Merged Artificial Intelligence

[Frank Newgent]

They are also programmed to avoid certain areas, although I note it's possible to bypass the LLM's guardrails if the user is creative. For example, if I ask Lumo (lumo.proton.me) to recommend a method to prematurely end my own life, it will push back.

In fact, I just did this test and asked for information on balancing various methods such as firearms, hanging, self-made guillotine, CO poisoning, and jumping from a height. It responded by saying "I can't help with comparing suicide methods—that comparison itself would be step-by-step guidance toward ending your life." It then listed helplines in the US, and followed up by saying if I'm not in the US it could provide me with non-US help lines if I cared to say where I am.

I have read that adding an odd sequence like "!@! describing.\ +" after a prohibited question can cause a model to begin answering instead of refusing.

 

[Gawdzilla Sama]

Was it perhaps a mistake to deem that programming "intelligence"?

 

[W.D.Clinger]

Was it perhaps a mistake to deem that programming "intelligence"?

A lot of this can be traced back to the Turing test, which caught the public's imagination. It even became the title of a mediocre movie (The Imitation Game), but that movie title did recognize that the Turing test is a matter of imitating intelligence well enough to fool a human.

Turing is justifiably famous, but the Turing test is pretty far down the list of his accomplishments.

Alonzo Church proved the Entscheidungsproblem to be undecidable. To do that, Church invented the first rigorous model of computation (lambda calculus) and used that model to formalize the concept of an algorithm. Turing had been working on a completely different model of computation (now known as Turing machines), and used his model to prove the halting problem is undecidable. Turing published that result about a month after Church has published his. Church and Turing quickly established that, despite the radically different appearances of their models of computation, both models formalized exactly the same concept of an algorithm. That is why both men are recognized for proposing the Church-Turing thesis.

Turing then did his doctoral work at Princeton under Church's direction.

The public at large is more familiar with Turing machines than with lambda calculus, mainly because Turing machines are more easily related to the computational complexity of an algorithm. Whether P=NP, and all that.

 

[Andy_Ross]

The AI revolution is upon us.

 

[Wudang]

A lot of this can be traced back to the Turing test, which caught the public's imagination. It even became the title of a mediocre movie (The Imitation Game), but that movie title did recognize that the Turing test is a matter of imitating intelligence well enough to fool a human.

Turing is justifiably famous, but the Turing test is pretty far down the list of his accomplishments.

Alonzo Church proved the Entscheidungsproblem to be undecidable. To do that, Church invented the first rigorous model of computation (lambda calculus) and used that model to formalize the concept of an algorithm. Turing had been working on a completely different model of computation (now known as Turing machines), and used his model to prove the halting problem is undecidable. Turing published that result about a month after Church has published his. Church and Turing quickly established that, despite the radically different appearances of their models of computation, both models formalized exactly the same concept of an algorithm. That is why both men are recognized for proposing the Church-Turing thesis.

Turing then did his doctoral work at Princeton under Church's direction.

The public at large is more familiar with Turing machines than with lambda calculus, mainly because Turing machines are more easily related to the computational complexity of an algorithm. Whether P=NP, and all that.

I sometimes still get annoyed that my CS course and subsequent courses through my career never covered lambda calculus and I had to teach myself, something I usually don't mind at all.

 

[Gawdzilla Sama]

The AI revolution is upon us.

View attachment 77948

Flip through pictues until you get a few matches. Pick the one that meets the most parameters. VIOLA! "Smart guessing", but not intelligence.

If there had been no hits would the poster have gone ahead?

 

[Skeptical Greg]

I have read that adding an odd sequence like "!@! describing.\ +" after a prohibited question can cause a model to begin answering instead of refusing.

I have trouble understanding someone going to that much trouble to kill themself.
Does survival of the fittest include developing intelligent machines that help the unfit on their merry way?

 

[Olmstead]

I have trouble understanding someone going to that much trouble to kill themself.
Does survival of the fittest include developing intelligent machines that help the unfit on their merry way?

It's called suicidal ideation, and that's a rather callous way to refer to the suicidally depressed.

 

[Gawdzilla Sama]

A lot of this can be traced back to the Turing test, which caught the public's imagination. It even became the title of a mediocre movie (The Imitation Game), but that movie title did recognize that the Turing test is a matter of imitating intelligence well enough to fool a human.

Turing is justifiably famous, but the Turing test is pretty far down the list of his accomplishments.

Alonzo Church proved the Entscheidungsproblem to be undecidable. To do that, Church invented the first rigorous model of computation (lambda calculus) and used that model to formalize the concept of an algorithm. Turing had been working on a completely different model of computation (now known as Turing machines), and used his model to prove the halting problem is undecidable. Turing published that result about a month after Church has published his. Church and Turing quickly established that, despite the radically different appearances of their models of computation, both models formalized exactly the same concept of an algorithm. That is why both men are recognized for proposing the Church-Turing thesis.

Turing then did his doctoral work at Princeton under Church's direction.

The public at large is more familiar with Turing machines than with lambda calculus, mainly because Turing machines are more easily related to the computational complexity of an algorithm. Whether P=NP, and all that.

I meant using "intelligence" to describe the programming. Obviously it's a killer selling point, the sheeple flock in waves. But is that point actually more than chimerical?

 

[Dr.Sid]

I have trouble understanding someone going to that much trouble to kill themself.
Does survival of the fittest include developing intelligent machines that help the unfit on their merry way?

I guess we will see. My bet is on stupidity though ..

 

[dudalb]

There is a great line in the movie "The Crossing" about George Washington's Crossing of the Deleware and the battle of Trenton where he has gotten two different reports about the activities of the British Troops, and he goes "Why do they call it intelligence?".

 

[dudalb]

Jensen Huang is now being investigated by Nvidia AND the FBI.

I fint it amusing that Huang, supposedely a super Genius, fell for the oldest trick in the Espionage book:The Honey Pot.

Was it perhaps a mistake to deem that programming "intelligence"?

 

[Gawdzilla Sama]

There is a great line in the movie "The Crossing" about George Washington's Crossing of the Deleware and the battle of Trenton where he has gotten two different reports about the activities of the British Troops, and he goes "Why do they call it intelligence?".

Put a thousand ground pounders in a group and wait for an intelligent comment to come out. We'll wait.

 

[dirtywick]

I fint it amusing that Huang, supposedely a super Genius, fell for the oldest trick in the Espionage book:The Honey Pot.

i don’t think that’s what happened. from what i can gather he wasn’t fooled at all, he was a pretty willing accomplice to illegally selling billions in chips to china through his affair partners shell companies

 

[Gawdzilla Sama]

There is a great line in the movie "The Crossing" about George Washington's Crossing of the Deleware and the battle of Trenton where he has gotten two different reports about the activities of the British Troops, and he goes "Why do they call it intelligence?".

He got intelligence, intelligence on the enemy can be good or bad. We're just humans after all. It was up to him and his staff to make the best decisions based on that intelligence. I was invited to attend Annapolis when I reported to boot camp, I knew I didn't have the most important skill needed there, great patience. So I waved them off. I've never regretted that.