[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/his/ - History & Humanities


Thread archived.
You cannot reply anymore.




Would a super computer as a president with a cabinet of super computers lead any nation to prosperity? Why hasn’t this been tried? Could this hypothetical nation out compete nations in geopolitics that don’t have a super computer in charge?
>>
To an extant. I believe this technocracy should have the computer as the God of sorts, presenting projections, with a human president and cabinet to decide the best course for prosperity.
>>
That would only work if you think politics is a science that has objective laws, e.g one correct answer to a problem, like a math equation.

Protip: It does not.
>>
*smashes the president computer with a club and declares myself president*

see?
>>
people wouldn't like what it says
>>
>>13755831
OpenAI just wrote a paper on cracking the game of stratego: https://arxiv.org/abs/2206.15378

It performs really well, but not perfectly. This is the very cutting edge of AI in terms of games - along with the projects of DeepMind. Thousands and thousands of hours of training on supercomputers filling entire buildings to produce an AI that barely outcompetes humans in one game.

So, in many strategic games, it seems AI is slowly but surely surpassing humans in every last one we can still hold over the best of AIs. Maybe someday that will extend to ruling a nation?

Well, ruling a nation is much more complicated than playing a strategy game, it turns out. To train an AI for a game like chess, go, or stratego, you can simulate the game perfectly - you can have the AI play against itself or against other people, and measure how good a position is based on winrates and other measurements. These types of metrics are practically nonexistent in real life.

How do you measure a good ruler? GDP? Happiness? A good policy might impact the GDP 40 years later - an AI would have to try millions of things to find an action like this, and has very few ways to tell which actions led to what. And it will learn patterns - only to find that these patterns only exist for a certain period of history due to a technology or weird aspect of human culture.

The fact is, AI as it exists now is highly dependent on ML, which just reads in a ton of data (inputs and outputs) and runs linear regression to find weights that achieve the best accuracy given some test data. This can be used to achieve amazing results in many tasks - but general intelligence? We;re far from that.

The truth is, general intelligence requires some advancement we just don't know. we can achieve vision, text generation, and similar tasks - but not yet an AI that can synthesize these things and produce results that aren't simply taken and rearranged from training data.
>>
>>13755831
>Would a super computer as a president with a cabinet of super computers lead any nation to prosperity?
It's communist cope for pretending resources allocation doesn't need creativity.
>>
There are times when policy needs to focus more on long-term gains than short-term gains. Given AI control of the government, would you choose to make it more in favor of the national interest or the citizen interest? These two interests don't intersect in many ways.
>>
>>13755831
>Would a super computer as a president with a cabinet of super computers lead any nation to prosperity?
>Why hasn’t this been tried?
>real automated luxury communism has never been tried.
Paul Cockshott pls go.
>>
>>13757402
>There are times when policy needs to focus more on long-term gains than short-term gains.

lolno
>>
>>13755831
How much does your tard wrangler make?
>>
Wouldn’t the IT technicians become the new eunuch rulers like in China?

>I fixed an.... error... virus... not we get 500% more pay. It’s what the computer wanted
>>
>>13755831
What happens if a wasteland vagabond strolls in and calls it an abortion of science?
>>
>>13755831
Computers have no will of their own. An AI would make a great civil servant but not a good politician.
>>
>I am your President, John Henry Eden
>>
>>13755831
Who decides the priorities of the AI?
>>
>>13757888
The value of friendship, i'm sure this machine without any understanding of what being human means (besides efficiency) would never exploit...
>>
>>13755831
https://m.youtube.com/watch?v=RJLA2_Ho7X0
>>
>>13755831
>>13757997
All modern corporations are governed by computer systems their purpose is to maximize profit but we could maximize any metric we wanted.
>>
>>13755864
/thread. spbp



Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.