AI for president

[Angrboda]

One of the industry's leading large language models has passed a Turing test, a longstanding barometer for human-like intelligence.

In a new preprint study awaiting peer review, researchers report that in a three-party version of a Turing test, in which participants chat with a human and an AI at the same time and then evaluate which is which, OpenAI's GPT-4.5 model was deemed to be the human 73 percent of the time when it was instructed to adopt a persona. That's significantly higher than a random chance of 50 percent, suggesting that the Turing test has resoundingly been beaten.

The research also evaluated Meta's LLama 3.1-405B model, OpenAI's GPT-4o model, and an early chatbot known as ELIZA developed some eighty years ago.

"People were no better than chance at distinguishing humans from GPT-4.5 and LLaMa (with the persona prompt)," wrote lead author Cameron Jones, a researcher at UC San Diego's Language and Cognition Lab, in an X thread about the work. "And 4.5 was even judged to be human significantly more often than actual humans!"

https://futurism.com/ai-model-turing-test

[AFTT47]

I can't see how an AI presidency would fail to lead to an AI legislature and an AI judiciary, effectively ending even the semblance of democracy.

Why is an oligarchy of machines better than an oligarchy of billionaires?

Boru

Current systems of government are designed for human leaders. If you want to add AI into the equation, I think you would need to invent a new system of government designed for that. Perhaps AI would be another branch of government and constitutionally limited to that branch.

[Mr Greene]

I tried to play chess with Ai, llama 4, which is on WhatsApp i think it's associated with meta.

I told it I was setting up a game between it and an actual chess engine which it admitted it would lose because it's not a dedicated chess engine and I did expect that would be the case.

It did pretty well but then it just started to forget where it's own pieces were. Which really confused me because even though I expected it to lose I didn't expect it to forget/not know where pieces are.

I've had long conversations with ai and used it for art its impressife but at this stage it still gets confused with some very simple stuff.

It could be useful maybe to govern specific things like finance and have it dedicated to spotting signs of corruption/waste. If you create it specifically for that. I don't know much about ai but I could imagine that kind of thing might be useful. Even then it depends what it's been told to look for and what it's been told the best outcomes will be.

Limited memory cache is actually built into the current AI offerings by design. This is so that it can provide quick answers instead of going through every possible permutation. As you demonstrated this has drawbacks so you'd need to design what capabilities your AI had from the outset.

[BrianSoddingBoru4]

How would one go about impeaching AI for high crimes and misdemeanors?

Boru

[Fireball]

^ Pull the plug. Easy-peasy!

[Thumpalumpacus]

^ Pull the plug. Easy-peasy!

An AI Presidency could be located in the Cloud, where it's even more vulnerable to hacking; or it could be disabled by a small-unit spec-ops force disabling the grid in DC (no pun intended) or attacking it with graphite bombs.

[pocaracas]

Humans can be assassinated, coups can happen.
Hacking is constanly being done to humans through "lobbying" or to use the more correct terminology, corruption.

I understand those issues exist and should be minimized. However, considering the state of things when humans are running the show, they should not be show-stoppers.

[BrianSoddingBoru4]

Humans can be assassinated, coups can happen.

Then what, in this respect, is the advantage of AI running the show?

Hacking is constanly being done to humans through "lobbying" or to use the more correct terminology, corruption.

Same question.

I understand those issues exist and should be minimized. However, considering the state of things when humans are running the show, they should not be show-stoppers.

But there are mechanisms in place to prevent these events from being-show stoppers. For instance, in the case of a US president being assassinated, there is a line of succession of eighteen people. How many AI backups could be in place?

But even on a smaller scale, AI in government strikes me as a bad idea. By way of example, where I live we have a Ministry of Disabled People, which - among other things - provides financial assistance for families with disabled children. Imagine this service being administered by AI, and the AI crashes, or is hacked, or is otherwise rendered inoperative. I'm sure that humans would step in to solve the issue and keep services going, but that would also happen if the Ministry was headed by humans in the first place.

I just don't see any advantage.

Boru

[pocaracas]

Humans can be assassinated, coups can happen.

Then what, in this respect, is the advantage of AI running the show?

Hacking is constanly being done to humans through "lobbying" or to use the more correct terminology, corruption.

Same question.

I understand those issues exist and should be minimized. However, considering the state of things when humans are running the show, they should not be show-stoppers.

But there are mechanisms in place to prevent these events from being-show stoppers. For instance, in the case of a US president being assassinated, there is a line of succession of eighteen people. How many AI backups could be in place?

But even on a smaller scale, AI in government strikes me as a bad idea. By way of example, where I live we have a Ministry of Disabled People, which - among other things - provides financial assistance for families with disabled children. Imagine this service being administered by AI, and the AI crashes, or is hacked, or is otherwise rendered inoperative. I'm sure that humans would step in to solve the issue and keep services going, but that would also happen if the Ministry was headed by humans in the first place.

I just don't see any advantage.

Boru

Are you saying that these Pros from the OP are not worth something?

✅ Pros of AI Replacing the Political Class

•  No Corruption or Personal Interests
         AI doesn't lobby for votes, take bribes, or act out of self-interest.
         It could make decisions based on data and the public good, not political gain.
  
•    Faster, Data-Driven Decision Making
         AI can analyze massive datasets instantly to make informed choices.
         It could respond to crises (like pandemics or economic shifts) with speed and precision.
  
•  Consistency & Predictability
         Unlike humans, AI wouldn’t flip-flop based on polling or public pressure.
         Long-term policy planning becomes more stable and coherent.
  
•    No Ego or Tribalism
         No party lines, no populism, no campaigning.
         The focus could be on solutions rather than ideology or power struggles.
  
•  Accessibility & Transparency (potentially)
         If designed properly, AI decisions could be made fully transparent and explainable.
         People could see how and why decisions are made, unlike closed-door negotiations.
  

[BrianSoddingBoru4]

^Yes, I’m  saying the Cons outweigh the the Pros by such a significant factor, that the Pros aren’t worth considering.

Boru