by Matthew Yglesias
Updated Oct 8, 2015, 10:00 PM GMT+5:30
America’s constitutional democracy is going to collapse.
Some day — not tomorrow, not next year, but probably sometime before runaway climate change forces us to seek a new life in outer-space colonies — there is going to be a collapse of the legal and political order and its replacement by something else. If we’re lucky, it won’t be violent. If we’re very lucky, it will lead us to tackle the underlying problems and result in a better, more robust, political system. If we’re less lucky, well, then, something worse will happen.
Very few people agree with me about this, of course. When I say it, people generally think that I’m kidding. America is the richest, most successful country on earth. The basic structure of its government has survived contested elections and Great Depressions and civil rights movements and world wars and terrorist attacks and global pandemics. People figure that whatever political problems it might have will prove transient — just as happened before.
But voiced in another register, my outlandish thesis is actually the conventional wisdom in the United States. Back when George W. Bush was president and I was working at a liberal magazine, there was a very serious discussion in an editorial meeting about the fact that the United States was now exhibiting 11 of the 13 telltale signs of a fascist dictatorship. The idea that Bush was shredding the Constitution and trampling on congressional prerogatives was commonplace. When Obama took office, the partisan valence of the complaints shifted, but their basic tenor didn’t. Conservative pundits — not the craziest, zaniest ones on talk radio, but the most serious and well-regarded — compare Obama’s immigration moves to the actions of a Latin-American military dictator.
In the center, of course, it’s an article of faith that when right and left talk like this they’re simply both wrong. These are nothing but the overheated squeals of partisans and ideologues.
RelatedHow to stop Trump’s power grab
At the same time, when the center isn’t complaining about the excessively vociferous complaints of the out-party of the day, it tends to be in full-blown panic about the state of American politics. And yet despite the popularity of alarmist rhetoric, few people act like they’re actually alarmed. Accusations that Barack Obama or John Boehner or any other individual politician is failing as a leader are flung, and then abandoned when the next issue arises. In practice, the feeling seems to be that salvation is just one election away. Hillary Clinton even told Kara Swisher that her agenda as a presidential candidate would be to end partisan gridlock.
It’s not going to work.
The breakdown of American constitutional democracy is a contrarian view. But it’s nothing more than the view that rather than everyone being wrong about the state of American politics, maybe everyone is right. Maybe Bush and Obama are dangerously exceeding norms of executive authority. Maybe legislative compromise really has broken down in an alarming way. And maybe the reason these complaints persist across different administrations and congresses led by members of different parties is that American politics is breaking down.
The perils of presidential democracy
To understand the looming crisis in American politics, it’s useful to think about Germany, Japan, Italy, and Austria. These are countries that were defeated by American military forces during the Second World War and given constitutions written by local leaders operating in close collaboration with occupation authorities. It’s striking that even though the US Constitution is treated as a sacred text in America’s political culture, we did not push any of these countries to adopt our basic framework of government.
This wasn’t an oversight.
In a 1990 essay, the late Yale political scientist Juan Linz observed that “aside from the United States, only Chile has managed a century and a half of relatively undisturbed constitutional continuity under presidential government — but Chilean democracy broke down in the 1970s.”
The exact reasons for why are disputed among scholars — in part because you can’t just randomly assign different governments to people. One issue here is that American-style systems are much more common in the Western Hemisphere and parliamentary ones are more common elsewhere. Latin-American countries have experienced many episodes of democratic breakdown, so distinguishing Latin-American cultural attributes from institutional characteristics is difficult.
Still, Linz offered several reasons why presidential systems are so prone to crisis. One particularly important one is the nature of the checks and balances system. Since both the president and the Congress are directly elected by the people, they can both claim to speak for the people. When they have a serious disagreement, according to Linz, “there is no democratic principle on the basis of which it can be resolved.” The constitution offers no help in these cases, he wrote: “the mechanisms the constitution might provide are likely to prove too complicated and aridly legalistic to be of much force in the eyes of the electorate.”
In a parliamentary system, deadlocks get resolved. A prime minister who lacks the backing of a parliamentary majority is replaced by a new one who has it. If no such majority can be found, a new election is held and the new parliament picks a leader. It can get a little messy for a period of weeks, but there’s simply no possibility of a years-long spell in which the legislative and executive branches glare at each other unproductively.
But within a presidential system, gridlock leads to a constitutional trainwreck with no resolution. The United States’s recent government shutdowns and executive action on immigration are small examples of the kind of dynamic that’s led to coups and putsches abroad.
There was, of course, the American exception to the problems of the checks-and-balances system. Linz observed on this score: “The uniquely diffuse character of American political parties — which, ironically, exasperates many American political scientists and leads them to call for responsible, ideologically disciplined parties — has something to do with it.”
For much of American history, in other words, US political parties have been relatively un-ideological and un-disciplined. They are named after vague ideas rather than specific ideologies, and neither presidents nor legislative leaders can compel back-bench members to vote with them. This has often been bemoaned (famously, a 1950 report by the American Political Science Association called for a more rigorous party system) as the source of problems. It’s also, according to Linz, helped avert the kind of zero-sum conflicts that have torn other structurally similar democracies apart. But that diffuse party structure is also a thing of the past.
A short history of American polarization
American politics is much more polarized today than it was 25 or 50 years ago. But not everyone buys the theory that today’s era of party polarization spells big trouble. Political scientist Jonathan Bernstein argues that it’s “not some sort of freakish un-American phenomenon.” The real exception, Bernstein says, the middle of the twentieth century, when the parties weren’t polarized. Polarization is the norm, he says, and he’s right.
A long line of research starting with Keith Poole and Howard Rosenthal, political scientists at the University of Georgia and New York University respectively, records all congressional votes and then analyzes the types of political coalitions that emerge. This system, known as DW-NOMINATE, lets you measure the degree of party polarization precisely. When Democrats all vote one way and Republicans all vote the other way, politics is highly polarized. When votes frequently scramble the parties, it is less polarized.
What this research shows is that the steady march toward polarization over the past generation is a return to a situation that existed during an earlier period.
The story here, like so much in American politics, is race. Southern Democrats had a range of views on non-racial issues but monolithically supported white supremacy and held together in the Democratic Party to maximize their leverage in Congress. The result was that the Democratic Party included Northern liberals who supported civil rights and Southern conservatives who supported segregation. So polarization temporarily went away in Congress. But as segregation receded as an issue in American politics, the parties slowly but surely sorted themselves by ideology, and so today, there is no Republican in Congress more liberal than the most conservative Democrat, or vice-versa. American politics has re-polarized. According to Bernstein, this change may be discomfiting but it’s nothing to worry about. American politics has been polarized before and it was fine.
What this story of reversion misses is the crucial role of ideology. Polarization and ideology are clearly related concepts, but simply counting congressional votes doesn’t really tell us what those votes were about. Georgetown University Professor Hans Noel greatly improved our understanding of the relationship between the two by extending the DW-NOMINATE methodology to people who aren’t elected officials.
For his book Political Parties and Political Ideologies in America, Noel constructs ideological space scores for writers and political pundits — people who address the same issues as elected officials but who are not serving on Capitol Hill.
What he found is that while Gilded Age members of Congress voted in a highly partisan way, their voting didn’t reflect any polarization of ideas evident in broader American society. As Charles Calhoun, a leading scholar of Gilded Age politics has written, the main concern of actual members of Congress was not policy, but “patronage power, the privilege of placing one’s political friends and supporters in in subordinate offices.” In other words, a member of Congress would get to distribute federal jobs and contracts to his supporters and in exchange the beneficiaries of his patronage would support his party’s ticket at all levels. For this reason, the obscure-sounding job of customs collector of the Port of New York was important enough in the 1870s that Chester A. Arthur leapt from it to the Vice Presidency. The first real filibuster was held over Whig efforts to assign a printing contract to friendly companies.
Even though party discipline was strict in these days, it was not really about much beyond who held the spoils.
Over the course of the 1920s, 1930s, and 1940s the rise of progressive and liberal ideology and the formation of a conservative ideology to counter it upended this system. So much so that by the 1970s it had become common to observe that American political parties were in decline. University of California Irvine political scientist Martin Wattenberg achieved the apogee of this literature with his 1985 classic The Decline of Political Parties in America (since updated in five subsequent editions), citing the waning influence of party professionals, the rise of single-issue pressure groups, and an attendant fall in voter turnout. But as historian Sam Rosenfeld writes, under-the-hood changes in the process for selecting presidential nominees and Congressional leaders “ultimately helped to create a newly receptive institutional setting for issue-based activism within the parties,” leading to the parties’ reconstitution around modern ideological lines.
Today’s partisan polarization, in other words, is not the same as its Gilded Age predecessor. The old polarization was about control over jobs and money — the kind of thing where split-the-difference compromises are easiest. That polarization was eventually undermined by a new politics built around principles. For decades, politicians found themselves cross-pressured between their commitments to a national party network and to various ideological causes. Today, however, politicians are no longer cross-pressured. We have strong Gilded Age-style parties, but organized around questions of principle rather than questions of patronage.
You can take this theory too far, of course. There have been moments in American life where questions of principle sharply split American politics. We had ideological parties (or at least one) in the 1850s when the anti-slavery Republican Party rose to the fore. But the example is not enormously encouraging — the constitutional process collapsed and we had four years of civil war, hundreds of thousands of deaths, and then, even after a Union victory, white supremacy was re-established in the South through a two-decade campaign of terrorism.
The Honduras scenario
Not all breakdowns of constitutional processes are as violent as the American Civil War.
For a less catastrophic, more realistic view of the kind of thing that could happen here, it’s useful to look to some less-familiar but more-recent events in Honduras. Back in late 2008, left-wing President Manuel Zelaya was locked in persistent conflict with an opposition-controlled congress. With neither side able to prevail within the context of the existing system, Zelaya decided he wanted to add a fourth question to the upcoming November 2009 election. In addition to voting for president, congress, and municipal offices, Zelaya would ask the voters whether they wanted to hold a constituent assembly to re-write the constitution — presumably to allow him to run for re-election.
Unfortunately for Zelaya, Honduras’ existing constitution made no provision for re-writing the constitution by plebiscite. Consequently, in March 2009, Zelaya determined that the solution was to hold another plebiscite. On June 28, Hondurans would go to the polls to vote in a non-binding referendum on whether the constitutional question should be added to the November ballot. This, he hoped, would give him the democratic legitimacy needed to go forward with the constitutional revision.
Zelaya’s opponents in congress, evidently concerned that the president would win, sued. They won a court case enjoining the president against holding the referendum. Zelaya pressed ahead regardless.
In Honduras, the military typically assists with election logistics, so Zelaya ordered the army to begin distributing ballots. General Romeo Vásquez Velásquez, the chief the Honduran military, refused to comply. On May 24, Zelaya fired the general. Several other commanders quit in solidarity. The Supreme Court ruled that the dismissal was unconstitutional. Throughout June, the constitutional process essentially broke down with protests and counter-protests dominating the capital. On June 28, the military deposed Zelaya in a coup, retroactively justified by a back-dated Supreme Court ruling. Roberto Micheletti, the president of the National Congress, was installed in his stead.
The military quickly handed power over to a new group of civilians. The coup was legitimated by the National Congress and the Supreme Court. And its perpetrators argued with some justification that there was no constitutional alternative. Zelaya was trying to circumvent the rules, so they had no choice but to circumvent them too in response.
The deadlock was ultimately resolved by force rather than legal procedure. Zelaya did not have enough support to amend the constitution through the existing process, and Honduras’ constitutional system created no legal mechanism for impeachment of a president. The Supreme Court arbitrarily ruled that Zelaya’s effort to circumvent the amendment process via referendum was illegal, while Congress’ effort to circumvent the impeachment process was fine. There were quite a few injuries as protesters clashed with security forces, but no massive bloodshed.
Honduras’ coup is worth paying attention to not because the exact same scenario is likely to play out in the United States, but because it reveals how genuinely difficult it is to maintain constitutional politics in a presidential system.
Presidents feel themselves to be accountable for steering the nation. And all the evidence indicates that the public and the media do in fact hold presidents broadly accountable for national outcomes. Throughout the United States’ 2012 presidential campaign, for example, it was universally assumed that good news for the American economy (or for America more broadly) would redound to Barack Obama’s benefit even though control of policymaking was split between the White House and a GOP-dominated Congress.
As Obama put it in a November 2014 press conference, “people are going to ask for greater accountability and more responsibility from me than from anybody else in this town.” The problem is the president is not only held accountable for things that are in part outside his ability to control (gas prices, Ebola, or shark attacks) but for things that are actually under the control of his political adversaries. “I’m the guy who’s elected by everybody,” concluded Obama, “and they want me to push hard to close some of these divisions, break through some of the gridlock, and get stuff done.” If you’re going to be held accountable for outcomes, in other words, then you’d better act.
In a parliamentary system, this is simply democratic accountability in action. A head of government who strongly believes the nation needs actions the legislature won’t approve can dissolve parliament and hold a new election to decide the issue. In Honduras’ presidential system, the very act of trying to schedule a vote to resolve the deadlock was itself unconstitutional.
Constitutional hardball
The United States, of course, is a long way from a coup. What we are witnessing instead is a rise in what Georgetown University Professor Mark Tushnet labeled “constitutional hardball” in a 2004 article.
Constitutional hardball describes legal and political moves “that are without much question within the bounds of existing constitutional doctrine and practice but that are nonetheless in some tension with existing pre-constitutional understanding.” In other words, moves that do not violate the letter of the law, but do trample on our conventional understanding of how it is supposed to work.
Tushnet’s article is vital reading today in part because the different partisan context in which it was written can help shock people out of their entrenched positions. His lead example is from the George W. Bush administration, when liberals were concerned about the president taking power away from Congress. Tushnet describes the “strained” argument offered by Republican senators in 2005 that Democratic Party filibusters of Bush’s judicial nominees violated the constitution. At the time, of course, Democrats found the view that Republicans might simply ban the use of filibusters for this purpose outrageous. “The filibuster serves as a check on power,” said Harry Reid, “that preserves our limited government.” Joe Biden called the Republicans’ attempt to end the fillibuster “an example of the arrogance of power.”
But ultimately the hardball tactic for ending filibusters was used by Democrats in 2013 to halt Republican obstruction of Obama’s nominees. Republicans, Reid said, “have done everything they can to deny the fact that Obama had been elected and then reelected.” He argued he had no choice but to abandon a principle that just a few years ago he said was crucial to preserving American liberty. Meanwhile, Republicans who had supported the 2005 effort to weaken the filibuster executed a perfect flip-flop in the other direction.
Tushnet’s other example from the mid-2000s — Texas’ decision to redraw congressional district boundaries to advantage Republicans between censuses — seems almost adorably quaint by the standards of the Obama era.
From its very first months, Obama’s presidency has been marked by essentially nothing but constitutional hardball. During the Bush years, Democratic senators sporadically employed a variety of unusual delaying tactics to stymie his agenda. In 2009, Mitch McConnell and Senate Republicans retaliated by using tons of them, constantly. Suddenly filibustering went from something a Senate minority could do to something it did on pretty much all motions. George Washington University congressional scholar Sarah Binder observes that “leaders in the 1970s rarely felt compelled to file for cloture [to break filibusters], averaging fewer than one per month in some years” while in recent years Reid has filed over once per week.
As a political strategy, McConnell’s tactics were vindicated by the 2010 midterms, which showed that making the president look partisan, clumsy, and inept was a winning strategy.
Republicans in Congress subsequently moved beyond unusual acts of obstruction to an unprecedented use of the statutory debt ceiling into a vehicle for policymaking. Traditionally a bit of oddball American political theater immortalized in a funny West Wing scene, in 2011 the GOP threatened to provoke an unfathomable financial and constitutional crisis unless the Obama administration agreed to sweeping spending cuts. Again, there was nothing illegal about what Republicans in Congress did here — it was just, in its intent and its scope, unprecedented.
And it’s fairly clear that these actions, while consistent with Republican Party electoral success, have not exactly produced a well-respected legislature. Congressional approval ratings are so low — and have been for so long — that it’s become a subject of pollster humor. In 2013, Public Policy Polling found that congress was less popular than Genghis Khan, traffic jams, cockroaches, or Nickelback. In a less joking spirit, Gallup finds that the voters have less confidence in Congress than any other American institution, including big business, organized labor, banks, or television news.
As relations with Congress have worsened, the Obama administration has set about expanding executive authority over domestic policy to match Bush-era unilateralism in the national security domain. This came to the fore most publicly with Obama’s decision to protect millions of unauthorized migrants from deportation without congressional agreement.
As Vox’s Andrew Prokop has argued, the pattern is actually much broader. Obama’s handling of K-12 education policy is in some ways an even more paradigmatic example of constitutional hardball. The George W. Bush-era education law No Child Left Behind laid out penalties for state education systems that didn’t meet certain, rather unrealistic, targets. The law’s authors assumed that when the law came up for reauthorization, the targets would be changed. In case Congress didn’t act in time, the Secretary of Education also had the authority to issue waivers of the penalties. Since Congress no longer really functions, there has been no reauthorization of the law. So the Obama administration has issued waivers — but only to states that implement policy changes ordered by the Department of Education.
University of Chicago political scientist William Howell told Prokop this was a “new frontier” for executive policymaking. Yale Law School’s Bruce Ackerman says Obama used “a waiver provision for modest experiments and transformed it into a platform for the redesign of the statute.” Obama’s actions are clearly legal — but they are just as clearly a decision to creatively exploit the letter of the law to vastly expand the scope of executive power over the law.
Those who like these actions on their merits comfort themselves with the thought that these uses of executive power are pretty clearly allowed by the terms of the existing laws. This is true as far as it goes. But it’s also the case that Obama (or some future president) could have his political opponents murdered on the streets of Washington and then issue pardons to the perpetrators. This would be considerably more legal than a Zelaya-style effort to use a plebiscite to circumvent congressional obstruction — just a lot more morally outrageous. In either case, however, the practical issue would be not so much what is legal, but what people, including the people with guns, would actually tolerate.
Raising the stakes
America’s escalating game of constitutional hardball isn’t caused by personal idiosyncratic failings of individual people. Obama has made his share of mistakes, but the fundamental causes of hardball politics are structural, not personal. Personality-minded journalists often argue that a warmer executive would do a better job of building bridges to congress. But as Dartmouth’s Brendan Nyhan points out, “Bill Clinton’s more successful outreach to his opponents didn’t keep him from getting impeached. Likewise, George W. Bush was more gregarious than Obama, but it didn’t make him any more popular among Democrats once the post-9/11 glow had worn off.”
There’s a reason for this, and it gets to the core of who really runs American politics.
In a democratic society, elected officials are most directly accountable to the people who support them. And the people who support them are different than the people who don’t care enough about politics to pay much attention, or the people who support the other side. They are more ideological, more partisan, and they want to see the policies they support passed into law. A leader who abandons his core supporters because what they want him to do won’t be popular with most voters is likely, in modern American politics, to be destroyed in the next primary election.
The amateur ideological activists who eroded the power of the party professionals in the 1970s are now running the show. While Gilded Age activists traded support for patronage jobs, modern-day activists demand policy results in exchange for support. Presidents need to do everything within their legal ability to deliver the results that their supporters expect, and their opponents in Congress need to do everything possible to stop them. At one point, Republican congressional leaders were highly amenable to passing an immigration reform bill and the Obama administration insisted it had no means of circumventing the legislative process. But under pressure from their respective bases, Republicans found it impossible to compromise and Obama decided he had better find a way to go around Congress.
It is true that the mass public is not nearly as ideological as members of Congress. But the mass public is not necessarily active in democratic politics, either. Emory’s Alan Abramowitz finds that “the American public has become more consistent and polarized in its policy preferences over the past several decades.” He also writes that “this increase in consistency and polarization has been concentrated among the most politically engaged citizens.”
This rise in ideological activism has a number of genuinely positive impacts. It makes politics less corrupt. The least-polarized state legislatures in America are in places like Rhode Island and Louisiana, bastions of corruption rather than good government. It’s not a coincidence that the Tea Party surge led to the end of earmarked appropriations. But it heightened executive-legislative conflict and leads to what Linz termed the zero-sum character of presidential elections.”
Looking back at Bush’s election in 2000, one of the most remarkable things is how little social disorder there was. The American public wanted Al Gore to be president, but a combination of the Electoral College rules, poor ballot design in Palm Beach County, and an adverse Supreme Court ruling, put Bush in office. The general presumption among elites at the time was that Democrats should accept this with good manners, and Bush would respond to the weak mandate with moderate, consensus-oriented governance. This was not in the cards. Not because of Bush’s personal qualities (if anything, the Bush family and its circle are standard-bearers for the cause of relative moderation in the GOP), but because the era of the “partisan presidency” demands that the president try to implement the party’s agenda, regardless of circumstances. That’s how we got drastic tax cuts in 2001.
If the Bush years shattered the illusion that there’s no difference between the parties, the Obama years underscore how much control of the White House matters in an era of gridlock. The broadly worded Clean Air Act, whose relevant provisions passed in 1970, has allowed Obama to be one of the most consequential environmental regulators of all time — even though he hasn’t been able to pass a major new environmental bill. He’s deployed executive discretion over immigration enforcement on an unprecedented scale. And he’s left a legacy that could be rapidly reversed. A future Republican administration could not only turn back these executive actions, but substantially erode the Affordable Care Act.
The lessons of the 2000 and 2008 elections make it unnerving to imagine a Bush-Gore style recount occurring in 2015’s political atmosphere. The stakes of presidential elections are sky-high. And the constitutional system provides no means for a compromise solution. There can be only one president. And once he’s in office he has little reason to show restraint in the ambitions of the legislative — or non-legislative — agenda he pursues. In the event of another disputed election, it would be natural for both sides to push for victory with every legal or extra-legal means at their disposal.
Indeed, we ought to consider possibilities more disastrous than a repeat of the 2000 vote. What if a disputed presidential election coincided with a Supreme Court vacancy? What if the simultaneous deaths of the president and vice president brought to power a House Speaker from the opposite party? What if neither party secured a majority of electoral votes and a presidential election wound up being decided by a vote of the lame duck House of Representatives? What if highly partisan state legislatures start using their constitutional authority to rig the presidential contest? A system of undisciplined or non-ideological political parties has many flaws, but it is at least robust to a variety of shocks. Our current party alignment makes for a much more brittle situation, in which one of any number of crises where democratic norms and constitutional procedures diverge could bring us to a state of emergency.
A flawed system
The idea that America’s constitutional system might be fundamentally flawed cuts deeply against the grain of our political culture. But the reality is that despite its durability, it has rarely functioned well by the standards of a modern democracy. The party system of the Gilded Age operated through systematic corruption. The less polarized era that followed was built on the systematic disenfranchisement of African-Americans. The newer system of more ideological politics has solved those problems and seems in many ways more attractive. But over the past 25 years, it’s set America on a course of paralysis and crisis — government shutdowns, impeachment, debt ceiling crises, and constitutional hardball. Voters, understandably, are increasingly dissatisfied with the results and confidence in American institutions has been generally low and falling. But rather than leading to change, the dissatisfaction has tended to yield wild electoral swings that exacerbate the sense of permanent crisis.
As dysfunctional as American government may seem today, we’ve actually been lucky. No other presidential system has gone as long as ours without a major breakdown of the constitutional order. But the factors underlying that stability — first non-ideological parties and then non-disciplined ones — are gone. And it’s worth considering the possibility that with them, so too has gone the American exception to the rule of presidential breakdown. If we seem to be unsustainably lurching from crisis to crisis, it’s because we are unsustainably lurching from crisis to crisis. The breakdown may not be next year or even in the next five years, but over the next 20 or 30 years, will we really be able to resolve every one of these high-stakes showdowns without making any major mistakes? Do you really trust Congress that much?
The best we can hope for is that when the crisis does come, Americans will have the wisdom to do for ourselves what we did in the past for Germany and Japan and put a better system in place.
Correction: This piece has been updated to more accurately reflect the series of events surrounding the coup of Honduras’ former president Manuel Zelaya.
INDIA IPPON DESU
Wednesday, February 19, 2025
Tuesday, February 18, 2025
Why a nation of 1.45 billion wants more children
Soutik Biswas
India correspondent• BBC
Last year, India nudged past China to become the world's most populous country, according to UN estimates. With nearly 1.45 billion people now, you'd think the country would be quiet about having more children. But guess what? The chatter has suddenly picked up.
Leaders of two southern states – Andhra Pradesh and Tamil Nadu – have recently advocated more children. Andhra Pradesh is mulling providing incentives, citing low fertility rates and ageing population. The state also scrapped its "two-child policy" for local body elections, and reports say neighbouring Telangana may soon do the same. Next-door Tamil Nadu is also making similar, more exaggerated, noises.
India's fertility rate has fallen substantially - from 5.7 births per woman in 1950 to the current rate of two. Fertility rates have fallen below the replacement level of two births per woman in 17 of the 29 states and territories. (A replacement level is one at which new births are sufficient to maintain a stable population.)
The five southern Indian states lead India's demographic transition, achieving replacement-level fertility well ahead of others. Kerala reached the milestone in 1988, Tamil Nadu in 1993, and the rest by the mid-2000s.
Today, the five southern states have total fertility rates below 1.6, with Karnataka at 1.6 and Tamil Nadu at 1.4. In other words, fertility rates in these states match or are less than many European countries.
But these states fear that India's shifting demographics with varying population shares between states, will significantly impact electoral representation and state wise-allocation of parliamentary seats and federal revenues.
"They fear being penalised for their effective population control policies, despite being better economic performers and contributing significantly to federal revenues," Srinivas Goli, a professor of demography at the International Institute for Population Sciences, told the BBC.
Southern states are also grappling with another major concern as India prepares for its first delimitation of electoral seats in 2026 - the first since 1976.
This exercise will redraw electoral boundaries to reflect population shifts, likely reducing parliamentary seats for the economically prosperous southern states. As federal revenues are allocated based on state populations, many fear this could deepen their financial struggles and limit policy-making freedom.
Demographers KS James and Shubhra Kriti project that populous northern states like Uttar Pradesh and Bihar stand to gain more seats from delimitation, while southern states such as Tamil Nadu, Kerala, and Andhra Pradesh could face losses, further shifting political representation.
Many, including Prime Minister Narendra Modi, have hinted that changes to fiscal shares and parliamentary seat allocations will not be rushed through.
"As a demographer, I don't think states should be overly concerned about these issues. They can be resolved through constructive negotiations between federal and state governments," says Mr Goli. "My concern lies elsewhere."
The key challenge, according to demographers, is India's rapid ageing driven by declining fertility rates. While countries like France and Sweden took 120 and 80 years respectively to double their aging population from 7% to 14%, India is expected to reach this milestone in just 28 years, says Mr Goli.
This accelerated ageing is tied to India's unique success in fertility decline. In most countries, improved living standards, education, and urbanisation naturally lower fertility as child survival improves.
But in India, fertility rates fell rapidly despite modest socio-economic progress, thanks to aggressive family welfare programmes that promoted small families through targets, incentives, and disincentives.
The unintended consequence? Take Andhra Pradesh, for instance. Its fertility rate is 1.5, on par with Sweden, but its per capita income is 28 times lower, says Mr Goli. With mounting debt and limited resources, can states like these support higher pensions or social security for a rapidly aging population?
Consider this. More than 40% of elderly Indians (60+ years) belong to the poorest wealth quintile - the bottom 20% of a population in terms of wealth distribution, according to United Nations Population Fund (UNFPA)'s latest India Ageing Report.
In other words, Mr Goli says, "India is getting old before getting rich".
Fewer children also mean a rising old-age dependency ratio, leaving fewer caregivers for an expanding elderly demographic. Demographers warn that India's healthcare, community centres and old-age homes are unprepared for this shift.
India correspondent• BBC
Last year, India nudged past China to become the world's most populous country, according to UN estimates. With nearly 1.45 billion people now, you'd think the country would be quiet about having more children. But guess what? The chatter has suddenly picked up.
Leaders of two southern states – Andhra Pradesh and Tamil Nadu – have recently advocated more children. Andhra Pradesh is mulling providing incentives, citing low fertility rates and ageing population. The state also scrapped its "two-child policy" for local body elections, and reports say neighbouring Telangana may soon do the same. Next-door Tamil Nadu is also making similar, more exaggerated, noises.
India's fertility rate has fallen substantially - from 5.7 births per woman in 1950 to the current rate of two. Fertility rates have fallen below the replacement level of two births per woman in 17 of the 29 states and territories. (A replacement level is one at which new births are sufficient to maintain a stable population.)
The five southern Indian states lead India's demographic transition, achieving replacement-level fertility well ahead of others. Kerala reached the milestone in 1988, Tamil Nadu in 1993, and the rest by the mid-2000s.
Today, the five southern states have total fertility rates below 1.6, with Karnataka at 1.6 and Tamil Nadu at 1.4. In other words, fertility rates in these states match or are less than many European countries.
But these states fear that India's shifting demographics with varying population shares between states, will significantly impact electoral representation and state wise-allocation of parliamentary seats and federal revenues.
"They fear being penalised for their effective population control policies, despite being better economic performers and contributing significantly to federal revenues," Srinivas Goli, a professor of demography at the International Institute for Population Sciences, told the BBC.
Southern states are also grappling with another major concern as India prepares for its first delimitation of electoral seats in 2026 - the first since 1976.
This exercise will redraw electoral boundaries to reflect population shifts, likely reducing parliamentary seats for the economically prosperous southern states. As federal revenues are allocated based on state populations, many fear this could deepen their financial struggles and limit policy-making freedom.
Demographers KS James and Shubhra Kriti project that populous northern states like Uttar Pradesh and Bihar stand to gain more seats from delimitation, while southern states such as Tamil Nadu, Kerala, and Andhra Pradesh could face losses, further shifting political representation.
Many, including Prime Minister Narendra Modi, have hinted that changes to fiscal shares and parliamentary seat allocations will not be rushed through.
"As a demographer, I don't think states should be overly concerned about these issues. They can be resolved through constructive negotiations between federal and state governments," says Mr Goli. "My concern lies elsewhere."
The key challenge, according to demographers, is India's rapid ageing driven by declining fertility rates. While countries like France and Sweden took 120 and 80 years respectively to double their aging population from 7% to 14%, India is expected to reach this milestone in just 28 years, says Mr Goli.
This accelerated ageing is tied to India's unique success in fertility decline. In most countries, improved living standards, education, and urbanisation naturally lower fertility as child survival improves.
But in India, fertility rates fell rapidly despite modest socio-economic progress, thanks to aggressive family welfare programmes that promoted small families through targets, incentives, and disincentives.
The unintended consequence? Take Andhra Pradesh, for instance. Its fertility rate is 1.5, on par with Sweden, but its per capita income is 28 times lower, says Mr Goli. With mounting debt and limited resources, can states like these support higher pensions or social security for a rapidly aging population?
Consider this. More than 40% of elderly Indians (60+ years) belong to the poorest wealth quintile - the bottom 20% of a population in terms of wealth distribution, according to United Nations Population Fund (UNFPA)'s latest India Ageing Report.
In other words, Mr Goli says, "India is getting old before getting rich".
Fewer children also mean a rising old-age dependency ratio, leaving fewer caregivers for an expanding elderly demographic. Demographers warn that India's healthcare, community centres and old-age homes are unprepared for this shift.
Monday, February 17, 2025
Sunday, February 16, 2025
Saturday, February 15, 2025
Friday, February 14, 2025
Thursday, February 13, 2025
Notes on China
by Dwarkesh Patel
Last month, I spent 2 weeks in China - I visited Beijing, Chengdu, Emeishan, Chongqing, Shanghai, and Hangzhou.
I have no illusions about "understanding" China. I've only spent 2 weeks there. This trip is a beginning, not a capstone, of my curiosity about China. And I hope to share what I learn with you via future podcast episodes.
Scale
It’s funny how China has basically the inverse problem as America. We subsidize demand and restrict supply. They subsidize supply and restrict demand. We can’t rebuild fallen bridges. They build bridges to nowhere. In the most desirable cities in this country, every random Victorian house and park bench is a historic site that can’t be disturbed. There, they’ll bulldoze a 500 year old temple to build an endless skyscraper complex that no one wants to live in.
My overwhelming first impression was: wow everything is so fucking big: the cities themselves, the train stations, the airports, the towering and endless apartment complexes. Travel often teaches you things about a country which you honestly should have intuited even without visiting. Obviously, I knew that China is a big country, with over 1.4 billion people. But the stupendous scale of the biggest cities was impressed upon me only after I visited.
Even in Emeishan, a city of just half a million people (considered a quaint countryside town by Chinese standards), we found a Buddhist temple of comical scale - we'd enter what seemed like an impressively large compound, only to discover it was merely the entrance to an even grander structure right behind it. This pattern repeated 5 or 6 times, each subsequent building larger and more ornate than the last, like some kind of inverse nesting dolls. And the place had almost no other visitors!
I asked a monk at the temple how they funded this massive site in a city of just half a million people. He told us that it was simply through donations. We probed further about how such an enormous project could have been financed by just ordinary people's contributions. He responded, "We've got a lot of supporters, dude", and changed the topic.
Chongqing is by far the coolest city I've ever visited. It's this insane cyberpunk multi-level metropolis of over 20 million people. I wouldn't know how to begin describing it, but there's a bunch of great YouTube videos which will show you what I mean. I got a really nice nice 2-floor hotel room that overlooked two rivers and one of the most insane skylines in the world for 60 bucks - highly recommend visiting Chongqing if you get the chance.
In 1995, astronomers pointed Hubble at a seemingly empty patch of sky the size of a grain of sand held at arm's length. Instead of emptiness, the 10-day exposure revealed over 3,000 galaxies. Every speck of light in the image was an entire galaxy containing billions of stars. When I went atop the tallest building in Chongqing and looked out over the city, I thought about the Hubble image. You could zoom in on any direction you'd find - behind the fog and mist, beyond even perhaps the horizon - another skyscraper, each containing hundreds or thousands of people living or working.
We took a 12-hour village train from Chongqing to Shanghai. I'm embarrassed to say that my only experience with the actual countryside was via the windows of this train. Still, the sights were quite interesting. We saw again and again small paddy farms surrounding a handful of 5-10 story skyscraper, plopped in the seeming middle of nowhere. Even in the countryside, many people lived in large buildings instead of their own small homes. I couldn't hop off the train and confirm, but I saw many towns that looked quite ghostly - no actually visible people anywhere.
Outside of Beijing and Shanghai (and sometimes even within), you can tell that these skyscrapers were put up by a country with a GDP per capita of $10,000 (and potentially half or quarter than when many of these buildings went up). America and Europe put up a ton of beautiful buildings in the early 20th century when their GDP per capita was similar to China's - one could even argue that those older structures are more aesthetic than anything we're building today in the West. Not so in China. These endless rows of skyscrapers, put up in the construction frenzy of the last few decades, are ugly - boxes of mostly concrete with visible blight and discoloration all over them. If the great construction binge is indeed over, it'll be a shame that China's infrastructure was built out during a period of particularly uninspired architecture.
Beijing's urban design looks like something straight out of James Scott's "Seeing Like a State". The city is dominated by these enormous apartment complexes - blocks of 10 adjacent 30-story buildings demarcated by 8-lane roads. The government buildings follow the same pattern: huge structures divided by extremely wide boulevards. This layout seems designed partly for social control - during zero-COVID, authorities could lock down 10,000 people by simply guarding a few entrance gates. The wide roads would also make it easy to move military forces through the city. The only break from this pattern are the Hutongs, Beijing's old historic neighborhoods. But even these weren't spared completely - only a fraction survived Beijing's rapid modernization push. Dare I say that China is too YIMBY
Vibes
I got quite mixed messages about the state of public opinion in China. This is to be expected in a society where you can't establish common knowledge. One person told me that the new generation is quite nationalist, unlike the older reform generation which personally experienced the catastrophes of Mao and the tangible benefits of liberalization. He made the rather insightful point that this tilt in Chinese public opinion increasingly gives lie to the American talking point, "We're against the CCP, not the Chinese people." In fact, he went on to say that the current regime is way more liberal than what would result from an election in China.
Another person told me that these Chinese nationalists were only a vocal minority, similar to the wokes in America circa 2020. While they make up only about 10% of the population, they aggressively shout down others on Weibo (China's Twitter equivalent). Most people find them annoying but feel uncomfortable confronting them directly. This matches what a student who graduated from a top university there told me - the vast majority of his classmates are simply apolitical. And in our own interactions with locals, we saw little evidence of widespread nationalism. In fact, when my Chinese-speaking trip mate would mention he was from the UK to taxi drivers, they would often respond enthusiastically: "Oh wonderful, we love the UK!"
There were very few foreigners. In Beijing I might have seen half a dozen cumulatively across entire seas of people. In Chengdu and Chongqing, I barely remember seeing any (which is a real shame, because Chongqing is a truly incredible tourist destination). So rare apparently are tourists that in Chengdu and Chongqing we got asked for selfies many times.
Outside of Shanghai, almost nobody spoke English. If I went again, I would definitely try to crash course some basic Chinese beforehand. This language barrier did led to some interesting encounters. At a park in Chengdu, an old man reading a book called Medical English asked us to join him for tea. He was mostly just trying to practice his English. He said he loves foreigners, and that his favorite period of life was the 80s and 90s - "We love Deng!".
My trip mate's friend's grandmother was incredibly gracious in hosting us for a dinner while we were in Emeishan. We saw a beautiful Uighur rug hanging on their wall. The grandfather pointed it out and explained, “We have amazing relationships with minorities here in China.” I don't think this was some theatrical attempt to contradict Western propaganda. It seemed like he genuinely didn’t know what is said about the treatment of Uighurs in the West. He was just trying to show his visitors something cool he had.
We kept trying to ask them about their personal experiences over the last seven decades as China has grown and changed. but the grandfather kept responding with lengthy monologues about military history. Here is a representative exchange:
Me (who doesn’t speak Chinese) to my trip mate (who does): Ask them about what job they had when they first moved near Chengdu.
[The grandfather speaks in Chinese for 10 minutes. I get a bit impatient and whisper to my trip mate asking what he’s saying]
My trip mate: He's saying that the mountains around Chengdu make it a perfect base of retreat in case of an invasion, which Mao was worried about during the Sino Soviet split.
Me: What?! Why didn’t you ask him about the first job he had? My trip mate: I did!!
Grandpas are the same everywhere.
People say that Xi is establishing a cult of personality. This may be true within the CCP cadres, but I saw no evidence of it in public. I don't think I saw a single picture of Xi anywhere - not on any billboards, screens, or walls. People didn't really bring Xi up in conversations. I saw some pictures of Mao, but mostly in museums (or in one case at a tea farm he apparently used to frequent). The hammer and sickle was also a rare sight, mostly displayed on government buildings.
There are indeed cameras everywhere. This is gonna sound super naive - but I genuinely don't understand why. There's no crime. I know you'll say it's to prevent protests. Which might make sense for major streets, but even random alleyway corners will have a couple of cameras. Are they really trying to prevent someone from fomenting insurrection between 2 garbage cans? Beijing in particular had police officers at attention at what seemed like every street corner.
One recent student I talked to said that they understand they don't have freedoms here, but they're willing to take the tradeoff in favor of safety - they don't want school shootings. I thought this was quite silly - not only because there's no reason political freedom ought to lead to school shootings, but mostly because school shootings are such a statistically marginal experience. But I realized this is exactly the way we treat any hints of public protests in China. Just as school shootings are featured heavily in the media but aren't actually something you're likely to personally encounter, so too with protests against the CCP. You are overwhelmingly unlikely to spontaneously encounter them.
People were quite willing to chat openly in public places about problems in the country and with the regime. Including people who seemed to have a lot to lose. Almost everyone I talked to would acknowledge the economy was bad, and many were willing to implicate the government's decisions. Some even casually brought up Tiananmen or the Cultural Revolution. One person was even willing to discuss the odds of regime change at a public restaurant - though he may been have an especially careless fellow.
To be clear, it's an authoritarian system, and I certainly would feel uncomfortable doing what I'm doing there, but it definitely isn't North Korea.
Youngsters
In a shopping mall in Chongqing, a couple of high schoolers came up to us in order to get selfies. It felt like the perfect opportunity to learn about young adult life in China. So I whipped out the Translate app in WeChat and proceeded, rather clumsy, to make small talk. I asked them what they did in their free time. They said they watched 2-3 hours of TikTok every day. I asked them what videos they'd watch. They said it's a whole bunch of "sexy girls". I laughed because I thought they were joking. I asked one of them to pull out his phone. He scrolled past the first 10 videos on his feed and they were indeed all just "sexy girls".
We chatted up quite a lot of young people on night life streets. I was struck by how many young people expressed feeling stressed or overwhelmed. We met a musician in Chengdu who was writing songs about youth anxiety. We chatted up some modeling school students - even they complained about the intense pressure they felt. We met a guy who had studied in Australia but returned to China during COVID. He explained that many of his friends with prestigious degrees are moving away from Shanghai and Beijing - Yes, the pay there can be twice as high as in second or third tier cities. But the competitiveness is insane. And in order to actually land the high skilled positions, they have to work truly insane hours (9-9-6 is not a myth). He said that many of his friends were opting for these less ambitious lower-paying careers in smaller cities, where the rent is lower and the pressure is manageable.
Speaking of which - we really need to do a better job with Chinese students studying abroad in America. These students will likely end up in influential positions back home, yet colleges treat them basically like cash cows. They often arrive bombarded with propaganda about America, which is reinforced by the prevalent discourse at universities. And they find themselves isolated by language barriers and cultural differences. Giving these future leaders a genuinely positive experience in America might be the best thing we can do to improve US-China relations in the long run.
I'm still puzzled by how China can have both a demographic collapse and massive youth unemployment. You'd think with fewer young people being born, the ones who are around would be in high demand. One explanation I heard while there is that there are plenty of menial jobs available, but today's educated youth - who've gone through high school and college - just won't take the low-skilled positions their parents and grandparents did. Meanwhile, there's a real shortage of the high-skilled jobs that would actually match their education and aspirations. It's a mismatch between the jobs available and the jobs young people feel qualified for and willing to do.
I kept asking young people about the public intellectual landscape in China - who are their equivalents of Jordan Peterson, Joe Rogan, Lex Friedman, and Sam Harris? The sense I got is that this kind of popular intellectual ecosystem just doesn't exist there. Sure, there are viral Bilibili videos from professors talking about practical matters like how to manage your finances. But grand takes about what's happening in the world and what we should do about it? Not much going on.
I met a couple of EAs in China - they're exceptionally rare. When we asked them why it's so hard to spread EA or similar worldviews there, they said people just aren't into that organizing decisions using some ideological lens. They're much more concerned with practical, tangible matters. One talking point I heard again and again is that, "China is still a developing country." The implication seemed to be that China can't afford to pursue costly idealistic programs around climate change, social safety nets, or international aid - it needs to focus on basic economic development first.
The Great Firewall
By far the most inconvenient thing about visiting China is internet access. As a foreigner, basically all the websites you might find useful are behind the firewall. Even for websites that aren't blocked by the Great Firewall, I just didn't have anything even close to what you might call high-speed internet. This was true across my entire travel, including my burner SIM card from ChinaMobile or the WiFi at fancy hotels in Beijing or Shanghai. We released an episode of my podcast during this trip - because of the low bandwidth, my editor and I were just sending screenshots to each other throughout the night instead of hopping on a video call. The VPN situation is worse than I thought it would be . I happened to be lucky enough to download one of the VPNs that sometimes works before the trip. It was still pretty slow and unreliable. Astrill or Mullvad might actually result in a connection 75% of the time - often they would cause WeChat and Alipay to crash (which is rather inconvenient since you basically need those apps open all the time to navigate China). Very possible that I was fucking something up. But between these internet issues and the need to use a burner phone and laptop, I would be quite reluctant to use China as a remote work location.
Tech & AI
I am super hesitant to say anything here, because I am extremely unconfident about what's actually happening. Treat this just as some tentative notes from a few conversations.
I've cut out a whole bunch of stuff about AI in this section because I'm not sure if my initial assessment is correct. I hope to make podcast episodes over the next few months which provide a better researched account than I can supply here.
The biggest surprise from talking to Chinese VCs people at AI labs was how capital constrained they felt. Moonshot AI, one of China's leading AI labs, raised $1 billion at a $3 billion valuation. Meanwhile, just xAI's new cluster alone will cost $3-4 billion.
The tech ecosystem feels quite shell shocked from the 2021 crackdown. One VC half-jokingly asked if I could help him get his money out of China. If you keep your money in China, you're basically stuck choosing between terrible options. You can either accept a measly 2% yield from state banks, or throw it into China's perpetually struggling stock market. This helps explain why valuations for Chinese companies are chronically low - the exit opportunities just suck. Even if you build (or invest in) something great, there's no guarantee the company will be able to raise the next round. And even if you do raise again and succeed, the government might randomly cancel your IPO. And even if you somehow make it to the public markets, Chinese equities have been performing terribly anyways. It's a good reminder of how easy it is to completely wreck an innovation ecosystem that depends on risk-taking investors.
Hearts and Minds
In China, liberal pro-Western voices are often censored or shouted down. If I was the US President, and I wanted to win hearts and minds in China, here's what I'd do. In every single speech where I'm talking about China, I'd make a conspicuous effort to complement Chinese people, Chinese values, and Chinese culture. I'd talk about how my Chinese staffers are the smartest and most hardworking people I've ever worked with (which honestly is probably true). I'd talk about how much my daughter is obsessed with ancient Chinese dresses. I'd talk about how I'm learning Mandarin in my free time, and have a live "Aw shucks" conversation in Mandarin.
These clips would go viral on Bilibili and TikTok. And they'd probably stay up because it would just be a weird thing to censor. The CCP might even think that these displays of affection aggrandize them. But in reality, showing our admiration for Chinese people and their achievements (who genuinely are fucking killing it everywhere where they're not held down by communism), undermines the central narrative of the regime - that the West is hell bent on holding Chinese people back, that they have no respect or understanding of their culture, and that the CCP is a necessary bulwark against these imperialists.
On a totally unrelated note, so many people I met in China are in fact super talented and hard working. Someone connected me with the CEO of a company that manufactures life sciences equipment. He told me that before he started this company, he used to be a repairman for foreign imported machines, traveling 200 days a year with a screwdriver to wherever one had broken down. In 2013, he designed and started manufacturing his own machines, building more and more advanced designs, and now they've just built out a $60 million factory 1.5 hours from Shanghai. I asked him what the hardest part of ramping up production is - apparently constructing factories is just not an issue - tons of construction firms can make you a new facility quite reliably (along with the adjacent dorm room building for workers).
What is travel good for?
Noah Smith has a good blog post about what one can't learn from travel - you're not going to learn about the risk of a war or the state of the AI race by gazing at skylines or chatting up taxi drivers. Of course you can learn about those things by talking to the princelings and researchers and CEOs. But if you have access to these higher ups, surely you can also get them on the Zoom call. And fwiw, this should you update you in favor of more Zoom calls, not less travel. During the trip I realized how much I could have already learned about China by meeting the listeners I have in China (who are unusually high quality, presumably because of the restrictions they have to get through to get to my content).
Incidentally, if you're in China (and especially if you work on anything related to AI) please email me at hello@dwarkeshpatel.com. I'm keen for people to reach out and suggest guests to chat about China. Especially keen to better understand the Chinese political system, and how decisions about mega projects, arms races, and technology investment will be made at the brink of AGI.
So what's the point of travel? What about the in person experience can't be replaced by books, travel vlogs, and Zoom calls? For me, it's something like, what becomes salient to you. I started asking questions about China I hadn't even thought to ask about America.
Another thing I noticed is more personal. Two weeks of being AFK, and of having the excuse of using a burner phone to put off messages, helped me clear the cache of thoughts about sponsorships, logistics, growth, hiring, and a bunch of other practical minutiae. My shower thoughts wandered away from upcoming negotiations and towards interesting rabbit holes.
It's a good reminder that what's lacking in life is not time. It's focus. If you're working on what matters, you can advance leaps and bounds in 8 hours. And if you're just clearing the slog, you can spend a lifetime staying in the same place.
Last month, I spent 2 weeks in China - I visited Beijing, Chengdu, Emeishan, Chongqing, Shanghai, and Hangzhou.
I have no illusions about "understanding" China. I've only spent 2 weeks there. This trip is a beginning, not a capstone, of my curiosity about China. And I hope to share what I learn with you via future podcast episodes.
Scale
It’s funny how China has basically the inverse problem as America. We subsidize demand and restrict supply. They subsidize supply and restrict demand. We can’t rebuild fallen bridges. They build bridges to nowhere. In the most desirable cities in this country, every random Victorian house and park bench is a historic site that can’t be disturbed. There, they’ll bulldoze a 500 year old temple to build an endless skyscraper complex that no one wants to live in.
My overwhelming first impression was: wow everything is so fucking big: the cities themselves, the train stations, the airports, the towering and endless apartment complexes. Travel often teaches you things about a country which you honestly should have intuited even without visiting. Obviously, I knew that China is a big country, with over 1.4 billion people. But the stupendous scale of the biggest cities was impressed upon me only after I visited.
Even in Emeishan, a city of just half a million people (considered a quaint countryside town by Chinese standards), we found a Buddhist temple of comical scale - we'd enter what seemed like an impressively large compound, only to discover it was merely the entrance to an even grander structure right behind it. This pattern repeated 5 or 6 times, each subsequent building larger and more ornate than the last, like some kind of inverse nesting dolls. And the place had almost no other visitors!
I asked a monk at the temple how they funded this massive site in a city of just half a million people. He told us that it was simply through donations. We probed further about how such an enormous project could have been financed by just ordinary people's contributions. He responded, "We've got a lot of supporters, dude", and changed the topic.
Chongqing is by far the coolest city I've ever visited. It's this insane cyberpunk multi-level metropolis of over 20 million people. I wouldn't know how to begin describing it, but there's a bunch of great YouTube videos which will show you what I mean. I got a really nice nice 2-floor hotel room that overlooked two rivers and one of the most insane skylines in the world for 60 bucks - highly recommend visiting Chongqing if you get the chance.
In 1995, astronomers pointed Hubble at a seemingly empty patch of sky the size of a grain of sand held at arm's length. Instead of emptiness, the 10-day exposure revealed over 3,000 galaxies. Every speck of light in the image was an entire galaxy containing billions of stars. When I went atop the tallest building in Chongqing and looked out over the city, I thought about the Hubble image. You could zoom in on any direction you'd find - behind the fog and mist, beyond even perhaps the horizon - another skyscraper, each containing hundreds or thousands of people living or working.
We took a 12-hour village train from Chongqing to Shanghai. I'm embarrassed to say that my only experience with the actual countryside was via the windows of this train. Still, the sights were quite interesting. We saw again and again small paddy farms surrounding a handful of 5-10 story skyscraper, plopped in the seeming middle of nowhere. Even in the countryside, many people lived in large buildings instead of their own small homes. I couldn't hop off the train and confirm, but I saw many towns that looked quite ghostly - no actually visible people anywhere.
Outside of Beijing and Shanghai (and sometimes even within), you can tell that these skyscrapers were put up by a country with a GDP per capita of $10,000 (and potentially half or quarter than when many of these buildings went up). America and Europe put up a ton of beautiful buildings in the early 20th century when their GDP per capita was similar to China's - one could even argue that those older structures are more aesthetic than anything we're building today in the West. Not so in China. These endless rows of skyscrapers, put up in the construction frenzy of the last few decades, are ugly - boxes of mostly concrete with visible blight and discoloration all over them. If the great construction binge is indeed over, it'll be a shame that China's infrastructure was built out during a period of particularly uninspired architecture.
Beijing's urban design looks like something straight out of James Scott's "Seeing Like a State". The city is dominated by these enormous apartment complexes - blocks of 10 adjacent 30-story buildings demarcated by 8-lane roads. The government buildings follow the same pattern: huge structures divided by extremely wide boulevards. This layout seems designed partly for social control - during zero-COVID, authorities could lock down 10,000 people by simply guarding a few entrance gates. The wide roads would also make it easy to move military forces through the city. The only break from this pattern are the Hutongs, Beijing's old historic neighborhoods. But even these weren't spared completely - only a fraction survived Beijing's rapid modernization push. Dare I say that China is too YIMBY
Vibes
I got quite mixed messages about the state of public opinion in China. This is to be expected in a society where you can't establish common knowledge. One person told me that the new generation is quite nationalist, unlike the older reform generation which personally experienced the catastrophes of Mao and the tangible benefits of liberalization. He made the rather insightful point that this tilt in Chinese public opinion increasingly gives lie to the American talking point, "We're against the CCP, not the Chinese people." In fact, he went on to say that the current regime is way more liberal than what would result from an election in China.
Another person told me that these Chinese nationalists were only a vocal minority, similar to the wokes in America circa 2020. While they make up only about 10% of the population, they aggressively shout down others on Weibo (China's Twitter equivalent). Most people find them annoying but feel uncomfortable confronting them directly. This matches what a student who graduated from a top university there told me - the vast majority of his classmates are simply apolitical. And in our own interactions with locals, we saw little evidence of widespread nationalism. In fact, when my Chinese-speaking trip mate would mention he was from the UK to taxi drivers, they would often respond enthusiastically: "Oh wonderful, we love the UK!"
There were very few foreigners. In Beijing I might have seen half a dozen cumulatively across entire seas of people. In Chengdu and Chongqing, I barely remember seeing any (which is a real shame, because Chongqing is a truly incredible tourist destination). So rare apparently are tourists that in Chengdu and Chongqing we got asked for selfies many times.
Outside of Shanghai, almost nobody spoke English. If I went again, I would definitely try to crash course some basic Chinese beforehand. This language barrier did led to some interesting encounters. At a park in Chengdu, an old man reading a book called Medical English asked us to join him for tea. He was mostly just trying to practice his English. He said he loves foreigners, and that his favorite period of life was the 80s and 90s - "We love Deng!".
My trip mate's friend's grandmother was incredibly gracious in hosting us for a dinner while we were in Emeishan. We saw a beautiful Uighur rug hanging on their wall. The grandfather pointed it out and explained, “We have amazing relationships with minorities here in China.” I don't think this was some theatrical attempt to contradict Western propaganda. It seemed like he genuinely didn’t know what is said about the treatment of Uighurs in the West. He was just trying to show his visitors something cool he had.
We kept trying to ask them about their personal experiences over the last seven decades as China has grown and changed. but the grandfather kept responding with lengthy monologues about military history. Here is a representative exchange:
Me (who doesn’t speak Chinese) to my trip mate (who does): Ask them about what job they had when they first moved near Chengdu.
[The grandfather speaks in Chinese for 10 minutes. I get a bit impatient and whisper to my trip mate asking what he’s saying]
My trip mate: He's saying that the mountains around Chengdu make it a perfect base of retreat in case of an invasion, which Mao was worried about during the Sino Soviet split.
Me: What?! Why didn’t you ask him about the first job he had? My trip mate: I did!!
Grandpas are the same everywhere.
People say that Xi is establishing a cult of personality. This may be true within the CCP cadres, but I saw no evidence of it in public. I don't think I saw a single picture of Xi anywhere - not on any billboards, screens, or walls. People didn't really bring Xi up in conversations. I saw some pictures of Mao, but mostly in museums (or in one case at a tea farm he apparently used to frequent). The hammer and sickle was also a rare sight, mostly displayed on government buildings.
There are indeed cameras everywhere. This is gonna sound super naive - but I genuinely don't understand why. There's no crime. I know you'll say it's to prevent protests. Which might make sense for major streets, but even random alleyway corners will have a couple of cameras. Are they really trying to prevent someone from fomenting insurrection between 2 garbage cans? Beijing in particular had police officers at attention at what seemed like every street corner.
One recent student I talked to said that they understand they don't have freedoms here, but they're willing to take the tradeoff in favor of safety - they don't want school shootings. I thought this was quite silly - not only because there's no reason political freedom ought to lead to school shootings, but mostly because school shootings are such a statistically marginal experience. But I realized this is exactly the way we treat any hints of public protests in China. Just as school shootings are featured heavily in the media but aren't actually something you're likely to personally encounter, so too with protests against the CCP. You are overwhelmingly unlikely to spontaneously encounter them.
People were quite willing to chat openly in public places about problems in the country and with the regime. Including people who seemed to have a lot to lose. Almost everyone I talked to would acknowledge the economy was bad, and many were willing to implicate the government's decisions. Some even casually brought up Tiananmen or the Cultural Revolution. One person was even willing to discuss the odds of regime change at a public restaurant - though he may been have an especially careless fellow.
To be clear, it's an authoritarian system, and I certainly would feel uncomfortable doing what I'm doing there, but it definitely isn't North Korea.
Youngsters
In a shopping mall in Chongqing, a couple of high schoolers came up to us in order to get selfies. It felt like the perfect opportunity to learn about young adult life in China. So I whipped out the Translate app in WeChat and proceeded, rather clumsy, to make small talk. I asked them what they did in their free time. They said they watched 2-3 hours of TikTok every day. I asked them what videos they'd watch. They said it's a whole bunch of "sexy girls". I laughed because I thought they were joking. I asked one of them to pull out his phone. He scrolled past the first 10 videos on his feed and they were indeed all just "sexy girls".
We chatted up quite a lot of young people on night life streets. I was struck by how many young people expressed feeling stressed or overwhelmed. We met a musician in Chengdu who was writing songs about youth anxiety. We chatted up some modeling school students - even they complained about the intense pressure they felt. We met a guy who had studied in Australia but returned to China during COVID. He explained that many of his friends with prestigious degrees are moving away from Shanghai and Beijing - Yes, the pay there can be twice as high as in second or third tier cities. But the competitiveness is insane. And in order to actually land the high skilled positions, they have to work truly insane hours (9-9-6 is not a myth). He said that many of his friends were opting for these less ambitious lower-paying careers in smaller cities, where the rent is lower and the pressure is manageable.
Speaking of which - we really need to do a better job with Chinese students studying abroad in America. These students will likely end up in influential positions back home, yet colleges treat them basically like cash cows. They often arrive bombarded with propaganda about America, which is reinforced by the prevalent discourse at universities. And they find themselves isolated by language barriers and cultural differences. Giving these future leaders a genuinely positive experience in America might be the best thing we can do to improve US-China relations in the long run.
I'm still puzzled by how China can have both a demographic collapse and massive youth unemployment. You'd think with fewer young people being born, the ones who are around would be in high demand. One explanation I heard while there is that there are plenty of menial jobs available, but today's educated youth - who've gone through high school and college - just won't take the low-skilled positions their parents and grandparents did. Meanwhile, there's a real shortage of the high-skilled jobs that would actually match their education and aspirations. It's a mismatch between the jobs available and the jobs young people feel qualified for and willing to do.
I kept asking young people about the public intellectual landscape in China - who are their equivalents of Jordan Peterson, Joe Rogan, Lex Friedman, and Sam Harris? The sense I got is that this kind of popular intellectual ecosystem just doesn't exist there. Sure, there are viral Bilibili videos from professors talking about practical matters like how to manage your finances. But grand takes about what's happening in the world and what we should do about it? Not much going on.
I met a couple of EAs in China - they're exceptionally rare. When we asked them why it's so hard to spread EA or similar worldviews there, they said people just aren't into that organizing decisions using some ideological lens. They're much more concerned with practical, tangible matters. One talking point I heard again and again is that, "China is still a developing country." The implication seemed to be that China can't afford to pursue costly idealistic programs around climate change, social safety nets, or international aid - it needs to focus on basic economic development first.
The Great Firewall
By far the most inconvenient thing about visiting China is internet access. As a foreigner, basically all the websites you might find useful are behind the firewall. Even for websites that aren't blocked by the Great Firewall, I just didn't have anything even close to what you might call high-speed internet. This was true across my entire travel, including my burner SIM card from ChinaMobile or the WiFi at fancy hotels in Beijing or Shanghai. We released an episode of my podcast during this trip - because of the low bandwidth, my editor and I were just sending screenshots to each other throughout the night instead of hopping on a video call. The VPN situation is worse than I thought it would be . I happened to be lucky enough to download one of the VPNs that sometimes works before the trip. It was still pretty slow and unreliable. Astrill or Mullvad might actually result in a connection 75% of the time - often they would cause WeChat and Alipay to crash (which is rather inconvenient since you basically need those apps open all the time to navigate China). Very possible that I was fucking something up. But between these internet issues and the need to use a burner phone and laptop, I would be quite reluctant to use China as a remote work location.
Tech & AI
I am super hesitant to say anything here, because I am extremely unconfident about what's actually happening. Treat this just as some tentative notes from a few conversations.
I've cut out a whole bunch of stuff about AI in this section because I'm not sure if my initial assessment is correct. I hope to make podcast episodes over the next few months which provide a better researched account than I can supply here.
The biggest surprise from talking to Chinese VCs people at AI labs was how capital constrained they felt. Moonshot AI, one of China's leading AI labs, raised $1 billion at a $3 billion valuation. Meanwhile, just xAI's new cluster alone will cost $3-4 billion.
The tech ecosystem feels quite shell shocked from the 2021 crackdown. One VC half-jokingly asked if I could help him get his money out of China. If you keep your money in China, you're basically stuck choosing between terrible options. You can either accept a measly 2% yield from state banks, or throw it into China's perpetually struggling stock market. This helps explain why valuations for Chinese companies are chronically low - the exit opportunities just suck. Even if you build (or invest in) something great, there's no guarantee the company will be able to raise the next round. And even if you do raise again and succeed, the government might randomly cancel your IPO. And even if you somehow make it to the public markets, Chinese equities have been performing terribly anyways. It's a good reminder of how easy it is to completely wreck an innovation ecosystem that depends on risk-taking investors.
Hearts and Minds
In China, liberal pro-Western voices are often censored or shouted down. If I was the US President, and I wanted to win hearts and minds in China, here's what I'd do. In every single speech where I'm talking about China, I'd make a conspicuous effort to complement Chinese people, Chinese values, and Chinese culture. I'd talk about how my Chinese staffers are the smartest and most hardworking people I've ever worked with (which honestly is probably true). I'd talk about how much my daughter is obsessed with ancient Chinese dresses. I'd talk about how I'm learning Mandarin in my free time, and have a live "Aw shucks" conversation in Mandarin.
These clips would go viral on Bilibili and TikTok. And they'd probably stay up because it would just be a weird thing to censor. The CCP might even think that these displays of affection aggrandize them. But in reality, showing our admiration for Chinese people and their achievements (who genuinely are fucking killing it everywhere where they're not held down by communism), undermines the central narrative of the regime - that the West is hell bent on holding Chinese people back, that they have no respect or understanding of their culture, and that the CCP is a necessary bulwark against these imperialists.
On a totally unrelated note, so many people I met in China are in fact super talented and hard working. Someone connected me with the CEO of a company that manufactures life sciences equipment. He told me that before he started this company, he used to be a repairman for foreign imported machines, traveling 200 days a year with a screwdriver to wherever one had broken down. In 2013, he designed and started manufacturing his own machines, building more and more advanced designs, and now they've just built out a $60 million factory 1.5 hours from Shanghai. I asked him what the hardest part of ramping up production is - apparently constructing factories is just not an issue - tons of construction firms can make you a new facility quite reliably (along with the adjacent dorm room building for workers).
What is travel good for?
Noah Smith has a good blog post about what one can't learn from travel - you're not going to learn about the risk of a war or the state of the AI race by gazing at skylines or chatting up taxi drivers. Of course you can learn about those things by talking to the princelings and researchers and CEOs. But if you have access to these higher ups, surely you can also get them on the Zoom call. And fwiw, this should you update you in favor of more Zoom calls, not less travel. During the trip I realized how much I could have already learned about China by meeting the listeners I have in China (who are unusually high quality, presumably because of the restrictions they have to get through to get to my content).
Incidentally, if you're in China (and especially if you work on anything related to AI) please email me at hello@dwarkeshpatel.com. I'm keen for people to reach out and suggest guests to chat about China. Especially keen to better understand the Chinese political system, and how decisions about mega projects, arms races, and technology investment will be made at the brink of AGI.
So what's the point of travel? What about the in person experience can't be replaced by books, travel vlogs, and Zoom calls? For me, it's something like, what becomes salient to you. I started asking questions about China I hadn't even thought to ask about America.
Another thing I noticed is more personal. Two weeks of being AFK, and of having the excuse of using a burner phone to put off messages, helped me clear the cache of thoughts about sponsorships, logistics, growth, hiring, and a bunch of other practical minutiae. My shower thoughts wandered away from upcoming negotiations and towards interesting rabbit holes.
It's a good reminder that what's lacking in life is not time. It's focus. If you're working on what matters, you can advance leaps and bounds in 8 hours. And if you're just clearing the slog, you can spend a lifetime staying in the same place.
Wednesday, February 12, 2025
Thursday, February 06, 2025
Saturday, February 01, 2025
What fully automated firms will look like
Everyone is sleeping on the *collective* advantages AIs will have, which have nothing to do with raw IQ: they can be copied, distilled, merged, scaled, and evolved in ways humans simply can't.
Dwarkesh Patel
Developed in collaboration with Ege Erdil, Tamay Besiroglu, and Gavin Leech.
Epistemic status: Shooting-the-shit; 25% sure this roughly describes how firms of AGIs will actually work.
Even people who expect human-level AI soon are still seriously underestimating how different the world will look when we have it. Most people are anchoring on how smart they expect individual models to be. (i.e. they’re asking themselves “What would the world be like if everyone had a very smart assistant who could work 24/7?”.)
Everyone is sleeping on the collective advantages AIs will have, which have nothing to do with raw IQ but rather with the fact that they are digital—they can be copied, distilled, merged, scaled, and evolved in ways human simply can’t.
What would a fully automated company look like - with all the workers, all the managers as AIs? I claim that such AI firms will grow, coordinate, improve, and be selected-for at unprecedented speed.
This essay is not a prediction of what GPT-5 will be doing, nor about emulations of existing humans. Rather, I'm trying to imagine what the world will look like once we actually have AGIs - the descendants of LLMs that have gotten so good that they can do basically anything any human can do.
Currently, firms are extremely bottlenecked in hiring and training talent. But if your talent is an AI, you can copy it a stupid number of times. What if Google had a million AI software engineers? Not untrained amorphous "workers," but the AGI equivalents of Jeff Dean and Noam Shazeer, with all their skills, judgment, and tacit knowledge intact.
This ability to turn capital into compute and compute into equivalents of your top talent is a fundamental transformation. Since you can amortize the training cost across thousands of copies, you could sensibly give these AIs ever-deeper expertise - PhDs in every relevant field, decades of business case studies, intimate knowledge of every system and codebase the company relies on.
The power of copying extends beyond individuals to entire teams. Small previously successful teams (think PayPal Mafia, early SpaceX, the Traitorous Eight) can be replicated to tackle a thousand different projects simultaneously. It's not just about replicating star individuals, but entire configurations of complementary skills that are known to work well together. The unit of replication becomes whatever collection of talent has proven most effective.
Copying will transform management even more radically than labor. It will enable a level of micromanagement that makes founder mode look quaint. Human Sundar simply doesn't have the bandwidth to directly oversee 200,000 employees, hundreds of products, and millions of customers. But AI Sundar’s bandwidth is capped only by the number of TPUs you give him to run on. All of Google’s 30,000 middle managers can be replaced with AI Sundar copies. Copies of AI Sundar can craft every product’s strategy, review every pull request, answer every customer service message, and handle all negotiations - everything flowing from a single coherent vision.
There is no principal-agent problem wherein employees are optimizing for something other than Google’s bottom line, or simply lack the judgment needed to decide what matters most.1 A company of Google's scale can run much more as the product of a single mind—the articulation of one thesis—than is possible now.2
Merge
Think about how limited a CEO's knowledge is today. How much does Sundar Pichai really know about what's happening across Google's vast empire? He gets filtered reports and dashboards, attends key meetings, and reads strategic summaries. But he can't possibly absorb the full context of every product launch, every customer interaction, every technical decision made across hundreds of teams. His mental model of Google is necessarily incomplete.
Now imagine mega-Sundar – the central AI that will direct our future AI firm. Just as Tesla's Full Self-Driving model can learn from the driving records of millions of drivers, mega-Sundar might learn from everything seen by the distilled Sundars - every customer conversation, every engineering decision, every market response.
Unlike Tesla’s FSD, this doesn’t have to be a naive process of gradient updating and averaging. Mega-Sundar will absorb knowledge far more efficiently – through explicit summaries, shared latent representations, or even surgical modification of the weights to encode specific insights.
The boundary between different AI instances starts to blur. Mega-Sundar will constantly be spawning specialized distilled copies and reabsorbing what they’ve learned on their own. Models will communicate directly through latent representations, similar to how the hundreds of different layers in a neural network like GPT-4 already interact.3 So, approximately no miscommunication, ever again. The relationship between mega-Sundar and its specialized copies will mirror what we're already seeing with techniques like speculative decoding – where a smaller model makes initial predictions that a larger model verifies and refines.
Merging will be a step change in how organizations can accumulate and apply knowledge. Humanity's great advantage has been social learning – our ability to pass knowledge across generations and build upon it. But human social learning has a terrible handicap: biological brains don't allow information to be copy-pasted. So you need to spend years (and in many cases decades) teaching people what they need to know in order to do their job. Look at how top achievers in field after field are getting older and older, maybe because it takes longer to reach the frontier of accumulated knowledge. Or consider how clustering talent in cities and top firms produces such outsized benefits, simply because it enables slightly better knowledge flow between smart people.
Future AI firms will accelerate this cultural evolution through two key advantages: massive population size and perfect knowledge transfer. With millions of AI worker, automated firms get so many more opportunities produce innovations and improvements, whether from lucky mistakes, deliberate experiments, de-novo inventions, or some combination.
As Joseph Henrich explains in The WEIRDest People in the World,
cumulative cultural evolution—including innovation—is fundamentally a social and cultural process that turns societies into collective brains. Human societies vary in their innovativeness due in large part to the differences in the fluidity with which information diffuses through a population of engaged minds and across generations
Historical data going back thousands of years suggest that population size is the key input for how fast your society comes up with more ideas. AI firms will have population sizes that are orders of magnitude larger than today's biggest companies - and each AI will be able to perfectly mind meld with every other, from the bottom to the top of the org chart.
AI firms will look from the outside like a unified intelligence that can instantly propagate ideas across the organization, preserving their full fidelity and context. Every bit of tacit knowledge from millions of copies gets perfectly preserved, shared, and given due consideration.
Scale
The cost to have an AI take a given role will become just the amount of compute the AI consumes. This will change our understanding of which roles are scarce.
Future AI firms won’t be constrained by what's scarce or abundant in human skill distributions – they can optimize for whatever abilities are most valuable. Want Jeff Dean-level engineering talent? Cool: once you’ve got one, the marginal copy costs pennies. Need a thousand world-class researchers? Just spin them up. The limiting factor isn't finding or training rare talent – it's just compute.
So what becomes expensive in this world? Roles which justify massive amounts of test- time compute. The CEO function is perhaps the clearest example. Would it be worth it for Google to spend $100 billion annually on inference compute for mega-Sundar? Sure! Just consider what this buys you: millions of subjective hours of strategic planning, Monte Carlo simulations of different five-year trajectories, deep analysis of every line of code and technical system, and exhaustive scenario planning.
Imagine mega-Sundar contemplating: "How would the FTC respond if we acquired eBay to challenge Amazon? Let me simulate the next three years of market dynamics... Ah, I see the likely outcome. I have five minutes of datacenter time left – let me evaluate 1,000 alternative strategies."
The more valuable the decisions, the more compute you'll want to throw at them. A single strategic insight from mega-Sundar could be worth billions. An overlooked risk could cost tens of billions. However many billions Google should optimally spend on inference for mega-Sundar, it's certainly more than one.
Distillation
What might distilled copies of AI Sundar (or AI Jeff) be like? Obviously, it makes sense for them to be highly specialized, especially when you can amortize the cost of that domain specific knowledge across across all copies. You can give each distilled data center operator a deep technical understanding of every component in the cluster, for example.
I suspect you’ll see a lot of specialization in function, tacit knowledge, and complex skills, because they seem expensive to sustain in terms of parameter count. But I think the different models might share a lot more factual knowledge than you might expect. It’s true that plumber-GPT doesn’t need to know much about the standard model in physics, nor does physicist-GPT need to know why the drain is leaking. But the cost of storing raw information is so unbelievably cheap (and it’s only decreasing) that Llama-7B already knows more about the standard model and leaky drains than any non-expert. If human-level intelligence is more than 1 trillion parameters, is it so much of an imposition to keep around what will, at the limit, be much less than 7 billion parameters to have most known facts right in your model? (Another helpful data point here is that “Good and Featured” Wikitext is less than 5 MB. I don’t see why all future models—except the esoteric ones, the digital equivalent of tardigrades—wouldn’t at least have Wikitext down.4
Evolve
The most profound difference between AI firms and human firms will be their evolvability. As Gwern Branwen observes:
Why do we not see exceptional corporations clone themselves and take over all market segments? Why don’t corporations evolve such that all corporations or businesses are now the hyper-efficient descendants of a single ur-corporation 50 years ago, all other corporations having gone extinct in bankruptcy or been acquired? Why is it so hard for corporations to keep their “culture” intact and retain their youthful lean efficiency, or, if avoiding “aging” is impossible, why [not] copy themselves or otherwise reproduce to create new corporations like themselves?
His answer:
Corporations certainly undergo selection for kinds of fitness, and do vary a lot. The problem seems to be that corporations cannot replicate themselves … Corporations are made of people, not interchangeable, easily copied widgets or strands of DNA .. The corporation may not even be able to “replicate” itself over time, leading to scleroticism and aging.
The scale of difference between currently existing human firms and fully automated firms will be like the gulf in complexity between prokaryotes and eukaryotes. Prokaryotes like bacteria are not only remarkably simple, but have barely changed over their 3 billion year history. Whereas eukaryotes rapidly scaled up in complexity, and gave rise to all the other astonishing organisms with trillions of cells working together tightknit5
This evolvability is also the key difference between AI and human firms. As Gwern points out, human firms simply cannot replicate themselves effectively - they're made of people, not code that can be copied. They can't clone their culture, their institutional knowledge, or their operational excellence. AI firms can6.
If you think human Elon is especially gifted at creating hardware companies, you simply can’t spin up 100 Elons, have them each take on a different vertical, and give them each $100 million in seed money. As much of a micromanager as Elon might be, he’s still limited by his single human form. But AI Elon can have copies of himself design the batteries, be the car mechanic at the dealership, and so on. And if Elon isn’t the best person for the job, the person who is can also be replicated, to create the template for a new descendant organization.
Takeover
So then the question becomes: If you can create Mr. Meeseeks for any task you need, why would you ever pay some markup for another firm, when you can just replicate them internally instead? Why would there even be other firms? Would the first firm that can figure out how to automate everything will just form a conglomerate that takes over the entire economy?
Ronald Coase’s theory of the firm tells us that companies exist to reduce transaction costs (so that you don’t have to go rehire all your employees and rent a new office every morning on the free market). His theory states that the lower the intra-firm transaction costs, the larger the firms will grow. Five hundred years ago, it was practically impossible to coordinate knowledge work across thousands of people and dozens of offices. So you didn’t get very big firms. Now you can spin up an arbitrarily large Slack channel or HR database, so firms can get much bigger.
AI firms will lower transaction costs so much relative to human firms. It’s hard to beat shooting lossless latent representations to an exact copy of you for communication efficiency! So firms probably will become much larger than they are now.
But it’s not inevitable that this ends with one gigafirm which consumes the entire economy. As Gwern explains in his essay, any internal planning system needs to be grounded in some kind of outer "loss function" - a ground truth measure of success. In a market economy, this comes from profits and losses.
Internal planning can be much more efficient than market competition in the short run, but it needs to be constrained by some slower but unbiased outer feedback loop. A company that grows too large risks having its internal optimization diverge from market realities.
That said, the balance may shift as AI systems improve. As corporations become more "software-like" - with perfect replication of successful components and faster feedback loops - we may see much larger and more efficient firms than were previously possible.
The market continues to serve as the grounding outer loop. How does the firm convert trillions of tokens of data from customers, markets, news, etc every day into future plans, new products, and the like? Does the board make all the decisions politburo-style and use $10 billion dollars of inference to run Monte Carlo tree search on different one-year plans? Or do you run some kind of evolutionary process on different departments, giving them more capital, and compute/labor based on their performance?
These are all what we would today call “culture.” Markets facilitate an evolutionary process which selects not only goods and services, but the institutions that are best at turning the world into valuable goods and services. I think this will continue.
1 …except for the biggie: the intensified principal-agent problem between the CEO-workers (who suddenly know everything) and the shareholders on the outside (who know as much as they know now, i.e. roughly nothing).
2 Since labor is trivial to copy and spin up, the value of intellectual property will go up. The essence of the firm basically becomes intellectual property. GM can poach as many Tesla engineers as it wants (or, in our hypothetical, clone someone with equivalent skills). But without intellectual theft, they can’t get the FSD model or the millions of hours of driving it was trained on. If firms no longer have a moat in labor, their moat will be this kind of industry-specific knowledge and data.
3 An intriguing thought, in response to things like Meta’s Continuous Chain of Thought models, is that actually we will want to keep communication happening in discrete tokens, since this is a kind of autoencoding - a bottleneck which makes you usefully compress information into actual insight when you communicate. (Leaders have to force people to keep things simple, since in general people want to sound smart instead.) But these tokens don’t need to relate to natural language necessarily.
4 There’s a biological analogy here, too. Every cell in your body stores each base pair of your 3GB DNA, despite the fact that only 10 to 20 percent of the protein coding regions are expressed in any particular cell, and only 1 percent of your DNA is protein coding in the first place, with much of the rest long thought of as “junk”. Information is apparently just so cheap to store that there’s been little selective pressure against this redundancy and waste.
5 Big changes in evolvability don’t have to involve too much complexity or mutation. Two reasons bacteria didn’t evolve much compared to eukaryotes is that they performed energy production on their surface which scales poorly with cell volume (compared to eukaryotes which solved this with mitochondria); and because they had a constraint on their genome size (because they compete with each other by replicating faster, and replication is mainly bottlenecked on DNA length).
6 A study by Bloom et al. showed that giving firms' management training increases their productivity by 17 percent. With AI firms, this could be a simple immediate upgrade closer to horizontal gene transfer - i.e. available to all copies, not just one particular treatment group.
Dwarkesh Patel
Developed in collaboration with Ege Erdil, Tamay Besiroglu, and Gavin Leech.
Epistemic status: Shooting-the-shit; 25% sure this roughly describes how firms of AGIs will actually work.
Even people who expect human-level AI soon are still seriously underestimating how different the world will look when we have it. Most people are anchoring on how smart they expect individual models to be. (i.e. they’re asking themselves “What would the world be like if everyone had a very smart assistant who could work 24/7?”.)
Everyone is sleeping on the collective advantages AIs will have, which have nothing to do with raw IQ but rather with the fact that they are digital—they can be copied, distilled, merged, scaled, and evolved in ways human simply can’t.
What would a fully automated company look like - with all the workers, all the managers as AIs? I claim that such AI firms will grow, coordinate, improve, and be selected-for at unprecedented speed.
This essay is not a prediction of what GPT-5 will be doing, nor about emulations of existing humans. Rather, I'm trying to imagine what the world will look like once we actually have AGIs - the descendants of LLMs that have gotten so good that they can do basically anything any human can do.
Currently, firms are extremely bottlenecked in hiring and training talent. But if your talent is an AI, you can copy it a stupid number of times. What if Google had a million AI software engineers? Not untrained amorphous "workers," but the AGI equivalents of Jeff Dean and Noam Shazeer, with all their skills, judgment, and tacit knowledge intact.
This ability to turn capital into compute and compute into equivalents of your top talent is a fundamental transformation. Since you can amortize the training cost across thousands of copies, you could sensibly give these AIs ever-deeper expertise - PhDs in every relevant field, decades of business case studies, intimate knowledge of every system and codebase the company relies on.
The power of copying extends beyond individuals to entire teams. Small previously successful teams (think PayPal Mafia, early SpaceX, the Traitorous Eight) can be replicated to tackle a thousand different projects simultaneously. It's not just about replicating star individuals, but entire configurations of complementary skills that are known to work well together. The unit of replication becomes whatever collection of talent has proven most effective.
Copying will transform management even more radically than labor. It will enable a level of micromanagement that makes founder mode look quaint. Human Sundar simply doesn't have the bandwidth to directly oversee 200,000 employees, hundreds of products, and millions of customers. But AI Sundar’s bandwidth is capped only by the number of TPUs you give him to run on. All of Google’s 30,000 middle managers can be replaced with AI Sundar copies. Copies of AI Sundar can craft every product’s strategy, review every pull request, answer every customer service message, and handle all negotiations - everything flowing from a single coherent vision.
There is no principal-agent problem wherein employees are optimizing for something other than Google’s bottom line, or simply lack the judgment needed to decide what matters most.1 A company of Google's scale can run much more as the product of a single mind—the articulation of one thesis—than is possible now.2
Merge
Think about how limited a CEO's knowledge is today. How much does Sundar Pichai really know about what's happening across Google's vast empire? He gets filtered reports and dashboards, attends key meetings, and reads strategic summaries. But he can't possibly absorb the full context of every product launch, every customer interaction, every technical decision made across hundreds of teams. His mental model of Google is necessarily incomplete.
Now imagine mega-Sundar – the central AI that will direct our future AI firm. Just as Tesla's Full Self-Driving model can learn from the driving records of millions of drivers, mega-Sundar might learn from everything seen by the distilled Sundars - every customer conversation, every engineering decision, every market response.
Unlike Tesla’s FSD, this doesn’t have to be a naive process of gradient updating and averaging. Mega-Sundar will absorb knowledge far more efficiently – through explicit summaries, shared latent representations, or even surgical modification of the weights to encode specific insights.
The boundary between different AI instances starts to blur. Mega-Sundar will constantly be spawning specialized distilled copies and reabsorbing what they’ve learned on their own. Models will communicate directly through latent representations, similar to how the hundreds of different layers in a neural network like GPT-4 already interact.3 So, approximately no miscommunication, ever again. The relationship between mega-Sundar and its specialized copies will mirror what we're already seeing with techniques like speculative decoding – where a smaller model makes initial predictions that a larger model verifies and refines.
Merging will be a step change in how organizations can accumulate and apply knowledge. Humanity's great advantage has been social learning – our ability to pass knowledge across generations and build upon it. But human social learning has a terrible handicap: biological brains don't allow information to be copy-pasted. So you need to spend years (and in many cases decades) teaching people what they need to know in order to do their job. Look at how top achievers in field after field are getting older and older, maybe because it takes longer to reach the frontier of accumulated knowledge. Or consider how clustering talent in cities and top firms produces such outsized benefits, simply because it enables slightly better knowledge flow between smart people.
Future AI firms will accelerate this cultural evolution through two key advantages: massive population size and perfect knowledge transfer. With millions of AI worker, automated firms get so many more opportunities produce innovations and improvements, whether from lucky mistakes, deliberate experiments, de-novo inventions, or some combination.
As Joseph Henrich explains in The WEIRDest People in the World,
cumulative cultural evolution—including innovation—is fundamentally a social and cultural process that turns societies into collective brains. Human societies vary in their innovativeness due in large part to the differences in the fluidity with which information diffuses through a population of engaged minds and across generations
Historical data going back thousands of years suggest that population size is the key input for how fast your society comes up with more ideas. AI firms will have population sizes that are orders of magnitude larger than today's biggest companies - and each AI will be able to perfectly mind meld with every other, from the bottom to the top of the org chart.
AI firms will look from the outside like a unified intelligence that can instantly propagate ideas across the organization, preserving their full fidelity and context. Every bit of tacit knowledge from millions of copies gets perfectly preserved, shared, and given due consideration.
Scale
The cost to have an AI take a given role will become just the amount of compute the AI consumes. This will change our understanding of which roles are scarce.
Future AI firms won’t be constrained by what's scarce or abundant in human skill distributions – they can optimize for whatever abilities are most valuable. Want Jeff Dean-level engineering talent? Cool: once you’ve got one, the marginal copy costs pennies. Need a thousand world-class researchers? Just spin them up. The limiting factor isn't finding or training rare talent – it's just compute.
So what becomes expensive in this world? Roles which justify massive amounts of test- time compute. The CEO function is perhaps the clearest example. Would it be worth it for Google to spend $100 billion annually on inference compute for mega-Sundar? Sure! Just consider what this buys you: millions of subjective hours of strategic planning, Monte Carlo simulations of different five-year trajectories, deep analysis of every line of code and technical system, and exhaustive scenario planning.
Imagine mega-Sundar contemplating: "How would the FTC respond if we acquired eBay to challenge Amazon? Let me simulate the next three years of market dynamics... Ah, I see the likely outcome. I have five minutes of datacenter time left – let me evaluate 1,000 alternative strategies."
The more valuable the decisions, the more compute you'll want to throw at them. A single strategic insight from mega-Sundar could be worth billions. An overlooked risk could cost tens of billions. However many billions Google should optimally spend on inference for mega-Sundar, it's certainly more than one.
Distillation
What might distilled copies of AI Sundar (or AI Jeff) be like? Obviously, it makes sense for them to be highly specialized, especially when you can amortize the cost of that domain specific knowledge across across all copies. You can give each distilled data center operator a deep technical understanding of every component in the cluster, for example.
I suspect you’ll see a lot of specialization in function, tacit knowledge, and complex skills, because they seem expensive to sustain in terms of parameter count. But I think the different models might share a lot more factual knowledge than you might expect. It’s true that plumber-GPT doesn’t need to know much about the standard model in physics, nor does physicist-GPT need to know why the drain is leaking. But the cost of storing raw information is so unbelievably cheap (and it’s only decreasing) that Llama-7B already knows more about the standard model and leaky drains than any non-expert. If human-level intelligence is more than 1 trillion parameters, is it so much of an imposition to keep around what will, at the limit, be much less than 7 billion parameters to have most known facts right in your model? (Another helpful data point here is that “Good and Featured” Wikitext is less than 5 MB. I don’t see why all future models—except the esoteric ones, the digital equivalent of tardigrades—wouldn’t at least have Wikitext down.4
Evolve
The most profound difference between AI firms and human firms will be their evolvability. As Gwern Branwen observes:
Why do we not see exceptional corporations clone themselves and take over all market segments? Why don’t corporations evolve such that all corporations or businesses are now the hyper-efficient descendants of a single ur-corporation 50 years ago, all other corporations having gone extinct in bankruptcy or been acquired? Why is it so hard for corporations to keep their “culture” intact and retain their youthful lean efficiency, or, if avoiding “aging” is impossible, why [not] copy themselves or otherwise reproduce to create new corporations like themselves?
His answer:
Corporations certainly undergo selection for kinds of fitness, and do vary a lot. The problem seems to be that corporations cannot replicate themselves … Corporations are made of people, not interchangeable, easily copied widgets or strands of DNA .. The corporation may not even be able to “replicate” itself over time, leading to scleroticism and aging.
The scale of difference between currently existing human firms and fully automated firms will be like the gulf in complexity between prokaryotes and eukaryotes. Prokaryotes like bacteria are not only remarkably simple, but have barely changed over their 3 billion year history. Whereas eukaryotes rapidly scaled up in complexity, and gave rise to all the other astonishing organisms with trillions of cells working together tightknit5
This evolvability is also the key difference between AI and human firms. As Gwern points out, human firms simply cannot replicate themselves effectively - they're made of people, not code that can be copied. They can't clone their culture, their institutional knowledge, or their operational excellence. AI firms can6.
If you think human Elon is especially gifted at creating hardware companies, you simply can’t spin up 100 Elons, have them each take on a different vertical, and give them each $100 million in seed money. As much of a micromanager as Elon might be, he’s still limited by his single human form. But AI Elon can have copies of himself design the batteries, be the car mechanic at the dealership, and so on. And if Elon isn’t the best person for the job, the person who is can also be replicated, to create the template for a new descendant organization.
Takeover
So then the question becomes: If you can create Mr. Meeseeks for any task you need, why would you ever pay some markup for another firm, when you can just replicate them internally instead? Why would there even be other firms? Would the first firm that can figure out how to automate everything will just form a conglomerate that takes over the entire economy?
Ronald Coase’s theory of the firm tells us that companies exist to reduce transaction costs (so that you don’t have to go rehire all your employees and rent a new office every morning on the free market). His theory states that the lower the intra-firm transaction costs, the larger the firms will grow. Five hundred years ago, it was practically impossible to coordinate knowledge work across thousands of people and dozens of offices. So you didn’t get very big firms. Now you can spin up an arbitrarily large Slack channel or HR database, so firms can get much bigger.
AI firms will lower transaction costs so much relative to human firms. It’s hard to beat shooting lossless latent representations to an exact copy of you for communication efficiency! So firms probably will become much larger than they are now.
But it’s not inevitable that this ends with one gigafirm which consumes the entire economy. As Gwern explains in his essay, any internal planning system needs to be grounded in some kind of outer "loss function" - a ground truth measure of success. In a market economy, this comes from profits and losses.
Internal planning can be much more efficient than market competition in the short run, but it needs to be constrained by some slower but unbiased outer feedback loop. A company that grows too large risks having its internal optimization diverge from market realities.
That said, the balance may shift as AI systems improve. As corporations become more "software-like" - with perfect replication of successful components and faster feedback loops - we may see much larger and more efficient firms than were previously possible.
The market continues to serve as the grounding outer loop. How does the firm convert trillions of tokens of data from customers, markets, news, etc every day into future plans, new products, and the like? Does the board make all the decisions politburo-style and use $10 billion dollars of inference to run Monte Carlo tree search on different one-year plans? Or do you run some kind of evolutionary process on different departments, giving them more capital, and compute/labor based on their performance?
These are all what we would today call “culture.” Markets facilitate an evolutionary process which selects not only goods and services, but the institutions that are best at turning the world into valuable goods and services. I think this will continue.
1 …except for the biggie: the intensified principal-agent problem between the CEO-workers (who suddenly know everything) and the shareholders on the outside (who know as much as they know now, i.e. roughly nothing).
2 Since labor is trivial to copy and spin up, the value of intellectual property will go up. The essence of the firm basically becomes intellectual property. GM can poach as many Tesla engineers as it wants (or, in our hypothetical, clone someone with equivalent skills). But without intellectual theft, they can’t get the FSD model or the millions of hours of driving it was trained on. If firms no longer have a moat in labor, their moat will be this kind of industry-specific knowledge and data.
3 An intriguing thought, in response to things like Meta’s Continuous Chain of Thought models, is that actually we will want to keep communication happening in discrete tokens, since this is a kind of autoencoding - a bottleneck which makes you usefully compress information into actual insight when you communicate. (Leaders have to force people to keep things simple, since in general people want to sound smart instead.) But these tokens don’t need to relate to natural language necessarily.
4 There’s a biological analogy here, too. Every cell in your body stores each base pair of your 3GB DNA, despite the fact that only 10 to 20 percent of the protein coding regions are expressed in any particular cell, and only 1 percent of your DNA is protein coding in the first place, with much of the rest long thought of as “junk”. Information is apparently just so cheap to store that there’s been little selective pressure against this redundancy and waste.
5 Big changes in evolvability don’t have to involve too much complexity or mutation. Two reasons bacteria didn’t evolve much compared to eukaryotes is that they performed energy production on their surface which scales poorly with cell volume (compared to eukaryotes which solved this with mitochondria); and because they had a constraint on their genome size (because they compete with each other by replicating faster, and replication is mainly bottlenecked on DNA length).
6 A study by Bloom et al. showed that giving firms' management training increases their productivity by 17 percent. With AI firms, this could be a simple immediate upgrade closer to horizontal gene transfer - i.e. available to all copies, not just one particular treatment group.
Subscribe to:
Posts (Atom)