“Maybe social media really is uniquely bad because of the network effect . . .”
The true cognitive cost of social media is seldom discussed: doom-scrolling their was to decay of cognitive patience. Intelligent people suddenly discover they can’t read a book. Imagine the loss of productivity, of someone who can’t read instructions or training materials, of the dangers caused in the workplace. Induced illiteracy leads to institutional incompetence.
It’s not for nothing that it’s called ‘the attention economy.” Impoverishment of intelligence: Social media gets your attention for free and sells it.
Adding to this, algorithms are also problematic. Social algorithms reward engagement, and negative engagement (anger, depressive information, outrage, etc.) outdraws positive engagement (good news, affirmations, etc.). People who are in tune with the incentive structure create inflammatory content to dominate engagement algorithms, either as personalities or through bots they created.
I remember a Twitter account that would aggregate the Top 10 Facebook engagements. Each day, Ben Shapiro and Dan Bongino would account for about 7 of the Top 10. Fox News or Prager U might account for 1 or 2, and one would be something completely random like a weather article or some really high-priced real estate sale.
TikTok is also rewarding the stupidest and most dangerous behavior. See numerous "Challenge" videos, or just this week when users made "I just read Osama bin Laden's 'Letter to America' and bin Laden was right!" takes.
My previous occupation- surgical pathology= diagnosing cancer -can soon be done better by AI. There are thousands of matrices for malignancies, so many there is hyper-specialization, eg renal pathologist, cytopathogist, etc
A well-trained AI can remember the near Infinite permutations better than our very frail human memory.
Nov 17, 2023·edited Nov 17, 2023Liked by Noah Smith
AI companies have been having trouble with "look at medical imaging and make a diagnosis" AI. Last I heard, AI starts making a lot of mistakes as soon as you start giving it images from an X-ray machine that wasn't used in its training data.
Yeah, although this seems like an artifact of bad data or training ("we only gave it one X-ray machine's output to train on") rather than a fundamental barrier. I don't think any task that is primarily based on visual processing will ultimately be done better by people.
Generally what we've seen is even if AI>people, people + AI > AI in medicine. Radiology is one of the leading fields in incorporating AI.
There are several Cancers that we lack good screening tests for. At some point AI will improve screening to the point that we image way more people... potentially increasing demand for radiologists to double check the AI.
I agree, for the time being it seems like imaging will get more useful prevalent by virtue of adding AI. Thus, demand will go up, and we'll need more radiologists as well.
I find cardiology interesting. By virtue of better screening (apple device following cardiac activity and flagging, easier portable telemetry) and more accurate coronary CT, we might need more interventional cardiologists to cath people, but have far fewer emergent interventions.
But then, once Ozempic follows the course of viagra and gets cheap, a lot of cardiologists will be out of a job.
For a future newsletter, I would like to hear your opinion on this October 30th article by Rogé Karma in the Atlantic: The Secretive Industry Devouring the U.S. Economy; Private equity has made one-fifth of the market effectively invisible to investors, the media, and regulators. He cites some rather alarming statistics, in particular, the number of publicly traded companies today is less than half of what it was in 1996. It also seems to me that we are flying blind with respect to what we really know about the economy, given that much of it is hidden from view.
PE is kind of fascinating in that it mixes real robber baron style shenanigans with fixing up companies and making them better. There are the scenarios where the PE firm lays off a bunch of people, saddles the company with debt, and runs. Then there are also the ones where a company has fundamental problems that management hasn't addressed, and going private can help work those out. Imagine a scenario where Twitter had been bought by someone less capricious who didn't alienate advertisers but still managed to cut operating costs by whatever amount. That would be a big success and would make it a healthier company overall.
Anyway, I get the impression there are lots of horror stories and also some success stories, but it's hard to tell what the mix is.
Agree. Non-public companies do seem sort of creepy, but what are the real costs and benefits? Are there reporting requirements that could mitigate the information loss?
Regarding #4 ("Is Social Media a Trap?"): I can't take any grand proclamations about social media seriously until and unless someone can truly define what constitutes "social media". Otherwise, it's just "a handful of very popular websites which happen to be very poorly run and are increasingly seen as a poor user experience, yet still stagger onward in a state of stagnation for lack of more exciting alternatives".
Livejournal was arguably the original "social media" back in the day, but nobody outside the Jack Thompson wingnut contingent ever seriously considered banning Livejournal.
MySpace was social media and very popular in its heyday, certainly more mainstream than Livejournal and definitely "FOMO territory" amongst 13-25 year olds, and had its fair share of sensationalistic negative media coverage, yet never attracted the same putrid miasma that oozes off every discussion about Facebook since 2018 or thereabouts.
LinkedIn is clearly social media, but it's also clearly not what people mean when they're talking about "the ills of social media". Youtube is often considered "social media" - does this mean Spotify is also social media? What about Medium and Substack? Should these sites fall under the hypothetical banhammer?
I personally think the issue is a bit nuanced, and if I had to boil it down to a quickly-digestible thesis, it'd be that the poor management of the current "big social sites" - which I'm going to semi-arbitrarily, but probably uncontroversially, define as "Facebook, Twitter (currently known as X (formerly known as Twitter), TikTok, Reddit, and Youtube" - lead to a mob rule environment, where a group of 50-100 people can get any "unprivileged" account immediately revoked by spamming reports. Reddit, to its credit, is less susceptible to this form of mob rule, but (to its equal demerit) mob rule is baked into the very nature of the discussion, where unpopular opinions are quickly downvoted and hidden, and heterogeneity withers.
To go a bit beyond mere thesis: This then incentivizes the melodramatic tabloid-headline approach to personal expression (if I want Noah to go away forever, I don't say his opinion is wrong, I say he himself is wrong, bad, and coming for you, yes YOU, with hate in his heart!) and, just as critically, penalizes the most effective counter-approach to a bunch of people being crazy and melodramatic at you: telling them to fuck off. Both "underpaid overseas modfarm workers who are forced to process hundreds of modqueue requests a day to make quota" and "algorithms" alike look very unkindly on a single account that's posting a large amount of negative-toned messages at other accounts.
The result is like a prison yard where standing up for one's self is banned, with similar results - the winners are those most inclined towards outrage, meanness, and being the first to attack.
Your choices are a suspension or embracing shellshock. It wasn't like this on Livejournal or Myspace. People might gang up on you, but you could freely tell them off, defend yourself (persuasively or not!) and carry on about your business, rather than wake up to find you've been muzzled by a looming shadowy authority figure that doesn't even know who you are because in 2014 you told your best friend to fuck off while trash-talking each other's favorite sports teams and ten years later a bunch of angry crazy computer people found it and reported it as harassment.
Both the media and academics are incentivized to avoid addressing this issue - for the mainstream media, histrionic outrage is better for impulsive clicks, and generates an infinite supply of content. Independent media outlets might be more inclined to poke the golden goose, except the threat of mob-reports doesn't ring "true" for them because they have contacts which can ensure that, should they ever find themselves on the receiving end of 50-100 angry people clicking "report" on a post they made in 2014 telling their best friend to "fuck off", their suspension appeal will be quickly looked at by an actual human in a non-superficial manner. And hence, "cancel culture isn't real".
Meanwhile, the academics whose field revolves around this sort of study are in the humanities, a field which is currently too captive to illiberal politics to risk an objective look at the totality of the situation - if the problem isn't straight white men then the paper's not worth publishing and you, the aspiring humanities student or non-tenured professor, are running interference for the "real" problem, and now that we, your ostensible colleagues, think about it, the land acknowledgement in your thesis was a little bit brief - perhaps mere non-publishment alone is too good for you... let's see what you talked about with your friends in 2014, shall we...?
I presume this was in reply to my labeling Livejournal "arguably the original social media", in which case I guess it depends on whether one considers USENET social media - I personally don't, because all the "friend"/"follower"/"feed" relationships were strictly client-side, and the emphasis was on the content itself, not the individual - closer to a newspaper than to following @Woodward_69 and seeing @Bear_Bernstein's content pushed into your feed via reshares or direct links from @Woodward_69. If USENET is social media, then a book club is social media.
But of course, "social media" is a vague term with room for debate, and I completely get where you're coming from.
Thank you for the simple yet very perceptive observation that bunnies are prey animals while my cat is a predator. I had not thought about that difference and how it makes those two animals so different. It is important. Thanks again.
As a fellow cat lover, I think a lot of the beautiful idiosyncrasies among our feline friends are from the tension that, while they are hyper efficient killing machines, they too are prey animals for a lot of slightly larger mammals (fisher cats, coyotes, wolves).
I liked the economic comparison you made between the Koreas. I am curious why people never draw the same conclusion when comparing China to Taiwan? Yes, China has done very well at bringing people out of poverty when compared to other disastrous communist dictatorships like North Korea but isn't their success undermined by the even greater success of Taiwan? The people in those respective countries are ethnically identical, they have roughly the same cultural heritage, and they have basically the same starting date (The Chinese Civil War).
Is there some economic thing that I don't understand about relative population sizes (or something else) that makes this comparison silly? I think Taiwan has roughly triple the GDP per capita of China but nobody ever references this fact as a failure of China's economic development.
As an aside, is the relative success of Taiwan a reason the CCP wants to destroy Taiwan (as they did with Hong Kong)? It's hard to claim the CCP's China is the best possible version of China when there are other Chinas doing better.
1. We don't know about real wages because BLS does not produce wage indexes. Wages of CEO's should affect only the CEO wage index. Disproportional layoffs of low-income workers do not affect a wage index; they affect a labor unit value index.
I can't quite put my finger on it, but the social media research seems like it has underlying conceptual issues. I'm not sure if it proves too much or too little. I think people would get value from "turning off" many parts of the economy that their friends or family enjoy but that they themselves don't. The stereotypical example would be non-sports fan spouses wanting to turn off sports. They either have to participate because of the network effect despite not enjoying it as much as whatever else they would be doing, or give up on that time with their sports fan significant other. Unless we have some understanding of the landscape of network effects and negative externalities for other commonly accepted interests, it's hard to tell whether social media is worse.
I've debated with others if dishwashing machines are automation. I was thinking of the comercial ones in resturant but the process is similar in your home. The human still has to load and unload the machine but the cleaning of the dishes is automated. Do we need a robot to load and unload before we can call it automation?
Re: North Korea and Vietnam normalization, I think the benefit Vietnam had is that most of the normalization efforts didn't make huge splashes in the news, and therefore could proceed without the histrionics of both sides that occurs with such efforts. People on the more liberal side are likely to freak out about siding with dictators/non-democratic countries, and the more right leaning side will freak out about negotiating with communists. Same thing with Nixon and China: most of the ground work happened behind the scenes and by the time the public became aware of it, it was already a done deal. Tough to imagine that could happen today with North Korea.
Trump had the right idea with North Korea. Selling them on liberal democracy was a nonstarter, so why not sell them on Singapore without the voting? But he lacked the diplomatic competence to follow through, and the media and Democrats did no favors to the peace efforts by chiding Trump for "cozying up to a dictator." He also had no support in Congress due to most GOP politicians still being Bush-era neoconservatives when it comes to foreign policy.
#5 In addition to Noah's questions there is the question of the effect on other kinds of labor. Let's suppose that profits of the firms that employed fewer free lancers and in Chat GPT providers increased. How do these incomes get spent?
I remember prior to the pandemic when Republicans would talk about how great the economy was under Trump they would bring up rising working class wages under Trump…of course they would ignore the fact the increase in wages mostly had to do with states and municipalities increasing the minimum wage to $15/hr. That said during the fracking boom working class wages did increase in regions with fracking…but that industry matured fairly quickly and so now we are increasing production with fewer workers.
The businesses that seem to be going the ChatGPT-as-panacea route are legacy and online journalism businesses. They are Radioshacking right now; cutting workforce and other costs just to keep the lights on another quarter.
Think of Radioshacking (or Searsing, or KMarting) as the non-tech approach to Cory Doctorow's enshittifcation principle.
Given that the Xi-Biden meeting was made possible by the fact APEC was on, I’d just like to remind everyone that from Bob Hawke putting together the APEC economic forum in the 80s followed by Paul Keating doing the work, starting with his first meeting with George HW Bush then having to start again with Clinton before getting Suharto onside to turn it into a strategic meeting and a leaders forum that APEC and all the benefits that flow from it being part of the international calendar is, in Keatings words ‘APEC is a little gift of Australian foreign policy to the world’
Wealth manager Barry Ritholtz originally thought "greedflation," or companies juicing profits during inflation, was overestimated. Now he's starting to rethink his position.
Ritholtz: "Any industry enjoying its most profitable period in history gets my attention.
My bias is that I was on Team Transitory from the beginning. For sure, transitory took longer than expected, but as we learned earlier this week, it asserted itself again. But the risk of “stickier” inflation remains, driven in large part by corporate profits, aka Rines’ PoV and PaM:
...
The “tell” about corporate profits and greedflation came after 2022 proved to be such a challenging year in the markets. Despite 500+ BPS of rate increases, a ~20% drop in the S&P500, and a 30+% drop in the Nasdaq 100, profits have remained much better than expected:
...
Corporate margins and profits could be the reason why price increases will stick, even as CPI falls back to normal."
“Maybe social media really is uniquely bad because of the network effect . . .”
The true cognitive cost of social media is seldom discussed: doom-scrolling their was to decay of cognitive patience. Intelligent people suddenly discover they can’t read a book. Imagine the loss of productivity, of someone who can’t read instructions or training materials, of the dangers caused in the workplace. Induced illiteracy leads to institutional incompetence.
It’s not for nothing that it’s called ‘the attention economy.” Impoverishment of intelligence: Social media gets your attention for free and sells it.
Adding to this, algorithms are also problematic. Social algorithms reward engagement, and negative engagement (anger, depressive information, outrage, etc.) outdraws positive engagement (good news, affirmations, etc.). People who are in tune with the incentive structure create inflammatory content to dominate engagement algorithms, either as personalities or through bots they created.
I remember a Twitter account that would aggregate the Top 10 Facebook engagements. Each day, Ben Shapiro and Dan Bongino would account for about 7 of the Top 10. Fox News or Prager U might account for 1 or 2, and one would be something completely random like a weather article or some really high-priced real estate sale.
TikTok is also rewarding the stupidest and most dangerous behavior. See numerous "Challenge" videos, or just this week when users made "I just read Osama bin Laden's 'Letter to America' and bin Laden was right!" takes.
My previous occupation- surgical pathology= diagnosing cancer -can soon be done better by AI. There are thousands of matrices for malignancies, so many there is hyper-specialization, eg renal pathologist, cytopathogist, etc
A well-trained AI can remember the near Infinite permutations better than our very frail human memory.
AI companies have been having trouble with "look at medical imaging and make a diagnosis" AI. Last I heard, AI starts making a lot of mistakes as soon as you start giving it images from an X-ray machine that wasn't used in its training data.
Yeah, although this seems like an artifact of bad data or training ("we only gave it one X-ray machine's output to train on") rather than a fundamental barrier. I don't think any task that is primarily based on visual processing will ultimately be done better by people.
Generally what we've seen is even if AI>people, people + AI > AI in medicine. Radiology is one of the leading fields in incorporating AI.
There are several Cancers that we lack good screening tests for. At some point AI will improve screening to the point that we image way more people... potentially increasing demand for radiologists to double check the AI.
I agree, for the time being it seems like imaging will get more useful prevalent by virtue of adding AI. Thus, demand will go up, and we'll need more radiologists as well.
I find cardiology interesting. By virtue of better screening (apple device following cardiac activity and flagging, easier portable telemetry) and more accurate coronary CT, we might need more interventional cardiologists to cath people, but have far fewer emergent interventions.
But then, once Ozempic follows the course of viagra and gets cheap, a lot of cardiologists will be out of a job.
You make a great point. I wonder if the frailty in our memory can be advantageous is some sense
What are you doing now? Are you retired?
For a future newsletter, I would like to hear your opinion on this October 30th article by Rogé Karma in the Atlantic: The Secretive Industry Devouring the U.S. Economy; Private equity has made one-fifth of the market effectively invisible to investors, the media, and regulators. He cites some rather alarming statistics, in particular, the number of publicly traded companies today is less than half of what it was in 1996. It also seems to me that we are flying blind with respect to what we really know about the economy, given that much of it is hidden from view.
Definitely need to write more about PE!
New Freakonomics podcast on this!
PE is kind of fascinating in that it mixes real robber baron style shenanigans with fixing up companies and making them better. There are the scenarios where the PE firm lays off a bunch of people, saddles the company with debt, and runs. Then there are also the ones where a company has fundamental problems that management hasn't addressed, and going private can help work those out. Imagine a scenario where Twitter had been bought by someone less capricious who didn't alienate advertisers but still managed to cut operating costs by whatever amount. That would be a big success and would make it a healthier company overall.
Anyway, I get the impression there are lots of horror stories and also some success stories, but it's hard to tell what the mix is.
Agree. Non-public companies do seem sort of creepy, but what are the real costs and benefits? Are there reporting requirements that could mitigate the information loss?
Christine did a spectacular job. The movie always reaches more people than the book, but your words inspired these lovely pictures.
Regarding #4 ("Is Social Media a Trap?"): I can't take any grand proclamations about social media seriously until and unless someone can truly define what constitutes "social media". Otherwise, it's just "a handful of very popular websites which happen to be very poorly run and are increasingly seen as a poor user experience, yet still stagger onward in a state of stagnation for lack of more exciting alternatives".
Livejournal was arguably the original "social media" back in the day, but nobody outside the Jack Thompson wingnut contingent ever seriously considered banning Livejournal.
MySpace was social media and very popular in its heyday, certainly more mainstream than Livejournal and definitely "FOMO territory" amongst 13-25 year olds, and had its fair share of sensationalistic negative media coverage, yet never attracted the same putrid miasma that oozes off every discussion about Facebook since 2018 or thereabouts.
LinkedIn is clearly social media, but it's also clearly not what people mean when they're talking about "the ills of social media". Youtube is often considered "social media" - does this mean Spotify is also social media? What about Medium and Substack? Should these sites fall under the hypothetical banhammer?
I personally think the issue is a bit nuanced, and if I had to boil it down to a quickly-digestible thesis, it'd be that the poor management of the current "big social sites" - which I'm going to semi-arbitrarily, but probably uncontroversially, define as "Facebook, Twitter (currently known as X (formerly known as Twitter), TikTok, Reddit, and Youtube" - lead to a mob rule environment, where a group of 50-100 people can get any "unprivileged" account immediately revoked by spamming reports. Reddit, to its credit, is less susceptible to this form of mob rule, but (to its equal demerit) mob rule is baked into the very nature of the discussion, where unpopular opinions are quickly downvoted and hidden, and heterogeneity withers.
To go a bit beyond mere thesis: This then incentivizes the melodramatic tabloid-headline approach to personal expression (if I want Noah to go away forever, I don't say his opinion is wrong, I say he himself is wrong, bad, and coming for you, yes YOU, with hate in his heart!) and, just as critically, penalizes the most effective counter-approach to a bunch of people being crazy and melodramatic at you: telling them to fuck off. Both "underpaid overseas modfarm workers who are forced to process hundreds of modqueue requests a day to make quota" and "algorithms" alike look very unkindly on a single account that's posting a large amount of negative-toned messages at other accounts.
The result is like a prison yard where standing up for one's self is banned, with similar results - the winners are those most inclined towards outrage, meanness, and being the first to attack.
Your choices are a suspension or embracing shellshock. It wasn't like this on Livejournal or Myspace. People might gang up on you, but you could freely tell them off, defend yourself (persuasively or not!) and carry on about your business, rather than wake up to find you've been muzzled by a looming shadowy authority figure that doesn't even know who you are because in 2014 you told your best friend to fuck off while trash-talking each other's favorite sports teams and ten years later a bunch of angry crazy computer people found it and reported it as harassment.
Both the media and academics are incentivized to avoid addressing this issue - for the mainstream media, histrionic outrage is better for impulsive clicks, and generates an infinite supply of content. Independent media outlets might be more inclined to poke the golden goose, except the threat of mob-reports doesn't ring "true" for them because they have contacts which can ensure that, should they ever find themselves on the receiving end of 50-100 angry people clicking "report" on a post they made in 2014 telling their best friend to "fuck off", their suspension appeal will be quickly looked at by an actual human in a non-superficial manner. And hence, "cancel culture isn't real".
Meanwhile, the academics whose field revolves around this sort of study are in the humanities, a field which is currently too captive to illiberal politics to risk an objective look at the totality of the situation - if the problem isn't straight white men then the paper's not worth publishing and you, the aspiring humanities student or non-tenured professor, are running interference for the "real" problem, and now that we, your ostensible colleagues, think about it, the land acknowledgement in your thesis was a little bit brief - perhaps mere non-publishment alone is too good for you... let's see what you talked about with your friends in 2014, shall we...?
I was on USENET in the 1990s...
I presume this was in reply to my labeling Livejournal "arguably the original social media", in which case I guess it depends on whether one considers USENET social media - I personally don't, because all the "friend"/"follower"/"feed" relationships were strictly client-side, and the emphasis was on the content itself, not the individual - closer to a newspaper than to following @Woodward_69 and seeing @Bear_Bernstein's content pushed into your feed via reshares or direct links from @Woodward_69. If USENET is social media, then a book club is social media.
But of course, "social media" is a vague term with room for debate, and I completely get where you're coming from.
Morning Noah,
Thank you for the simple yet very perceptive observation that bunnies are prey animals while my cat is a predator. I had not thought about that difference and how it makes those two animals so different. It is important. Thanks again.
As a fellow cat lover, I think a lot of the beautiful idiosyncrasies among our feline friends are from the tension that, while they are hyper efficient killing machines, they too are prey animals for a lot of slightly larger mammals (fisher cats, coyotes, wolves).
Hey Noah,
I liked the economic comparison you made between the Koreas. I am curious why people never draw the same conclusion when comparing China to Taiwan? Yes, China has done very well at bringing people out of poverty when compared to other disastrous communist dictatorships like North Korea but isn't their success undermined by the even greater success of Taiwan? The people in those respective countries are ethnically identical, they have roughly the same cultural heritage, and they have basically the same starting date (The Chinese Civil War).
Is there some economic thing that I don't understand about relative population sizes (or something else) that makes this comparison silly? I think Taiwan has roughly triple the GDP per capita of China but nobody ever references this fact as a failure of China's economic development.
As an aside, is the relative success of Taiwan a reason the CCP wants to destroy Taiwan (as they did with Hong Kong)? It's hard to claim the CCP's China is the best possible version of China when there are other Chinas doing better.
I chose a world where we decide to protect the weak, especially the children who are victims of warfare.
Ayn Rand disapproves!
1. We don't know about real wages because BLS does not produce wage indexes. Wages of CEO's should affect only the CEO wage index. Disproportional layoffs of low-income workers do not affect a wage index; they affect a labor unit value index.
The Employment Cost Index, which I mentioned in this post, is a wage index.
I can't quite put my finger on it, but the social media research seems like it has underlying conceptual issues. I'm not sure if it proves too much or too little. I think people would get value from "turning off" many parts of the economy that their friends or family enjoy but that they themselves don't. The stereotypical example would be non-sports fan spouses wanting to turn off sports. They either have to participate because of the network effect despite not enjoying it as much as whatever else they would be doing, or give up on that time with their sports fan significant other. Unless we have some understanding of the landscape of network effects and negative externalities for other commonly accepted interests, it's hard to tell whether social media is worse.
I've debated with others if dishwashing machines are automation. I was thinking of the comercial ones in resturant but the process is similar in your home. The human still has to load and unload the machine but the cleaning of the dishes is automated. Do we need a robot to load and unload before we can call it automation?
Re: North Korea and Vietnam normalization, I think the benefit Vietnam had is that most of the normalization efforts didn't make huge splashes in the news, and therefore could proceed without the histrionics of both sides that occurs with such efforts. People on the more liberal side are likely to freak out about siding with dictators/non-democratic countries, and the more right leaning side will freak out about negotiating with communists. Same thing with Nixon and China: most of the ground work happened behind the scenes and by the time the public became aware of it, it was already a done deal. Tough to imagine that could happen today with North Korea.
Trump had the right idea with North Korea. Selling them on liberal democracy was a nonstarter, so why not sell them on Singapore without the voting? But he lacked the diplomatic competence to follow through, and the media and Democrats did no favors to the peace efforts by chiding Trump for "cozying up to a dictator." He also had no support in Congress due to most GOP politicians still being Bush-era neoconservatives when it comes to foreign policy.
#5 In addition to Noah's questions there is the question of the effect on other kinds of labor. Let's suppose that profits of the firms that employed fewer free lancers and in Chat GPT providers increased. How do these incomes get spent?
I remember prior to the pandemic when Republicans would talk about how great the economy was under Trump they would bring up rising working class wages under Trump…of course they would ignore the fact the increase in wages mostly had to do with states and municipalities increasing the minimum wage to $15/hr. That said during the fracking boom working class wages did increase in regions with fracking…but that industry matured fairly quickly and so now we are increasing production with fewer workers.
The businesses that seem to be going the ChatGPT-as-panacea route are legacy and online journalism businesses. They are Radioshacking right now; cutting workforce and other costs just to keep the lights on another quarter.
Think of Radioshacking (or Searsing, or KMarting) as the non-tech approach to Cory Doctorow's enshittifcation principle.
Given that the Xi-Biden meeting was made possible by the fact APEC was on, I’d just like to remind everyone that from Bob Hawke putting together the APEC economic forum in the 80s followed by Paul Keating doing the work, starting with his first meeting with George HW Bush then having to start again with Clinton before getting Suharto onside to turn it into a strategic meeting and a leaders forum that APEC and all the benefits that flow from it being part of the international calendar is, in Keatings words ‘APEC is a little gift of Australian foreign policy to the world’
Edifying as alway, thank you. And as always the bunny notes are the highlight of every post.
Wealth manager Barry Ritholtz originally thought "greedflation," or companies juicing profits during inflation, was overestimated. Now he's starting to rethink his position.
https://ritholtz.com/2023/11/revisiting-greedflation/
Ritholtz: "Any industry enjoying its most profitable period in history gets my attention.
My bias is that I was on Team Transitory from the beginning. For sure, transitory took longer than expected, but as we learned earlier this week, it asserted itself again. But the risk of “stickier” inflation remains, driven in large part by corporate profits, aka Rines’ PoV and PaM:
...
The “tell” about corporate profits and greedflation came after 2022 proved to be such a challenging year in the markets. Despite 500+ BPS of rate increases, a ~20% drop in the S&P500, and a 30+% drop in the Nasdaq 100, profits have remained much better than expected:
...
Corporate margins and profits could be the reason why price increases will stick, even as CPI falls back to normal."