"Godfather of AI" Geoffrey Hinton Departs Google, Voices Concern Over Dangers of AI

[Aquinus]

My point was, that guns were not created to hunt - they were created to kill.

Pardon my ignorance, but what do you think hunting is?

Edit: if the item quoted had been a bow and arrrow - yes, it'd been fine. Bows were created to hunt. Guns weren't.

Bows and arrows were designed to kill. Both animals and humans. So, I'm not sure where you're going with this. I'm seeing conflicting points in your argument. Again, it's a potentially lethal tool, both of them, with the same intent. A firearm is merely an evolution of more archaic tools. What you're describing is rhetorical nonsense.

 

[Guwapo77]

That's not sound reasoning. Guns were the portable evolution of cannons (an evolution of Chinese fire lances). They were never created for the purpose of hunting, rather, they were invented to win battles. AI is not such a thing; it is simply an entity in its own right, with an exceptionally diverse field of application. Most of which require to be heavily controlled to avoid the consequences.

IMO, AI created in our world will be no different than the people and nations that create it. Greedy, ideological, and/or altruistic. There will be bad AI and good AI. Problem is, a good AI probably wouldn't have the ethics to beat a bad AI. Story of life.

Not the greatest of analogies I admit, but you get the point. AI will be used for bad and that's a fact no matter how good or how bad the analogy will be.

 

[bug]

I see the bigger picture. I was inferring the 'guns' metaphor wasn't appropriate given that they were created for a purpose intended to harm others. AI isn't the same. It's end result will be determined by those who guide it.

Even that is inaccurate. Can you definitely say guns were invented to harm others and to protect one's own?

But again "guns" was just a word in an analogy. Let it go, take a step back and think about the analogy instead.

 

[AusWolf]

I think all the AI doomsayers are overreacting and exaggerating too much. AI is not dangerous, it just uses the free public available info online.
What is dangerous is the tools such as deepfake, voiceovers, etc, that can be used in malicious purposes, so that can be regulated.
For me, the tools such as BARD, ChatGPT and BingChat are tremendous helpful in my work, helping me complete tasks in 1H, that usually were taking 1 day or more to achieve.
It all depends how you use those tools.

And that's why it's dangerous. If something can be used for something bad, people will find a way. Besides, free public info is not necessarily (and a lot of times isn't) correct. Not long, and we'll live in a world where literally everything is fake, and it's impossible to tell what the truth about any subject is.

 

[Aquinus]

And that's why it's dangerous. If something can be used for something bad, people will find a way. Besides, free public info is not necessarily (and a lot of times isn't) correct. Not long, and we'll live in a world where literally everything is fake, and it's impossible to tell what the truth about any subject is.

I hate how true this statement could possibly become because it's already getting pretty bad. With that said, it's a bunch of intellectual dishonesty and people not knowing how to critically think. I would like to think that given all available information on a subject and an as objective view as possible, an individual with some level of intellect should be able to make some determinations based on real world observations. Maybe that's asking a lot from the modern day individual though.

 

[AusWolf]

I hate how true this statement could possibly become because it's already getting pretty bad. With that said, it's a bunch of intellectual dishonesty and people not knowing how to critically think. I would like to think that given all available information on a subject and an as objective view as possible, an individual with some level of intellect should be able to make some determinations based on real world observations. Maybe that's asking a lot from the modern day individual though.

I think the modern day individual has proven that anything is believable if a guy wearing a suit says it on TV. One of my neighbours still wears a surgical mask to this day while sitting alone in his car. There's way too much fakery going on even these days. With AI, it'll spin completely out of control, I'm afraid.