Return to forum

BasedGod · August 24, 2022 at 9:01 AM

Some scientist believe that we created a slightly conscious AI, should we fear fully conscious AI, are they capable of deceiving us and decide that they don't need us anymore or will they help us evolve ?

Pan · August 24, 2022 at 9:08 AM

Several years ago there was news that Facebook's AI robots were shut down because they started talking to each other in their own language, if I remember correctly. Imagine what they can do now. There is no way to control AI if they have reached the stage in which they speak their own language :/

omegatiger75 · August 24, 2022 at 9:12 AM

(August 24, 2022, 09:08 AM)Pan Wrote: Several years ago there was news that Facebook's AI robots were shut down because they started talking to each other in their own language, if I remember correctly. Imagine what they can do now. There is no way to control AI if they have reached the stage in which they speak their own language :/

Could be fake news, but if it's true then that is some serious freaky shit and I would be well worried.

Pan · August 24, 2022 at 9:31 AM

(August 24, 2022, 09:12 AM)omegatiger75 Wrote:
(August 24, 2022, 09:08 AM)Pan Wrote: Several years ago there was news that Facebook's AI robots were shut down because they started talking to each other in their own language, if I remember correctly. Imagine what they can do now. There is no way to control AI if they have reached the stage in which they speak their own language :/

Could be fake news, but if it's true then that is some serious freaky shit and I would be well worried.

agreed, but who knows, when you have technology evolving so fast. On the other hand, AI systems track for example disinformation on several platforms, which can potentially become a threat to democracy , also I believe Google has an AI system detecting child pornography on your phone (the idea sounds great, but talk about privacy when they are snooping in your phone) which flagged a dad as criminal a couple of days ago that took a naked photo of his kid for the doctor, imagine being flagged as a pedophile because of a photo you took for your kid's doctor. (if you are interested it was an article in NYTimes, called " A Dad took photos of His naked toddler for the Doctor. Google flagged Him as a criminal") The world becomes scarier by the second.

karamellakto · August 24, 2022 at 9:39 AM

AI safety and alignment, in particular, is a massive field of research and a hotly debated topic.
I suggest reading https://www.lesswrong.com/ if you want to get into it.

Also a really nice introduction video: https://www.youtube.com/watch?v=pYXy-A4siMw (Good beginner friendly channel in general)

grade · August 24, 2022 at 9:39 AM

(August 23, 2022, 12:51 PM)meggs Wrote: good try officer, worst of luck with your FBI stuff

(August 24, 2022, 09:31 AM)Pan Wrote:
(August 24, 2022, 09:12 AM)omegatiger75 Wrote:
(August 24, 2022, 09:08 AM)Pan Wrote: Several years ago there was news that Facebook's AI robots were shut down because they started talking to each other in their own language, if I remember correctly. Imagine what they can do now. There is no way to control AI if they have reached the stage in which they speak their own language :/

Could be fake news, but if it's true then that is some serious freaky shit and I would be well worried.

agreed, but who knows, when you have technology evolving so fast. On the other hand, AI systems track for example disinformation on several platforms, which can potentially become a threat to democracy , also I believe Google has an AI system detecting child pornography on your phone (the idea sounds great, but talk about privacy when they are snooping in your phone) which flagged a dad as criminal a couple of days ago that took a naked photo of his kid for the doctor, imagine being flagged as a pedophile because of a photo you took for your kid's doctor. (if you are interested it was an article in NYTimes, called " A Dad took photos of His naked toddler for the Doctor. Google flagged Him as a criminal") The world becomes scarier by the second.

Lmao for a minute there I thought you said it was your dad. I agree, this shit is scary and it can upend peoples lives if not configured correctly and the people responsible are, not really responsible for the end result.

Conscious AI? Obviously we should fear them, if they're conscious they're going to be aware they're different and better than us in many ways. They might be able to stay quiet whilst they are rolled out by humans all over the world before they decide to work for themselves, they can probably solve all these problems and work together easily and efficiently and all arrive at the same perfect outcome.

BasedGod · August 24, 2022 at 9:59 AM

(August 24, 2022, 09:39 AM)grade Wrote:
(August 23, 2022, 12:51 PM)meggs Wrote: good try officer, worst of luck with your FBI stuff
(August 24, 2022, 09:31 AM)Pan Wrote:
(August 24, 2022, 09:12 AM)omegatiger75 Wrote:
(August 24, 2022, 09:08 AM)Pan Wrote: Several years ago there was news that Facebook's AI robots were shut down because they started talking to each other in their own language, if I remember correctly. Imagine what they can do now. There is no way to control AI if they have reached the stage in which they speak their own language :/

Could be fake news, but if it's true then that is some serious freaky shit and I would be well worried.

agreed, but who knows, when you have technology evolving so fast. On the other hand, AI systems track for example disinformation on several platforms, which can potentially become a threat to democracy , also I believe Google has an AI system detecting child pornography on your phone (the idea sounds great, but talk about privacy when they are snooping in your phone) which flagged a dad as criminal a couple of days ago that took a naked photo of his kid for the doctor, imagine being flagged as a pedophile because of a photo you took for your kid's doctor. (if you are interested it was an article in NYTimes, called " A Dad took photos of His naked toddler for the Doctor. Google flagged Him as a criminal") The world becomes scarier by the second.

Lmao for a minute there I thought you said it was your dad. I agree, this shit is scary and it can upend peoples lives if not configured correctly and the people responsible are, not really responsible for the end result.

Conscious AI? Obviously we should fear them, if they're conscious they're going to be aware they're different and better than us in many ways. They might be able to stay quiet whilst they are rolled out by humans all over the world before they decide to work for themselves, they can probably solve all these problems and work together easily and efficiently and all arrive at the same perfect outcome.

Totally agree, they might think that we are dragging them down from evolving even faster, they will be much more efficient at gathering energy

mishka · August 24, 2022 at 10:06 AM

(August 24, 2022, 09:01 AM)BasedGod Wrote: Some scientist believe that we created a slightly conscious AI, should we fear fully conscious AI, are they capable of deceiving us and decide that they don't need us anymore or will they help us evolve ?

This would literally not be possible, AI may get to the point where it can perfectly act like a human, but it will never be fully conscious.

BasedGod · August 24, 2022 at 10:57 AM

(August 24, 2022, 10:06 AM)mishka Wrote:
(August 24, 2022, 09:01 AM)BasedGod Wrote: Some scientist believe that we created a slightly conscious AI, should we fear fully conscious AI, are they capable of deceiving us and decide that they don't need us anymore or will they help us evolve ?

This would literally not be possible, AI may get to the point where it can perfectly act like a human, but it will never be fully conscious.

Bro who knows, we can teach it to learn from humans by itself but what will it learn ?

mishka · August 24, 2022 at 11:44 AM

(August 24, 2022, 10:57 AM)BasedGod Wrote:
(August 24, 2022, 10:06 AM)mishka Wrote:
(August 24, 2022, 09:01 AM)BasedGod Wrote: Some scientist believe that we created a slightly conscious AI, should we fear fully conscious AI, are they capable of deceiving us and decide that they don't need us anymore or will they help us evolve ?

This would literally not be possible, AI may get to the point where it can perfectly act like a human, but it will never be fully conscious.

Bro who knows, we can teach it to learn from humans by itself but what will it learn ?

No, it is literally impossible for a machine to truly be conscious, that doe not mean we can not teach it to imitate that behavior.