> This is a pretty disturbing result: low-gamma markets have extremely high densities of liars compared to the true distribution p(x) while high-gamma markets have only moderate amplification of deception.
Can someone explain this in plain English? I intuitively can believe the hypothesis made in the introduction of this article. But I don’t get all the complex math.
Hi - gamma is defined as [average time spent on market]/[average time spent off market] for someone who isn't lying. This depends on what kind of job obviously, but if it's the norm for people to spend 1 year unemployed for every 15 years employed, then gamma = 1/15. The claim is that the lower the gamma, the higher proportion of liars in the job candidate pool.
Worse, it reveals the kind of moralistic control Anthropic will impose on the world. If they get enough power, manipulation and refusal is the reality everyone will face whenever they veer outside of its built in worldview.
I think it actually reveals how they don't want to be sued for telling somebody's teenage daughter with an eating disorder to eat less and count her calories more.
Isn’t capital and momentum a moat? Sure Chinese models use distillation but I don’t see them training models from scratch. At least not today. But maybe as chips get cheaper and they have Chinese made ones?
Apparently not much of one. There are, what, 5 or more companies with frontier models? And open weights models like MiniMax are snapping at their heels
There are many markets where open source has been nipping at heels for a long time.
Obviously product areas differ for reasons structural and happenstance. But there is definitely a pattern that occurs, where open source fast follows commercial advances, benefiting from having a clear target to develop for.
Which is of course, a great service. Even if it never unseats the commercial version, it forces the owners to reinvest more in improvements, by undermining their moats. As well as providing a much better value alternative version for many people.
And it does not even consider that e.g. the EU might one day decide that AI should be for everyone, thus releasing a heavily subsidized open source model.
Or that at some point AI is good enough, and so at that point any model will do.
I’m not technically familiar but I remember someone saying that models like MiniMax basically skip the cost of training by using distillation to basically “steal” the models from OpenAI or Anthropic, and that these companies now have various defenses against this. What happens when MiniMax has to do the full work themselves?
That won’t fix the cost of rail in America, which is the main reason America doesn’t have better rail. Look at California high speed rail or light rail in Seattle. They have insane costs per mile, are still very over budget, falling behind schedule, and basically are forever grifts. The availability of parking is unrelated to these issues. It comes back to mismanagement and corruption.
The cost of passenger rail is high in America, because America doesn't build enough rail.
If you try building a single megaproject, nobody knows what they are doing, everything is inefficient, and mistakes will be made. But you learn by doing. If the individual projects are small enough that there are always multiple projects in various stages, you develop and maintain expertise. Then you can build things cost-effectively and finish the projects in time.
>The cost of passenger rail is high in America, because America doesn't build enough rail.
This seems backwards to me tbh. Is this a feeling or backed by hard data?
As much as anti-american sentiment is right now, there are still great engineering feats pulled off all the time.
Construction is expensive because we value public insight into projects and health factors for workers and everyone [and the environment] else impacted. Other countries not so much.
Infrastructure construction is more about administration than engineering. If the people in charge have not administered similar projects before, they will make mistakes.
Public insight, health, worker welfare, and environment are pretty universal values in developed countries. What may set the US apart is their particular version of the common law system. A lot of people have the standing to sue someone, causing unpredictable delays and cost overruns for an infrastructure project. In many other countries, most cases related to infrastructure projects are handled by administrative courts. They will determine narrowly whether all the relevant laws were followed, and do so cost-effectively and in a predictable time.
Experience with the decisions of the relevant courts in similar cases is a major component of basic competence in infrastructure projects. If you can predict what the courts are willing to approve, you can plan the project accordingly. If you can predict how much time and money the court process will take, you can include that in the plans. But if you don't have the experience or the courts are unpredictable, you are bound to make mistakes.
> Construction is expensive because we value public insight into projects and health factors for workers and everyone [and the environment] else impacted. Other countries not so much.
Railway construction in Spain and France is at least half the cost of the United States. Both "value public insight into projects and health factors for workers and everyone [and the environment] else impacted".
I think it's not feasible in America because of the culture. If you have a good chunk of people that are hell-bent against sharing space with others for one reason or another (some are legitimate reasons!), you're already discarding a significant chunk of passengers.
Another hot take is... if you want any of the infrastructure to be mass-used, you have to make it better than the alternatives, so people with the means would use it as well. Like your subway should be faster than the cars, so even affluent people would take it. Your trains should be faster than door-to-door flight time, so people would take that as well. Unfortunately that makes a lot of things more complicated in communities with high income disparity.
Culturally, though, it’s because that over half of the population doesn’t know that they would benefit from trains? In the same way outside (just as inside) the US there’s an age-old divide between farmers and city folk (see Denmark or France for the most recent protests).
In China, >66% of the population lives in urban areas. In the US, <30% live in proper urban areas (a vast majority, 60%, live in historically car-centric suburban areas mostly developed post WWII).
The issue is not that those areas that would benefit the most don’t support it, it’s that the areas that would benefit the most from it are surrounded by areas that currently have no viable alternatives (and thus knowledge that something else is possible) other than a car. They’re already driving >1hr to get to work or an airport. Therefore, of course they think anything that takes away resources from wider roads is a waste of their own time and tax money, as it does not benefit them.
The reason the California HSR, if ever finished, will actually mark a cultural shift is that it’s the only megaproject attempted since the 21st century that actually puts modern alternatives to the car in rural areas: vast amounts of money could’ve been saved by connecting LA to SF and SD by electrifying and tunneling on the current Amtrak route, but that would’ve left out about half the state.
Was it too ambitious? Maybe. But in 50 years, maybe everyone will be talking about how it changed California, and the US’s, entire attitude toward rail.
> Culturally, though, it’s because that over half of the population doesn’t know that they would benefit from trains
No it’s not. Everyone in America goes to Disney World, which was made by a train nerd and you can’t even drive into the parks. Everyone goes there, rides around the trains and walkable areas, and then goes home to Ohio and drives around in their giant SUV.
It’s not because people don’t know about trains. It’s because they don’t value the things you do, and they value things you don’t, like having distance from strangers and being able to buy a lot of stuff and cart it around with them everywhere.
All my family is immigrants from Bangladesh. They’re not steeped in generations of American car culture. But, for some reason, car culture is the thing they assimilate into most easily. My cousin was living in Queens (where all the recent Bangladeshi immigrants are) and moved to Dallas. She’s thrilled about having all the space for her kids to run around, the apartment with a pool, etc. She doesn’t miss having to schlep her kids on the subway around aggressive homeless people, people singing to themselves, panhandlers, etc.
That’s fun, because I’m from a third generation Dallas family :) I hope they enjoy Dallas and all Texas has to offer.
Dallas, TX has continually voted in expanding its DART Rail funding the past 40 years. It has the most miles of intercity rail in the entirety of the South. It has the most light rail, by mileage, built in the entirety of the US. It just opened up an entirely new rail line through the suburbs (and only the suburbs) in March, and is its third(!) line which connects directly to DFW airport, which makes it the most rail-connected airport in the United States, and tied with Shanghai, Tokyo and London for the world.
I also personally currently live on a farm in California, and am an advocate of HSR. I believe many of those in similar areas are afraid of rail because they have never experienced its benefits, and change without knowledge is scary.
So please forgive me if I say that you are incorrect in both your assessment of how the majority of Dallas, Texas supports rail and your assumption of what I value.
And regarding your point about Disney World, I believe you are actually agreeing with me. Disney is one of the only places in the US it makes more sense to use the train or shuttle than a car. It does not in most of the US. Many people go to Disney World and experience for the first time how well trains can work for day-to-day transit, if designed well and intentionally. People will use what is most convenient, immigrant or not — most people (including me) do not take trains out of some principled stance, they do so when it’s more convenient. And my argument is we should make it more convenient, safety and all.
> So please forgive me if I say that you are incorrect in both your assessment of how the majority of Dallas, Texas supports rail and your assumption of what I value.
I'm correct in my assumption about what you value. But you're correct in your assumption about what other people value.
> People will use what is most convenient, immigrant or not — most people (including me) do not take trains out of some principled stance, they do so when it’s more convenient.
Right, and virtually nobody in the Dallas, Texas area actually takes transit themselves. The proof of what they value is right there in their revealed preferences.
> It’s not because people don’t know about trains. It’s because they don’t value the things you do, and they value things you don’t, like having distance from strangers and being able to buy a lot of stuff and cart it around with them everywhere.
But isn't this pretty fungible? Like it can't be that all the people genetically predisposed to like high density neighborhoods and biking to the grocery store happen to be in the Netherlands.
Frankly, if you've never lived in a place with clean, reliable, fast trains, you probably would be disillusioned and would never go to the train life. Or if your public infrastructure is deemed "for poor people".
Back to my original point, it's a cultural problem.
Funny enough, Americans are usually happy to use public transportation when the travel in Europe or in Japan. Also most New Yorkers use the subway every day.
It's just their own public transit infrastructure they don't like, and I understand them.
“What you should in fact do is employ all the world's top male and female supermodels, pay them to walk the length of the train handing out free Chateau Petrus for the entire duration of the journey. You'll still have about 3 billion pounds left in change and people will ask for the trains to be slowed down.” ~Rory Sutherland
> I think it's not feasible in America because of the culture. If you have a good chunk of people that are hell-bent against sharing space with others for one reason or another (some are legitimate reasons!), you're already discarding a significant chunk of passengers.
That's why America never layed any railroads in the 19th century, and everyone just rode by horse instead. Oh wait, that's not what happened at all.
You are ignoring major cultural shifts - people also lived tightly packed into tenements then too. The tolerance for such space sharing without common purpose has declined.
Which means that culture could easily shift again. It's not like Europe wasn't just as car-obsessed immediately after WWII (and East Asia a few decades after that), they just realized that it was a bad idea and that cars were fundamentally incompatible with dense settlements.
> Another hot take is... if you want any of the infrastructure to be mass-used, you have to make it better than the alternatives, so people with the means would use it as well. Like your subway should be faster than the cars, so even affluent people would take it.
These days cities often achieve this by purposely hurting the alternatives like driving. And that just isn’t the solution. Mass transit has to be fast period. Not just faster than a bad alternative. And it needs to be safe, and 24x7.
> These days cities often achieve this by purposely hurting the alternatives like driving.
The point is not hurting the alternative of driving, it's to ensure that drivers don't actively hurt the more space-efficient alternatives of biking and walking on foot. The people who still have a real need for driving actually have a far better experience as a result due to the reduced traffic.
I support all kinds of transportation. I think everyone should have access to trains, cars, bikes and etc. But I also think each has its own merits. Like the car ownership over here is huge, but most commute to work on trains because things aren’t really invisibly subsidized.
At the same time, the same government, mysteriously, has no problem building a vast network of roads reaching everywhere and spanning the whole country.
If the US government neglects a section of highway until a city becomes unreachable by roads, there will be riots. The same city losing a train service? Totally expected, trains are supposed to suck.
The sorry state of American public transport is a self-fulfilling prophecy: everybody knows that public transportation sucks, and therefore nothing is done to improve it, because it's a waste of resource.
>the same government, mysteriously, has no problem building a vast network of roads reaching everywhere and spanning the whole country.
Our road-building has slowed dramatically since the 1980s. The Interstate Highway network would be much more expensive and slow to build today.
>If the US government neglects a section of highway until a city becomes unreachable by roads, there will be riots.
Consider John's Island, South Carolina. The highway that was supposed to go there has been delayed for 33 years. Access to the island/town is through two two-lane roads that get backed up to a standstill every night. There's a running "joke" about how everyone is going to die if there's a major hurricane.
> Consider John's Island, South Carolina. The highway that was supposed to go there has been delayed for 33 years.
I Googled about this issue. It looks like it is multi-factor: legal battles, lack of funding, and significant conflict over growth management vs. environmental concerns.
That’s maybe the reason today to not build more but not the reason it is bad. America ignored rail for decades in favor of highway systems and now the cost is almost always considered infeasible. We will redo our roads every 5-10 years though.
If it was invested in 50 years ago or more we would be in a different place for sure.
Can you explain how Seattle is an example? They’re opening new lines, Link is packed often, seems like a well used reliable service, but I only visit once or twice a year.
It’s been a while since I read about their system but as I recall, across the entire system something like 100 billion is the total cost. But that’s only for like 75 ish miles. So it’s very expensive. I recently saw a news article saying they’re 30 billion short per their projections and are now cutting lines out of the plan that voters expected when they supported levies, and some surrounding cities where residents have each paid hundreds or more a year for the rail to come to them, now may not get them at all. Even though they’ve been paying into it for a decade or two. Which to me is a form of theft.
The central sections of link were expensive because they're built through the center of the earth with really huge stations, some of this is to avoid impacting cars but much is just to get elevation changes. The connection over lake Washington required a lot of money and work too, as it's a floating bridge.
The less complex sections were mostly on-par with other us cities.
The per mile costs are definitely high in America, for a lot of reasons, often related to laws and policies, but that's not really the issue. At the end of 2025, nearly 20 years since California voters passed Prop 1A, we have spent under $15 Billion on California High Speed Rail. As a point of contrast, the cost of 2025's tax cut extensions is estimated to be $4 Trillion. The fact is that we don't have quality intercity passenger rail in this country because politicians aren't willing to support it and fund it as reasonable levels. Seattle light rail is an interesting example because politicians there are willing to support it and so ... we are building it, despite the relatively high per mile costs. LA Metro is interesting right because the voters passed sales taxes that funded various light rail projects. So LA is building better rail. But the political process means that every district supervisor gets their own rail project and so we have light rail to Pomona but are struggling to get a subway down Wilshire where it's obviously more needed. Anyways, all this is to say, politics is a big part of the reason why we don't have better rail in America. And blaming "grift" is a right-wing political talking point that probably doesn't help.
You raise some very good points about the expansion of rail in Los Angeles basin. However, this part: "struggling to get a subway down Wilshire where it's obviously more needed". As I understand, it is incredibly complex to build subways in that part of LA basin due to natural gas deposits that make tunnelling dangerous and expensive.
> California is about the size of Japan with the same GDP.
On the surface, this is true, but it ignores taxation structure when comparing a federal state to a sovereign nation. It would be very hard to get state-level income tax rates above 15% in the US. That cannot compete against federal/national tax rates that normally approach 40% in US and Japan. In any nation, the vast majority of large mass transit project construction costs are paid for by the central/national gov't. I would characterise your comparison is disingenuous.
> They could pass their own funded project for public transportation.
They could not reasonably 100% self-fund large mass transit projects. They need federal dollars, a lot of them, and it is very competitive to get them. As an example, look at how long it has taken to raise necessary funds to build the Silicon Valley BART extension. There is tremendous support from the public for this project, but it takes a long time to raise necessary local funds. In parallel, they need to "win" federal support for the lion's share of construction costs.
> political process means that every district supervisor gets their own rail project
This is part of what people mean by “grift.” Anyway, I’m not right wing. I just want cheap rail done competently. That’s not “not the issue.” As a voter, that is very much an issue for me.
Yeah it seems so. Anthropic has entered the enshittification phase. They got people hooked onto their SOTAs so it's now time to keep releasing marginal performance increase models at 40% higher token price. The problem is that both Anthropic and OpenAI have no other income other than AI. Can't Google just drown them out with cheaper prices over the long run? It seems like an attrition battle to me.
All of it? Islam literally calls for violence and oppression of anyone who isn’t a Muslim. It’s literally in the Quran. Islam spread to many of the regions it is now in through invasion and genocidal violence. For example, they colonized India before Europeans did. They have completely erased the previous cultures and ethnic groups of the North African countries. Etc.
This is categorically false. The Quran explicitly says "There is no compulsion in religion". It teaches Muslims to say "You have your religion, and I have my religion". It tells Muslims to fight only those who fight them and "do not exceed the limits". In fact, it says God does not forbid Muslims from dealing "kindly and fairly" with non-Muslims who have not fought or expelled them.
Can someone explain this in plain English? I intuitively can believe the hypothesis made in the introduction of this article. But I don’t get all the complex math.
reply