effects are epigenetic, and therefore heritable. You can't reverse these effects without inducing a massive genetic bottleneck, which takes time to bear fruit "naturally" (meaning, no eugenics).
Yes, eventually humans will evolve to cope with microplastics by overcompensating with ridiculous amounts of testosterone production and excessive sperm counts. But that type of evolution would take several generations, and I don't expect to see the selective effects becoming visible in this century. But that's an assertion that needs to be proven by mathematical models. The problem is that most of these models contain assertions that don't account for the heritability of fertility, and instead rely on ad hoc sociological models in an area where we have no historical data, because fertility rates have never been this low before.
>What this shows us is that, despite record immigration between 2007 and 2024, birth rates have declined at consistently faster rates. There was no "Trump baby boom." The economic recovery after 2014 didn't help. The South Korean fertility projections seem like wishful thinking.
This is a global chart, measuring fertility_year1 - fertility_year0
Deepleftanalysis what do you think is causing LGBT increases? People have fewer siblings now than ever before so birth order effect is out of the window. Maybe a lot of people were always LGBT and only now are expressing themselves, but if that was the case, wouldn't there have been way more pressure for decriminalisation historically.
My best guess would be that human sexuality is somewhat fluid/is a latent desire and now that people have their Maslow hierarchy of need taken care of, they can spend more time worrying about such stuff while their ancestors were thinking more about getting the next meal.
I find it a big enigma and I think the main question would be whether these tendencies have always been there (but just very dormant) or have sprung up at such a scale very recently.
> My best guess would be that human sexuality is somewhat fluid/is a latent desire
Related to this point, I think one thing we might have discovered in modernity is that men and women don't really like each other all that much. I bring this point/hypothesis up because my world model was always built off of an assumption that men and women are generally good friends with each other And that getting them to love each other is a just a matter of some pushes and nudges here and there.
Now I kind of more think that getting men and women to love each other is a matter of setting up infrastructure so that they can get along in spite of overall mutual dislike. That is a much more depressing and hostile world to live in.
How much of this hostility is naturally extant versus induced by lack of privacy from the internet where ee all see our worst qualities? Who knows? But it does bum me out...
> People have fewer siblings now than ever before so birth order effect is out of the window.
How about the "sole child with no other-sex siblings to help them understand the other sex and sex differences" effect?
I did a third-level degree in the late 20th century and not to have done so was literally unthinkable for someone of my social class and IQ. None of my grandparents and only one of my parents had gone to third level mainly for economic reasons and there was huge social pressure. All of my fourteen first cousins have or will complete their level.
I keep a casual eye on local academic output and it seems to have gone downhill over the last 25 years since I was engaged . Large parts of the academy are just doing grievance studies and over in STEM everyone is in a micro-specialism. No one has Big Ideas anymore.
For my own kids - who will leave school over the course of the next decade - I’m not so sure. I can see merit in them doing something technical/vocational like medicine or engineering. But for much of the humanities I really see very little point given that everyone (even the professors) will rely on LLM-generated content and the curriculum is actively getting worse.
I would trust a successful businessman over a random person with a PhD in gender studies.Even if a plumber has a slightly lower iq that a colonial oppression studies scholar,I would still favor him to better complete 99% of given tasks,considering lower neuroticism and narcissistic that the activist upper cllass is cultivated with.If universities grade people on leftists instead of skills/competence/test scores why should we assign value to their judgements
Seb I think that fertility will continue declining throughout this century. Here's why
1) Africa's TFR will continue decreasing. Currently the only continent with an above replacement TFR is Africa. It stands at around 4, while everywhere else is <2.1. African TFR falls slowly, but it is falling and it is fully reasonable to believe that it will reach ~2. Therefore, the fall in African TFR will result in lower TFR globally, even if other regions stagnate/have modest increases in fertility.
2) There appears to be no 'fertility floor'. South Korea et al show us that there is no TFR that humans cannot go below. There is not some hard-wired level at which people must reproduce at. While the West experienced a respectable TFR boost in the pre-2008 boom, there seems to be a sharp decline post-Covid. Fertility also appears to be cultural/mimetic and very fluid so people getting used to smartphone use will probably not have an effect.
3) I am not a scientist but I do not believe that fertility is (moderately) genetically heritable because every ethnic group and every religion demonstrates similar fertility under similar conditions. If it were highly genetically heritable, we might see the emergence of groups who are unaffected by liberalism/technology, but we do not see this.
That might be true, but it's hard for anyone to know for sure since there's so many different variables involved in what could happen.
Regardless, it's not necessarily a bad thing to have some declining fertility. Overpopulation is still a major potential risk, even if fertility is declining: https://zerocontradictions.net/FAQs/overpopulation-FAQs
For sure. I remember reading that the optimal TFR for developed countries is around 1.8, so sub 2.1 rates are not something to worry about atm, but of course a TFR of 0.7 would be.
I looked at your link. Well done on such diligence; although I would disagree with the idea that some genetically ultra high fertility population group would emerge.
TFR for developed countries I find very difficult to predict, however with Africa, I could bet a lot of money that its TFR will continue decreasing.
Does it matter? Government funded orgs and corporate need to see that price of paper to get their payroll. Foreign influx will remain steady because they take over the useful stuff and don't have to build their own system. Change has to come from the grassroots or this process will continue.
I am highly skeptical that innovation peaked in the mid 16th century
> To some extent I cannot fault them for thinking this; my own prediction is that fertility will stop declining around 2030, as the drop is largely driven by the increase in smartphones and internet use.
The Korean plot shows huge fertility drop between 1950 and 1990, a period without either smartphones or internet available...
Most countries are like that too
> Admittedly, in the old days, people predicted that the internet would make higher education obsolete because anybody could look up anything. Which did not exactly pan out.
It kind of did. Although people do go to university and study computer science the reality is that most people in our industry are primarily self-taught. You can get a high paid prestigious job in the software industry with either no degree at all or a degree in an irrelevant subject, purely on the back of self teaching, which itself is 95% just looking things up on the internet.
This hasn't happened in other professions but this is largely thanks to occupational licensing requirements preventing it from happening, rather than any failure of the prediction or internet itself. The best scientific and medical analysis during COVID reliably came from people who were not public health or medical professionals, mostly because the former are fools and the latter had the sword of license revocation forcing them to agree with the fools. It wasn't hard for "amateurs" who just looked things up to beat their predictive accuracy.
LLMs will destroy academia, not because LLMs themselves are particularly special (though they are), but because the prestige of academia is already in free-fall and thus people's minds are now open to the possibility that there's no way to stop cheating except by forcing offline essay writing in a closely supervised room. Schools will do this (more), universities won't, and thus everyone will intuit that everyone else is probably cheating because why wouldn't you? Professors can't tell and employers don't care.
The decline in innovation outside of computing does feel real. It's hard to think of examples of big quality-of-life improvements that come from other fields in the past 20 years (don't say mRNA vaccines please, they made COVID worse not better). My suspicion is that this is related to the decline of institutions more generally. Computing is unusual in that it has low startup costs and ready access to capital, so it's easy to leave declining institutions and create fresh companies. This is so much the case that "startup" is essentially synonymous with "computing startup". Other industries have very large high barriers to entry, so innovation has to take place in the context of existing institutions. If they decay then innovation slows down or becomes outright impossible.
As to why we're seeing across-the-board institutional decay, I think it can be traced back to the left's strategic switch in the 1970s-era away from trying to change society through anti-capitalist revolution / social democratic takeover, towards the broader "social justice" movement, which necessarily entails capturing companies and NGOs as much as or more than governments. Because the left believes in equity over excellence, any institution so captured becomes ever less competent and ever more willing to punish risk taking innovators. Innovation slowly dies as people learn it's not worth the risks, whilst also finding there's nowhere else to go.
The engine of innovation therefore won't be restarted by AGI even if it were to be achieved. Captured institutions ignore their existing humans who innovate, why would they care about humans who innovative better using AI? The only thing that can restore the march of progress is a comprehensive defeat of leftist ideology, e.g. a rollback of DEI and occupational licensing regimes.
Care to elaborate on your AGI denial?
Microplastics! You were making sense then you fell off the rail with that deus ex machina. I fear that unexpected effects of the continuing RNA Jabs will prove more deleterious to humanity's future than microplastics. But nice try.
My two younger kids went to a major state university - the University of Washington, in the US. We never considered the ivy's even though my daughter certainly would have had no problem getting into them. But they graduated with their Masters by age 21 with no or minimal student debt. My daughter with a MS in Structural Engineering and my son with a MS - MIS - Data Security. The major state universities are the primary producers of engineering and STEM talent in the economy - and the employers know it. Compared to what I saw as a teaching assistant in Mechanical Engineering 40 years ago, I don't think the expectations in engineering are significanly less than they were half a century ago. Yes we have more tools now - but the problems assigned are correspondingly harder.
> The method by which students are graded is up to the professor, and professors don’t really have much of an incentive to make those systems effective.
The incentive is literally the title of this article.
The professors don't care about the average prestige of their school, or even their entire profession, more than they care about not appearing racist, for which they are treated the more the demographics desparities of their classrooms are apparent.
The only way to do this is to degrade the standards in some form or another.
>"now less valuable now"
AI is a good argument, because AI does have the potential to dramatically disrupt academia, and there is a possibility that AI will make all human cognitive work obsolete. Who knows what that kind of world will look like?
As for Republicans' distrust of institutions, it's unclear whether this matters in the long-term or by how much it matters, since conservatives almost always lose and Republicans are increasingly stupid. There is also the possibility that Republicans issues with low trust are just part of a transitional period, that populism will fade and that American politics will see a new equilibrium with the two parties coming closer together again.
As for innovation and productivity, it's unclear whether innovation has slowed down, some of the arguments used in favor of this idea are really unconvincing, but even if innovation has slowed down it's possible that we are about to experience very significant innovations in the fields of medicine and tech, see Ozempic, embryo selection research, gene editing research and anti-aging research for the former and AI and new products like AR glasses for the latter. You also have the fact that solar power and batteries are getting very cheap, and cheaper energy could be the key to significantly increasing productivity. Productivity in the US has increased in recent years by the way.
Demographic aging is a problem for all aspects of society, but the US has the ability to continue importing high IQ people in relatively large numbers, and it is very plausible that eugenic technologies coming in the next few decades could cause very significant IQ gains. As for academia becoming less selective, that's true, but people probably already understand that and take it into account. People understand that different fields and different universities have different selection processes, and that there can be differences in the quality of individuals' work even for individuals who obtain the same degrees from the same universities.
A.I. will never make all human cognitive work obsolete. The smartest humans will still have the advantage of being more clever in what they ask of the computer, and the computer still shows no sign of having a will of its own for which to drive its own cognitive labors.
I find really hard to grasp the average PHD student have IQs of 100 pr even 90 according with razib khan's.
This measures are applied for STEM too?
No