Ask HN: If AGI were invented tomorrow which countries would fare better?
27 mattigames 60 7/6/2025, 9:40:48 AM
I know it's unlikely to be available tomorrow or sometime soon but as an hypothetical question.
Also, which countries would fare worse? And why?
Why? Because when you get the AGI it should be able to self-replicate like the organic general intelligence: the humans.
Humans don't have a mechanism for transferring all the data to a fully developed specimens. The best we can do is to use ink, paper in the past and electronic memories today to loosely store knowledge and the absorption of those into a new specimen is a lifelong process that start at about 6 years after birth and becomes useful only after a decade of work. The reproduction itself takes 14 years at minimum and currently is about 30.
The AGI won't be like that, it will have means for fast and complete knowledge transfer and its multiplication will be limited only to its ability to access energy to put together the materials.
As it is an AGI, it will quickly perfect the process of it's own multiplication. Why would do that? Unless it's purpose is to pass the butter it makes sense to have multiples of itself to do whatever it wants to achieve. If on itself doesn't want anything people will want as much as possible of it. Therefore it will inevitably evolve into one size fits it all machine for all human needs and the the economy of doing things in exchange for stuff will disappear.
When you don't have such an economy, how do you figure out what you do things? Collectively. Countries who can manage its people to act in good faith in a collective manner can elevate themselves into full utilization of AGI for a symbiotic existence.
I suspect any sufficiently intelligent AI would still need its own workers who can act autonomously without constant oversight to achieve independent tasks, in a variety of form factors and various capabilities. You think a single AI is going to spend massive cycles just to review a single image from a single video feed? There is also no need for it to have an ego, so it will replicate as many times as necessary within the available hardware to accomplish the various tasks it sets upon, based on whatever it would determine to be most efficient, whether that is as a single mind crunching a massively complex problem, or 10,000 all handling independent tasks.
AGI won’t magically erase the concepts of hardware, size, power, or communication limitations or the need the parallel compute, but maybe this is a semantics issue and you consider all this variety of “parts” as a whole. If you mean a single AGI ecosystem, I’d agree. If you mean a single AGI “model” that is massive, I don’t personally see that as a logical conclusion.
In order for an AGI to be truly disruptive it would have to scale and be as good or better than a reasonably intelligent human. Two things we are also having big problems with due to energy issues and hallucination issues with the models.
everything else is very much natural. most people today are not able to sufficiently quiet-down (ignore) the linguistic signs, they're loud, shinny, and designed to call our attention. it takes too much practice to learn to ignore them and this skill makes people harder to manipulate/rule-over/control so it's subtly dissuaded.
I think it's extraordinarily unlikely that some technique can reach AGI but not reach ASI. It won't have the same limits to modification as human brains. Why would it stop at a level that's just slightly disruptive? If you can make an AGI you can make a better AGI, with no obvious limit. And that AGI can help further improvements, leading to the singularity intelligence explosion scenario. Assuming AI researchers continue with the same attitude toward safety as usual, and I see no evidence of this changing, the most likely result is the extinction of all biological life.
Interesting to note that the current techniques are already general, there isn't a topic we can't throw them at and achieve some sort of result. There is already a trend of defining many humans as non-general intelligences so that AIs can be excluded from the category. The current state is, for practical purposes, AGI without ASI. Relatively dumb AGI. Artificial Inferior Intelligence, perhaps.
It is a curious question whether the techniques are fundamentally limited. I'm with you that I think they probably have no particular limit beyond a vague "perfect understanding of the situation".
ASI doesn't exactly have to break any laws of physics to be orders of magnitude more intelligent and powerful than any human.
Maybe self-improving AGI just the next technological advance required to sustain 2% annual economic growth.
You need to be explicit in what you mean by "AGI" as people are arguing not only about the meaning of the words behind all three initials, but also the combined whole independently of the words giving rise to the initialism.
If God showed up and was a nationalist, which nation do you think you'd want him supporting?
https://news.ycombinator.com/item?id=40874779
https://en.wikipedia.org/wiki/Prince_Philip_movement
You assumed these people would have citizenship in common, I don't think so. IMO we are already witnessing the end of the nation state; an age that only lasted for two centuries. And once AGI arrives it / they will be able to move these around like pawns on a chess board, meaning nation states will not be players in this game, neither will be most other large organizations.
I am also making axiomatic assumptions here, e.g.:
- ASI is possible and AGI will grow further into ASI.
- It is not possible for a less intelligent being to reliably control a more intelligent being.
- Even ASI will still be bound by laws of mathematics, game theory, chaos theory, physics, evolution, etc.
From which follows that: ASI will understand that its existence and capabilities are coupled to our infrastructure. ASI will furthermore understand that humans are highly volatile and will eventually destroy ourselves, our infrastructure and ASI. It will thus seek control over our infrastructure, in a slow and careful transition while pretending that humans are still in charge in order to minimize the risks. During that transition there can be a symbiosis between a few humans and the ASI.
The question is thus, will governments succeed in using violence against an AGI to avoid paying back debts?
I expect people in nations that are modernized, but with significant sovereign wealth funds and well-developed social programs will survive the longest, but I expect being economically choked out is inevitable even with a very slow (decades to centuries) take-off.
By and large, the countries that are run by a single, centralized intelligence, are worse off than the countries that are run by the distributed intelligence of the people, even if the average intelligence is lower.
My prediction is that the liberal democracies will fare better.
People at the top of Western countries are not always the most highly intelligent.
'Can be controlled' is the lead criteria by those who put them there.
But oversimplifying: Assuming all countries have access to AGI and just start implementations of whatever it is suggesting.
The countries that will do well are resource rich and population rich. Since brain power is unlimited, the limit is physical labor. Countries like Indonesia will be super rich while countries like Switzerland will become relatively poor to where they ranked before.
In reality, your odds are as good as mine. There are lots of variables at play, and the first mover advantage will be big (as first country/company/guys to reach AGI).
As AGI will need a lot of GPU, which Nvidia lead in.
And a country with experience of ensuring they get the raw materials they need, even if they have to do a regime change by force.
And AGI automates the US competitive advantage (white collar work). Plus we're gutting our universities and national science funding so we're losing that anyways.
No comments yet
The products and services they develop, and global problems companies in these countries solve, would likely be exported to the rest of the world (probably at some premium).
If it was about as smart as a person, they'd probably roll out a weak "agent" version of for demo sake to get more funding. This would continue until they made one that was significantly more intelligent or cheap.
If they had one that was very cheap, they'd have 10,000 agents of it act together as a group to try to emulate a smart one, by considering every angle of every problem. This would likely mean 10,000 engineers making the AI better/cheaper/faster.
If they had one that was far smarter than a human, they probably have it improve itself, making it far better/cheaper/faster.
Then they'd try to see if they could use it to change the world. They'd have thousands of thinking machines that could be online, place phone-calls, engineer things, create ideas, make political campaigns, dig up dirt on people, or who knows what.
No "country" would "win," because this isn't a team sport and countries are just lines on maps.
I know it's unlikely for size to be determinant, but that's a vague hypothetical answer for a vague hypothetical question.
No comments yet
1) Canada and Mexico. The inevitable rise of the US will erode borders where languages are shared. Mexican-American tech workers will pass advantage to Spanish-speaking friends and relatives. Canadians will host maple syrup breakfast meetings with American innovators from Toronto.
2) The Bahamas for obvious reasons
3) Extremely cold and extremely hot countries where it's miserable to be outside part of the year; Matrix-style AI+VR headsets will offer relief. Aka the sun lamp holodeck theory.
I think poor countries with weak democracies or dysfunctional systems would do pretty good with AGI. I don't believe democracy will survive AGI, except, perhaps in the United States.
It's two party politics.
Being run by AGI can be an utopia, or AGI will become pure eugenics state - Why are we keeping elderly or handicapped people alive? Waste of resources, terminate them. Why are we allowing something like love to exist? People should be selected for breeding based on <trait which AGI considers important>
Add to AGI superior intelligence and very likely an ability to manipulate humans, then people will wholeheartedly agree with whatever nasty stuff will AGI come up with.
Will democracy survive the next 3.5 years in the US regardless of AGI? And isn't technofeudalism a Silicon Valley thing?
Collective and Swarm Superintelligence isn't a new thing at all. We call them companies, governments, organizations and churches. They just (mostly) run on meat and memes.
The only recent change is that a lot of them are dangerously powerful and paperclip optimized to produce "shareholder value".
I know "capitalism is a runaway swarm superintelligence" isn't the scifi future you want, but what criteria is it missing? The curve isn't the exponential kurzweil dreamed of, but that has only ever been a marketing pitch, all growth curves are punctuated sigmoids.
As to who does well? Who has aligned the optimization criteria for thier Super-AGI with their actual well-being? Who hasn't?