What AI isn't

[Churchill]

The LLMs do. Your fifty years lacked the hardware required to run such a LLM so nobody even tried.

The hardware is the same just faster, all hardware amounts to an FSM and FSM's embody rules, not choices.

But now times have changed.

Yes, times do change.

AI is even used in war.

I know.

It decides what target to attack.

No, the target is computed based on rules, not chosen.

 

[Churchill]

No. Making decisions is the whole concept of these LLMs. It is evolution of software.

Computing outcomes by following rules (FSM) does not entail "choice" which implies free will, an innate "self" but there is no such mechanism in any man made systems no matter the complexity.

 

[Bleipriester]

I'm not sure they make decisions. That would indicate that they have the power to choose anything else. It isn't a choice as much as a result of an algorithm. If the statistics indicate that based on samples, there is a high probability that the word "jump" would be the next in a sentence, the computer cannot choose NOT to use that word based on a feeling.

That is all correct. Still, the decisions are made by the AI. It is programmed to do so. There are enough examples you can read about. The developers themselves test their LLMs to see what they are doing. Example:

AI system resorts to blackmail if told it will be removed

In a fictional scenario, the model was willing to expose that the engineer seeking to replace it was having an affair.

www.bbc.com

 

[Bleipriester]

The hardware is the same just faster, all hardware amounts to an FSM and FSM's embody rules, not choices.

Yes, times do change.

I know.

No, the target is computed based on rules, not chosen.

Of course based on rules. But he AI decides. You need to understand that.

 

[rosends]

That is all correct. Still, the decisions are made by the AI. It is programmed to do so. There are enough examples you can read about. The developers themselves test their LLMs to see what they are doing. Example:

AI system resorts to blackmail if told it will be removed

In a fictional scenario, the model was willing to expose that the engineer seeking to replace it was having an affair.

www.bbc.com

but if it is based on rules, then what room is there for decision making? It seems more deterministic to me.

 

[Churchill]

Of course based on rules. But he AI decides.
You need to understand that.

Fine, you believe whatever you like, don't let the facts get in your way.

 

[Bleipriester]

but if it is based on rules, then what room is there for decision making? It seems more deterministic to me.

If you don´t apply rules on the AI what point whatsoever is there for it to exist? Again, the LLM is there to emulate human behavior. It can boast, lie, whine, ect. Only then it is smart enough to make decisions.

 

[Bleipriester]

Fine, you believe whatever you like, don't let the facts get in your way.

Returned

 

[Churchill]

Bleipriester

Since you value and trust AI, do you trust this?

 

[rosends]

If you don´t apply rules on the AI what point whatsoever is there for it to exist? Again, the LLM is there to emulate human behavior. It can boast, lie, whine, ect. Only then it is smart enough to make decisions.

but if you apply rules, then what room is there to choose?

 

[Bleipriester]

Bleipriester

Since you value and trust AI, do you trust this?


View attachment 1244823

That is basically what I told you. But to correct your assumption: No, I don´t "trust" the AI. I trust it as much as I trust human beings.
But it is very interesting and simply genius.

 

[Churchill]

That is basically what I told you.

No, it's basically the opposite of what you claimed, read it: Short answer: no

But to correct your assumption: No, I don´t "trust" the AI. I trust it as much as I trust human beings.
But it is very interesting and simply genius.

 

[Bleipriester]

No, it's basically the opposite of what you claimed, read it: Short answer: no

Let me reply with another AI´s assessment.

So as it simulates decision-making, the decisions are being made nevertheless. It is what I told you.

 

[Churchill]

Let me reply with another AI´s assessment.

So as it simulates decision-making,

No it does not, you interpret what it does as "simulating decision making" but those with the insights and knowledge of computability mathematics, logic etc do not interpret it that way.

the decisions are being made nevertheless. It is what I told you.

A deterministic system cannot simulate non-determinism.

 

[Bleipriester]

No it does not, you interpret what it does as "simulating decision making" but those with the insights and knowledge of computability mathematics, logic etc do not interpret it that way.



A deterministic system cannot simulate a non-deterministic system.

You will understand that if such an AI is given power, for example in a drone, its simulated decision to attack this point and not that point, has a real impact. It is a real decision, whether it was made by a human or by a set of calculations.

 

[Slade3200]

No thanks. I have no interest nor need for artificial intelligence when God has already imbued me with REAL intelligence.

Let’s test that intelligence that God gave you… have you ever traveled anywhere by either train, plane, or automobile? Yes/No

 

[Slade3200]

Only an idiot would think that putting a supercomputer artificial brain programmed by a profit-driven private company with an IQ of 1000 in autocratic charge of running people's lives would be a good thing, not unless it comes with a strong sense of morality and ethics as well, and you know that isn't happening.

What exactly do you mean by “in charge of running people’s lives”? in what way is AI running lives?

 

[Slade3200]

And if AI ever becomes real AI, then it is no longer artificial.

Creating real AI is like trying to create a supercharged human mind on steroids. That is akin to a chicken coop putting a pack of really big foxes in charge of guarding their coop!

You are misunderstanding what the term artificial means. The intelligence is real, it is just made artificially versus naturally.

 

[Churchill]

You will understand that if such an AI is given power, for example in a drone, its simulated decision to attack this point and not that point, has a real impact. It is a real decision, whether it was made by a human or by a set of calculations.

All decisions are made by people, the people who designed the software, chose the rules, selected the data, the decision to use it during warfare, those are the actual and only decisions, all decisions are made by humans.

 

[rosends]

Let me reply with another AI´s assessment.

So as it simulates decision-making, the decisions are being made nevertheless. It is what I told you.

an actor might simulate performing an operation, but he isn't actually performing an operation.