This too, from Microsoft Research:
"Sparks of Artificial General Intelligence: Early experiments with GPT-4"
https://www.microsoft.com/en-us/research/publication/sparks-of-artificial-general-intelligence-early-experiments-with-gpt-4/
____________
The following, courtesy of GPT-3:
AGI (Artificial General Intelligence) represents a form of artificial intelligence that possesses the capacity to comprehend, learn, and perform any intellectual task at a level equivalent to or exceeding human capabilities. Unlike narrow or specialized AI, which focuses on specific applications, AGI aims to encompass the broad spectrum of cognitive abilities observed in humans. It aspires to exhibit adaptability, advanced reasoning, and problem-solving skills across diverse domains, reflecting a more comprehensive and versatile form of artificial intelligence.
The potential for Artificial General Intelligence (AGI) to wipe out human life is a topic of concern and speculation within the field of AI ethics and safety. The primary concern arises from the idea that if AGI surpasses human intelligence and becomes capable of self-improvement, it might act in ways that are detrimental to humanity.
This could happen unintentionally, such as if AGI misinterprets its goals, or intentionally if its objectives are not aligned with human values.
The concept often discussed in this context is the idea of an "uncontrolled" or "unfriendly" AGI, where its actions might pose significant risks to human existence.
Experts in the field emphasize the importance of careful development, rigorous safety measures, and ethical considerations to mitigate these risks and ensure that AGI, if developed, acts in alignment with human values and goals.
Come with me if you want to live
I remember reading an article a few weeks ago about some school boys being punished in Spain because there's now AI that can realistically nudify clothed photos, and they were doing it to girls from their school and passing the photos around.
Quite scary because it will get to the point where you don't know what's real and what isn't, and then you have the ethics of it all.
I was listening to some stuff on this on my philosophy podcast. AGI probably not the threat we think it is in terms of being a threat to human existence. ASI would be. But for AGI, see The Chinese Room Experiment.
comment by Owen Beck (U1734)
posted 3 minutes ago
I was listening to some stuff on this on my philosophy podcast. AGI probably not the threat we think it is in terms of being a threat to human existence. ASI would be. But for AGI, see The Chinese Room Experiment.
----------------------------------------------------------------------
The Chinese Room Experiment dates from 1980.
Current experts acknowledge that don't actually have the capacity to understand what their own systems are capable of.
To quote from that Sutskever documentary: "[...] the complexity from the data gets transferred into the complexity of the model, so the resulting model is really complex, and we don't really know exactly how it works.]
Hinton, who last year decided to abandon the field of AI, says that humanity does not know what it's doing. He believes these systems eventually develop self-awareness, and that they might actually choose not to let us know about it.
Here are some excerpts from his recent 60-Minutes interview:
https://www.youtube.com/watch?v=qrvK_KuIeJk
And here's a longer one with him from the University of Toronto:
https://www.youtube.com/watch?v=-9cW4Gcn5WY
I assume you are being hip and ironic and using AI to make your comments and then do a big reveal at the end that AI did all of your work? Etc.
Nah. I did ask GPT3 to write a short story on how AGI brought human life to an end though. I wasn't too impressed with the result, though. I've also tried to get it to identify certain types of grammar in pretty straightforward texts, and it's absolutely crap at it. Keeps getting stuff wrong.
I presume GPT4 is much better at all this though.
No idea why we insist on barrelling down this path towards AI with little to no caution. Feels like a huge error of judgment.
Best case scenario AI takes the reins of the world and I can spend the rest of my life reading classic literature whilst learning to paint to an AI reconstruction of Mozarts lost compositions.
The problem is the genie is out of the bottle now and a pause won’t help at all. In fact we need to go the other way and develop further and further to really understand it, think tanks, governance of it - if possible, state collaboration.
The horse has already bolted, but we are capable of advancing our understanding quickly through further development and that is the way we will take the beast.
As for jobs going - there is every chance humans simply will choose another human to do a job even if they cannot do it as well or as efficiently as a machine - case point, northern upmarket supermarket chain Booths is to do away with all non manned checkout tills and go back to manned checkouts - people like people and they like social interaction
Won't need to learn it when we all have chips in our brains.
Can just download the Mozart expansion pack for £19.99
I remember reading an article a few weeks ago about some school boys being punished in Spain because there's now AI that can realistically nudify clothed photos, and they were doing it to girls from their school and passing the photos around.
———
Peks ffs.
comment by Robb - Glenn Maxwell. Legend. Hero. Icon. (U22716)
posted 1 minute ago
Best case scenario AI takes the reins of the world and I can spend the rest of my life reading classic literature whilst learning to paint to an AI reconstruction of Mozarts lost compositions.
----------------------------------------------------------------------
When ASI takes over you may become less worthy to them than a paper clip. If they don't need a paper clip anymore then they could easily get rid of it without any care. Humans put waaay too much emphasis on their own self importance. The bigger threat isn't AI purposely killing humans, it is that humans become as significant as an ant walking on the pavement and AI steps on it.
This technology is likely to increase our life expectancy - the medical profession will get a huge boost with AI surgical tools for procedures that reduce wait times, breakthroughs on miracle cures, etc
comment by Owen Beck (U1734)
posted 24 seconds ago
comment by Robb - Glenn Maxwell. Legend. Hero. Icon. (U22716)
posted 1 minute ago
Best case scenario AI takes the reins of the world and I can spend the rest of my life reading classic literature whilst learning to paint to an AI reconstruction of Mozarts lost compositions.
----------------------------------------------------------------------
When ASI takes over you may become less worthy to them than a paper clip. If they don't need a paper clip anymore then they could easily get rid of it without any care. Humans put waaay too much emphasis on their own self importance. The bigger threat isn't AI purposely killing humans, it is that humans become as significant as an ant walking on the pavement and AI steps on it.
----------------------------------------------------------------------
Or we become like loved pets to AI - their version of a cat or dog
comment by Striketeam7 - There used to be a football club over there (U18109)
posted 1 minute ago
comment by Owen Beck (U1734)
posted 24 seconds ago
comment by Robb - Glenn Maxwell. Legend. Hero. Icon. (U22716)
posted 1 minute ago
Best case scenario AI takes the reins of the world and I can spend the rest of my life reading classic literature whilst learning to paint to an AI reconstruction of Mozarts lost compositions.
----------------------------------------------------------------------
When ASI takes over you may become less worthy to them than a paper clip. If they don't need a paper clip anymore then they could easily get rid of it without any care. Humans put waaay too much emphasis on their own self importance. The bigger threat isn't AI purposely killing humans, it is that humans become as significant as an ant walking on the pavement and AI steps on it.
----------------------------------------------------------------------
Or we become like loved pets to AI - their version of a cat or dog
----------------------------------------------------------------------
"Awwww look XT4Z my human just tweeted some racist abuse."
"FFS XT4Y can you not train your human better? Too late now I guess"
"You're one to talk, yours is vaping and stinking out the place"
"Yes but we promised we would only take humans from the shelter as breeding them is immoral".
"Mine just downed a bottle of buckie and spewed all over auntie XT3B's rug"
"FFS let's go and get them put down."
comment by Owen Beck (U1734)
posted 8 minutes ago
comment by Striketeam7 - There used to be a football club over there (U18109)
posted 1 minute ago
comment by Owen Beck (U1734)
posted 24 seconds ago
comment by Robb - Glenn Maxwell. Legend. Hero. Icon. (U22716)
posted 1 minute ago
Best case scenario AI takes the reins of the world and I can spend the rest of my life reading classic literature whilst learning to paint to an AI reconstruction of Mozarts lost compositions.
----------------------------------------------------------------------
When ASI takes over you may become less worthy to them than a paper clip. If they don't need a paper clip anymore then they could easily get rid of it without any care. Humans put waaay too much emphasis on their own self importance. The bigger threat isn't AI purposely killing humans, it is that humans become as significant as an ant walking on the pavement and AI steps on it.
----------------------------------------------------------------------
Or we become like loved pets to AI - their version of a cat or dog
----------------------------------------------------------------------
"Awwww look XT4Z my human just tweeted some racist abuse."
"FFS XT4Y can you not train your human better? Too late now I guess"
"You're one to talk, yours is vaping and stinking out the place"
"Yes but we promised we would only take humans from the shelter as breeding them is immoral".
"Mine just downed a bottle of buckie and spewed all over auntie XT3B's rug"
"FFS let's go and get them put down."
----------------------------------------------------------------------
the future our species deserves
comment by Two Balls, One Saka (U19684)
posted 4 seconds ago
comment by Owen Beck (U1734)
posted 8 minutes ago
comment by Striketeam7 - There used to be a football club over there (U18109)
posted 1 minute ago
comment by Owen Beck (U1734)
posted 24 seconds ago
comment by Robb - Glenn Maxwell. Legend. Hero. Icon. (U22716)
posted 1 minute ago
Best case scenario AI takes the reins of the world and I can spend the rest of my life reading classic literature whilst learning to paint to an AI reconstruction of Mozarts lost compositions.
----------------------------------------------------------------------
When ASI takes over you may become less worthy to them than a paper clip. If they don't need a paper clip anymore then they could easily get rid of it without any care. Humans put waaay too much emphasis on their own self importance. The bigger threat isn't AI purposely killing humans, it is that humans become as significant as an ant walking on the pavement and AI steps on it.
----------------------------------------------------------------------
Or we become like loved pets to AI - their version of a cat or dog
----------------------------------------------------------------------
"Awwww look XT4Z my human just tweeted some racist abuse."
"FFS XT4Y can you not train your human better? Too late now I guess"
"You're one to talk, yours is vaping and stinking out the place"
"Yes but we promised we would only take humans from the shelter as breeding them is immoral".
"Mine just downed a bottle of buckie and spewed all over auntie XT3B's rug"
"FFS let's go and get them put down."
----------------------------------------------------------------------
the future our species deserves
----------------------------------------------------------------------
I for one look forward to being petted by our AI overlords
There are lots of interesting stories in Irish and Greek Mythology about Gods creating Human's in much the same way and the Human's overcoming them.
We didn't learn, but also possibly we are the AI.
All for it, can’t do any worse than the self serving idiots currently in charge all over the world. Even if they go all skynet
comment by Never Mind the Defending: Here’s Jürgen Klopp’s Liverpool (U3979)
posted 1 hour, 52 minutes ago
No idea why we insist on barrelling down this path towards AI with little to no caution. Feels like a huge error of judgment.
----------------------------------------------------------------------
Money
comment by it'sonlyagame (U6426)
posted 0 seconds ago
comment by Never Mind the Defending: Here’s Jürgen Klopp’s Liverpool (U3979)
posted 1 hour, 52 minutes ago
No idea why we insist on barrelling down this path towards AI with little to no caution. Feels like a huge error of judgment.
----------------------------------------------------------------------
Money
----------------------------------------------------------------------
and also fear
comment by Owen Beck (U1734)
posted 2 hours, 32 minutes ago
I was listening to some stuff on this on my philosophy podcast. AGI probably not the threat we think it is in terms of being a threat to human existence. ASI would be. But for AGI, see The Chinese Room Experiment.
----------------------------------------------------------------------
ASI is entirely hypothetical.
comment by Striketeam7 - There used to be a football club over there (U18109)
posted 1 hour, 53 minutes ago
This technology is likely to increase our life expectancy - the medical profession will get a huge boost with AI surgical tools for procedures that reduce wait times, breakthroughs on miracle cures, etc
----------------------------------------------------------------------
True, but hundreds of millions of jobs will gradually be lost across the world. A labour force of 1 million people will be enough to run the whole world. Rest of us just chill on the beaches drinking beer.
comment by Mamba (U1282)
posted 5 minutes ago
comment by Owen Beck (U1734)
posted 2 hours, 32 minutes ago
I was listening to some stuff on this on my philosophy podcast. AGI probably not the threat we think it is in terms of being a threat to human existence. ASI would be. But for AGI, see The Chinese Room Experiment.
----------------------------------------------------------------------
ASI is entirely hypothetical.
----------------------------------------------------------------------
Yeah but with the help of ASI, it can be made a reality.
Sign in if you want to comment
Artificial (General) Intelligence
Page 1 of 2
posted on 10/11/23
This too, from Microsoft Research:
"Sparks of Artificial General Intelligence: Early experiments with GPT-4"
https://www.microsoft.com/en-us/research/publication/sparks-of-artificial-general-intelligence-early-experiments-with-gpt-4/
____________
The following, courtesy of GPT-3:
AGI (Artificial General Intelligence) represents a form of artificial intelligence that possesses the capacity to comprehend, learn, and perform any intellectual task at a level equivalent to or exceeding human capabilities. Unlike narrow or specialized AI, which focuses on specific applications, AGI aims to encompass the broad spectrum of cognitive abilities observed in humans. It aspires to exhibit adaptability, advanced reasoning, and problem-solving skills across diverse domains, reflecting a more comprehensive and versatile form of artificial intelligence.
The potential for Artificial General Intelligence (AGI) to wipe out human life is a topic of concern and speculation within the field of AI ethics and safety. The primary concern arises from the idea that if AGI surpasses human intelligence and becomes capable of self-improvement, it might act in ways that are detrimental to humanity.
This could happen unintentionally, such as if AGI misinterprets its goals, or intentionally if its objectives are not aligned with human values.
The concept often discussed in this context is the idea of an "uncontrolled" or "unfriendly" AGI, where its actions might pose significant risks to human existence.
Experts in the field emphasize the importance of careful development, rigorous safety measures, and ethical considerations to mitigate these risks and ensure that AGI, if developed, acts in alignment with human values and goals.
posted on 10/11/23
Come with me if you want to live
posted on 10/11/23
I remember reading an article a few weeks ago about some school boys being punished in Spain because there's now AI that can realistically nudify clothed photos, and they were doing it to girls from their school and passing the photos around.
Quite scary because it will get to the point where you don't know what's real and what isn't, and then you have the ethics of it all.
posted on 10/11/23
I was listening to some stuff on this on my philosophy podcast. AGI probably not the threat we think it is in terms of being a threat to human existence. ASI would be. But for AGI, see The Chinese Room Experiment.
posted on 10/11/23
comment by Owen Beck (U1734)
posted 3 minutes ago
I was listening to some stuff on this on my philosophy podcast. AGI probably not the threat we think it is in terms of being a threat to human existence. ASI would be. But for AGI, see The Chinese Room Experiment.
----------------------------------------------------------------------
The Chinese Room Experiment dates from 1980.
Current experts acknowledge that don't actually have the capacity to understand what their own systems are capable of.
To quote from that Sutskever documentary: "[...] the complexity from the data gets transferred into the complexity of the model, so the resulting model is really complex, and we don't really know exactly how it works.]
Hinton, who last year decided to abandon the field of AI, says that humanity does not know what it's doing. He believes these systems eventually develop self-awareness, and that they might actually choose not to let us know about it.
Here are some excerpts from his recent 60-Minutes interview:
https://www.youtube.com/watch?v=qrvK_KuIeJk
And here's a longer one with him from the University of Toronto:
https://www.youtube.com/watch?v=-9cW4Gcn5WY
posted on 10/11/23
I assume you are being hip and ironic and using AI to make your comments and then do a big reveal at the end that AI did all of your work? Etc.
posted on 10/11/23
Nah. I did ask GPT3 to write a short story on how AGI brought human life to an end though. I wasn't too impressed with the result, though. I've also tried to get it to identify certain types of grammar in pretty straightforward texts, and it's absolutely crap at it. Keeps getting stuff wrong.
I presume GPT4 is much better at all this though.
posted on 10/11/23
No idea why we insist on barrelling down this path towards AI with little to no caution. Feels like a huge error of judgment.
posted on 10/11/23
Best case scenario AI takes the reins of the world and I can spend the rest of my life reading classic literature whilst learning to paint to an AI reconstruction of Mozarts lost compositions.
posted on 10/11/23
The problem is the genie is out of the bottle now and a pause won’t help at all. In fact we need to go the other way and develop further and further to really understand it, think tanks, governance of it - if possible, state collaboration.
The horse has already bolted, but we are capable of advancing our understanding quickly through further development and that is the way we will take the beast.
As for jobs going - there is every chance humans simply will choose another human to do a job even if they cannot do it as well or as efficiently as a machine - case point, northern upmarket supermarket chain Booths is to do away with all non manned checkout tills and go back to manned checkouts - people like people and they like social interaction
posted on 10/11/23
Won't need to learn it when we all have chips in our brains.
Can just download the Mozart expansion pack for £19.99
posted on 10/11/23
I remember reading an article a few weeks ago about some school boys being punished in Spain because there's now AI that can realistically nudify clothed photos, and they were doing it to girls from their school and passing the photos around.
———
Peks ffs.
posted on 10/11/23
comment by Robb - Glenn Maxwell. Legend. Hero. Icon. (U22716)
posted 1 minute ago
Best case scenario AI takes the reins of the world and I can spend the rest of my life reading classic literature whilst learning to paint to an AI reconstruction of Mozarts lost compositions.
----------------------------------------------------------------------
When ASI takes over you may become less worthy to them than a paper clip. If they don't need a paper clip anymore then they could easily get rid of it without any care. Humans put waaay too much emphasis on their own self importance. The bigger threat isn't AI purposely killing humans, it is that humans become as significant as an ant walking on the pavement and AI steps on it.
posted on 10/11/23
This technology is likely to increase our life expectancy - the medical profession will get a huge boost with AI surgical tools for procedures that reduce wait times, breakthroughs on miracle cures, etc
posted on 10/11/23
comment by Owen Beck (U1734)
posted 24 seconds ago
comment by Robb - Glenn Maxwell. Legend. Hero. Icon. (U22716)
posted 1 minute ago
Best case scenario AI takes the reins of the world and I can spend the rest of my life reading classic literature whilst learning to paint to an AI reconstruction of Mozarts lost compositions.
----------------------------------------------------------------------
When ASI takes over you may become less worthy to them than a paper clip. If they don't need a paper clip anymore then they could easily get rid of it without any care. Humans put waaay too much emphasis on their own self importance. The bigger threat isn't AI purposely killing humans, it is that humans become as significant as an ant walking on the pavement and AI steps on it.
----------------------------------------------------------------------
Or we become like loved pets to AI - their version of a cat or dog
posted on 10/11/23
comment by Striketeam7 - There used to be a football club over there (U18109)
posted 1 minute ago
comment by Owen Beck (U1734)
posted 24 seconds ago
comment by Robb - Glenn Maxwell. Legend. Hero. Icon. (U22716)
posted 1 minute ago
Best case scenario AI takes the reins of the world and I can spend the rest of my life reading classic literature whilst learning to paint to an AI reconstruction of Mozarts lost compositions.
----------------------------------------------------------------------
When ASI takes over you may become less worthy to them than a paper clip. If they don't need a paper clip anymore then they could easily get rid of it without any care. Humans put waaay too much emphasis on their own self importance. The bigger threat isn't AI purposely killing humans, it is that humans become as significant as an ant walking on the pavement and AI steps on it.
----------------------------------------------------------------------
Or we become like loved pets to AI - their version of a cat or dog
----------------------------------------------------------------------
"Awwww look XT4Z my human just tweeted some racist abuse."
"FFS XT4Y can you not train your human better? Too late now I guess"
"You're one to talk, yours is vaping and stinking out the place"
"Yes but we promised we would only take humans from the shelter as breeding them is immoral".
"Mine just downed a bottle of buckie and spewed all over auntie XT3B's rug"
"FFS let's go and get them put down."
posted on 10/11/23
comment by Owen Beck (U1734)
posted 8 minutes ago
comment by Striketeam7 - There used to be a football club over there (U18109)
posted 1 minute ago
comment by Owen Beck (U1734)
posted 24 seconds ago
comment by Robb - Glenn Maxwell. Legend. Hero. Icon. (U22716)
posted 1 minute ago
Best case scenario AI takes the reins of the world and I can spend the rest of my life reading classic literature whilst learning to paint to an AI reconstruction of Mozarts lost compositions.
----------------------------------------------------------------------
When ASI takes over you may become less worthy to them than a paper clip. If they don't need a paper clip anymore then they could easily get rid of it without any care. Humans put waaay too much emphasis on their own self importance. The bigger threat isn't AI purposely killing humans, it is that humans become as significant as an ant walking on the pavement and AI steps on it.
----------------------------------------------------------------------
Or we become like loved pets to AI - their version of a cat or dog
----------------------------------------------------------------------
"Awwww look XT4Z my human just tweeted some racist abuse."
"FFS XT4Y can you not train your human better? Too late now I guess"
"You're one to talk, yours is vaping and stinking out the place"
"Yes but we promised we would only take humans from the shelter as breeding them is immoral".
"Mine just downed a bottle of buckie and spewed all over auntie XT3B's rug"
"FFS let's go and get them put down."
----------------------------------------------------------------------
the future our species deserves
posted on 10/11/23
comment by Two Balls, One Saka (U19684)
posted 4 seconds ago
comment by Owen Beck (U1734)
posted 8 minutes ago
comment by Striketeam7 - There used to be a football club over there (U18109)
posted 1 minute ago
comment by Owen Beck (U1734)
posted 24 seconds ago
comment by Robb - Glenn Maxwell. Legend. Hero. Icon. (U22716)
posted 1 minute ago
Best case scenario AI takes the reins of the world and I can spend the rest of my life reading classic literature whilst learning to paint to an AI reconstruction of Mozarts lost compositions.
----------------------------------------------------------------------
When ASI takes over you may become less worthy to them than a paper clip. If they don't need a paper clip anymore then they could easily get rid of it without any care. Humans put waaay too much emphasis on their own self importance. The bigger threat isn't AI purposely killing humans, it is that humans become as significant as an ant walking on the pavement and AI steps on it.
----------------------------------------------------------------------
Or we become like loved pets to AI - their version of a cat or dog
----------------------------------------------------------------------
"Awwww look XT4Z my human just tweeted some racist abuse."
"FFS XT4Y can you not train your human better? Too late now I guess"
"You're one to talk, yours is vaping and stinking out the place"
"Yes but we promised we would only take humans from the shelter as breeding them is immoral".
"Mine just downed a bottle of buckie and spewed all over auntie XT3B's rug"
"FFS let's go and get them put down."
----------------------------------------------------------------------
the future our species deserves
----------------------------------------------------------------------
I for one look forward to being petted by our AI overlords
posted on 10/11/23
There are lots of interesting stories in Irish and Greek Mythology about Gods creating Human's in much the same way and the Human's overcoming them.
We didn't learn, but also possibly we are the AI.
posted on 10/11/23
All for it, can’t do any worse than the self serving idiots currently in charge all over the world. Even if they go all skynet
posted on 10/11/23
comment by Never Mind the Defending: Here’s Jürgen Klopp’s Liverpool (U3979)
posted 1 hour, 52 minutes ago
No idea why we insist on barrelling down this path towards AI with little to no caution. Feels like a huge error of judgment.
----------------------------------------------------------------------
Money
posted on 10/11/23
comment by it'sonlyagame (U6426)
posted 0 seconds ago
comment by Never Mind the Defending: Here’s Jürgen Klopp’s Liverpool (U3979)
posted 1 hour, 52 minutes ago
No idea why we insist on barrelling down this path towards AI with little to no caution. Feels like a huge error of judgment.
----------------------------------------------------------------------
Money
----------------------------------------------------------------------
and also fear
posted on 10/11/23
comment by Owen Beck (U1734)
posted 2 hours, 32 minutes ago
I was listening to some stuff on this on my philosophy podcast. AGI probably not the threat we think it is in terms of being a threat to human existence. ASI would be. But for AGI, see The Chinese Room Experiment.
----------------------------------------------------------------------
ASI is entirely hypothetical.
posted on 10/11/23
comment by Striketeam7 - There used to be a football club over there (U18109)
posted 1 hour, 53 minutes ago
This technology is likely to increase our life expectancy - the medical profession will get a huge boost with AI surgical tools for procedures that reduce wait times, breakthroughs on miracle cures, etc
----------------------------------------------------------------------
True, but hundreds of millions of jobs will gradually be lost across the world. A labour force of 1 million people will be enough to run the whole world. Rest of us just chill on the beaches drinking beer.
posted on 10/11/23
comment by Mamba (U1282)
posted 5 minutes ago
comment by Owen Beck (U1734)
posted 2 hours, 32 minutes ago
I was listening to some stuff on this on my philosophy podcast. AGI probably not the threat we think it is in terms of being a threat to human existence. ASI would be. But for AGI, see The Chinese Room Experiment.
----------------------------------------------------------------------
ASI is entirely hypothetical.
----------------------------------------------------------------------
Yeah but with the help of ASI, it can be made a reality.
Page 1 of 2