There is already a term for this, and ironically enough, it's often thrown around in discussions of machine learning; it's called emergence. As scales change, new properties appear, which is why we can describe a Chimpanzee as "swinging between the trees" even though at the level of quantum field theory there is no such thing as trees, Chimpanzees, or swinging.
Likewise, people shouldn't be surprised that as AI compute scales up, new forms of harm can be created, thereby introducing new moral quandaries. It's like comparing GPT-1 against today's frontier models. One is a fun albeit useless toy. The other is effecting categorical changes in the way knowledge work is done. In both cases the underlying technology is the same, but their impacts are totally different.
Just to elaborate a bit: companies register in Delaware so that legal disputes will be resolved by the Delaware Court of Chancery [1]. Although there are more business-friendly states, the Court of Chancery draws from a much larger body of business-related case law than any other, making legal challenges more predictable. The purpose of registering in Delaware is not to avoid legal scrutiny altogether.
Needless to say, this is categorically different from a company "moving" to Panama, without even maintaining a physical presence there, for the express purpose of avoiding American regulators. It's a false equivalence.
I'm surprised this article is gaining traction on HN when it's propping up such obvious conspiratorial drivel. For a counterpoint I would recommend this article [1], but I'll summarize the main points here:
- The investigation concerns somewhere between four and a dozen people spanning nearly half a decade. A dozen people dying or disappearing over the course of 4 years is hardly the statistical anomaly the articles claims it to be.
- Despite attempts to link these scientists together, there really is no common thread. One person was a biologist, not a rocket scientist; and two of the "scientists" weren't even scientists at all.
- Many of these purported "mysterious" deaths are hardly that mysterious. Two likely died of natural causes, one was murdered by a former classmate, and one disappeared while hiking. Most of the others appeared to have suffered from psychological distress.
And look, I don't want to minimize these people. These deaths and disappearances are all tragedies. The families and friends deserve closure. But dragging them into the conspiracy theory circuit is not going to do them any favors. If anything, it will likely make matters worse.
And as a scientist myself, the administration's "concern" about missing scientists feels like a slap in the face. This administration has been more hostile towards us than any other in modern history. I'll leave the article with the last word because I couldn't have worded it any better.
> Ironically, America doesn’t seem to need much help when it comes to disappearing scientists. About 1,000 employees have been laid off from NASA’s JPL in the past few years. One senior scientist who is still there told my colleague Ross Andersen last October that he’d never seen the place so empty and lifeless. In the meantime, the Trump administration has repeatedly proposed cutting NASA’s science research funding in half, a plan that would surely lead to further loss of staff at JPL, not to mention the abandonment of probes that have been sent into our solar system.
> And while the FBI looks into potential foreign involvement in professors’ deaths at MIT and Caltech, the Trump administration says that it intends to halve the budget of the National Science Foundation, which in recent years has furnished those two schools with hundreds of millions of dollars in research grants. Already, more than 40 percent of the NSF’s scientific staff have left or been fired.
> This is just a subset of the harms that have been done to the U.S. research enterprise since the start of 2025. In response, some top scientists have been getting up and walking out the door. Their absence can’t be blamed on China, Russia, or Iran. Maybe the White House should look into it.
I think a better formulation is the so-called "platinum rule", i.e. to treat people as they want to treated (with the important qualification that you ∈ people). But even then it's not without issue (what if someone's wants are harmful to them, e.g. a child refusing to eat anything but candy?), and it's still a far cry from illuminating "objective moral principles" and fairly useless as a calculus for balancing different people's competing interests.
Fair enough, I should have mentioned that the points-based system is for an accelerated application. The fact was on my mind as I was writing but I see that I forgot to mention it. My bad.
But I will point out that ten years is a major commitment. Surely if someone can hold a job for ten years the default assumption should be that they're contributing to society, not leeching off it.
It's not as if Japan (or any other country, for that matter) doesn't already have immigration restrictions. Japan uses a points-based merit system for permanent residence [1], not unlike the criteria you suggested. Just to give an example, having a PhD and speaking Japanese at an N1 level (~equivalent to B2 CEFR) is barely sufficient to qualify (unless you're older than 30, in which case it won't be).
The more interesting question to ask is: Why has Japan decided to tighten immigration requirements now? But in my opinion the answer is rather obvious, especially when you consider the current Prime Minister's nationalistic beliefs: It's much easier to blame foreigners for insufficient welfare, ailing infrastructure, etc than to actually improve welfare, infrastructure, etc.
Also, the example of "a 68 year old alcoholic high school dropout with multiple violent criminal convictions" is rather ridiculous. You're arguing a strawman. It's already impossible for such a candidate to immigrate almost anywhere barring some other exceptional circumstances.
Over the last 100 years, almost certainly not. For the most recent decade? Yes, of course I would expect these statistics to be fairly accurate.
Between 2021 and 2025 (inclusive), Wikipedia lists 68 dead in Germany versus 5882 dead in the US, despite the US only being ~4 times larger. More people have been killed by police in the US this year than in Germany in the past ten years, and it's not even April yet.
What a disingenuous comparison. The wiki article you've linked ("List of killings by law enforcement officers in Germany") sums to 552 people over the last 100 years. In contrast, the corresponding wiki article on the US ("Lists of killings by law enforcement officers in the United States" [1]) estimates more than 900 deaths per year. Indeed, the number of slayings is so great that the article does not tabulate the sum in a single table (as the German article does) but instead links to separate wiki articles with tabulated results by month.
> People underestimate how effective direct personal accountability is when it comes with harsh consequences like jail time. That's how you fix all issues in society and enforce law abiding behavior. You make the cost of the crime greater than the gains from it, then crucify some people in public to set an example for everyone else
And yet criminals still commit crimes. Obviously jail is not the ultimate deterrent you think it is. Nobody commits crimes with the expectation that they'll get caught, and if you only "crucify some people", then most criminals are going to (rightfully) assume that they'll be one of the lucky ones.
Essentially all of your concerns concerns can be mitigated by building somewhere else.
Worried about natural disasters? Build some place less prone to natural disasters.
Worried about the strain on local communities? Build some place more remote.
Worried about energy availability? Build near a nuclear power plant or hydroelectric power station.
Worried about hostile governments? Don't build data centers within the territories of hostile governments. (If you consider every country a hostile government, that is a you-problem.)
For the cost of building a data center in space, you could instead build a second (or third, or fourth, ...) data center somewhere else.
> A friend of mine has a terrible Instagram addiction, yet has developed for himself a certain degree of cinephilia lately -- we've watched long movies together in theaters and not once has he been on his phone during the screenings. When one has faith that sustained attention might hold more value than that gained by interruption, they tend to prioritize the former.
I'm not convinced that you've fingered the reason. Pulling out your phone at the theater is considered anti-social behavior, comparable to conversing with your seatmate, and that sort of normative pressure can overcome a compulsion. It's like claiming that someone couldn't possibly be an alcoholic because they don't drink on the job.
A better test would be: What does your friend do when you watch a movie at one of your homes, where there's a lesser expectation to tuck away one's phone? Does he still watch the movie attentively, or does he check his phone every so often?
It's satire. It's supposed to be absurd. Why else do students still read A Modest Proposal nearly three hundred years after its publication?
Regardless, LLMs are already being abused to mass produce spam, and some of that spam has almost certainly been employed to separate the elderly from their savings, so there's nothing particularly implausible about the satirical product, either.
No, the cause is structural. Even if one could identify the sources of rot (money in politics, an outdated electoral college, the collapse of our information environment, whatever), Congress would deadlock, the Courts would block any meaningful reform, and the President would be left trimming the blight while the rot festered underneath.
> Whatever power you put into the hands of the government is guaranteed to fall into your enemy's hands some day.
Only if there's a functioning system of checks and balances. Unfortunately, there is not. This Court is willing to use motivated reasoning to achieve its preferred outcomes; to slow-walking favorable rulings for Democrats while expediting favorable rulings for Republicans (often without explanation via the "shadow docket"); and to throw out decades of precedent in the process by ignoring stare decisis, a bedrock legal principle which ensures stability of the judicial process.
Just to give an example, consider the ban on universal (national) injunctions. One might be surprised to learn that it was the Biden administration that initially petitioned the Court for the ban. However, the Court found such a ban unnecessary then (i.e., when lower Courts were blocking the Biden administration's agenda), but conveniently found it necessary during the second Trump administration (when lower courts started blocking the Trump administration's agenda). And just as another kick in the balls, they used the birthright citizen case as a vehicle to bring the matter to Court, strengthening the President without even deigning to address the Trump administration's obviously illegal executive order.
The result of this mess is that, if the Trump administration is eventually voted out, it is highly unlikely that an incoming Democratic administration would be able to capitalize on the expansion of executive powers that this Court has given to this President. We see a similar situation in Poland. After ~a decade in power the Law & Justice party was voted out, but the new coalition government has not inherited the same ability to government, with its agenda constantly curtailed by Law & Justice appointees embedded throughout the government (including the highest court).
Even this is too charitable. A short timeline of January 2025 would be something like this:
- Jan 16: The Supreme Court issues its opinion, upholding the legality of the TikTok ban. The Biden administration declines to enforce it, preferring to let the incoming Trump administration handle the matter.
- Jan 18: TikTok voluntarily turns off its services. Google and Apple remove the app from their respective app stores. Trump declares on social media that he will sign an executive order "to extend the period of time before the law’s prohibitions take effect".
- Jan 19: TikTok restores it service after being assured by the incoming Trump administration that TikTok would not face penalties.
- Jan 20: The Trump administration signs the aforementioned executive order.
However, Trump's executive order was untimely (the law already should have gone into effect), and at any rate it's dubious that the executive order would've been legal regardless. The TikTok ban (PAFACA) had a specific provision for when an extension could be granted. From Wikipedia:
> The president may grant a one-time extension of the divestiture deadline by as long as 90 days if a path to a qualified divestiture has been identified, "significant" progress has been made to executing the divestiture, and legally binding agreements for facilitating the divestiture are in place.
Notably, none of these requirements had been met. There were no identified buyers; there were no binding agreements. The Trump administration's refusal to enforce the TikTok ban might have been the first lawless act of the second administration, and it happened only within hours of Trump being sworn in.
CAIR is an American organization established to protect the rights of Muslims in America. The people who work at this organization are, presumably, Americans by and large. So why would an American civil rights group divert its limited resources to something squarely outside its scope, especially when such advocacy would require entirely different, non-overlapping expertise in Moroccan/Turkish/whatever law?
Likewise, people shouldn't be surprised that as AI compute scales up, new forms of harm can be created, thereby introducing new moral quandaries. It's like comparing GPT-1 against today's frontier models. One is a fun albeit useless toy. The other is effecting categorical changes in the way knowledge work is done. In both cases the underlying technology is the same, but their impacts are totally different.