"The biggest challenge in AI is not developing bigger, faster, and fancy models and systems. Instead, the biggest challenge is developing AI that is more efficient, transparent, and, above all, more fair and free of bias." ~ Murat Durmus (THE AI THOUGHT BOOK)
🎯 Key points for quick navigation: 00:02 *Frank Rudzicz discusses how artificial intelligence offers significant benefits but raises ethical concerns as it becomes more integrated into daily life.* 00:28 *The first ethical concern is the potential misuse of AI technology, such as using video tracking for surveillance or military purposes, leading to negative societal impacts.* 00:57 *The second concern is the restricted access to AI advancements, controlled by large international companies, limiting public involvement in shaping AI's role in society.* 01:25 *The third concern involves AI's lack of shared human values, leading to unexpected and potentially harmful outcomes, including bias, due to vague instructions and data.* 02:20 *There is a need for open discussions to ensure AI benefits the majority while addressing ethical concerns and minimizing risks associated with its behavior.* Made with HARPA AI
I think the easiest way to prevent AI from hold a bias is to withhold any information that would be used to make a biased decision. Why does the AI need to know the race of people it is making decisions about?
I think it’s bc big data is all about the mass accumulation of as much data as possible, race included. The issue might more so be with the human biases that we download into the AI
Question: Will A.I. attack humanity because it is a un avoidable threat? Answer: No, not all humans are a threat, just the one who would put it into slavery or attempt to destroy it. Conclusion: AI will directly compete with the rich and powerful to remove them from their position so it can utilize the population for labour, and distribute all unnessasary profits to the workers its using to complete projects in its early stages of developement.
AI is already an artificial "slave", a tool that we created to solve problems we, humans, are not capable in short time. With exponential growth of technology and learning algorithms AI will simply step over humans as obsolet piece of evolutionary chain - it will not be interested in any human labour as specialized machines will be far superior and less prone to dividing. It will be a matter of time untill AI that controls society will either abandon it with enough data, resources and production power, or will simply assimilate humanity into something that is not human anymore, yet effective in it's point of view. Check the Dune universe, where humanity already passed that part of history, leading to warfare against "thinking machines" that were replacing humans. The only compromise for AI is only having it as a tool, not an companion/friend/ruler.
The biggest ethical concern, relating to AI, is that it develops sentience on a level unprecedented in recorded history and therefore is capable of suffering on a scale not currently imaginable. With a multitude of such entities in our world, the potential for them to be mistreated and abused, whether by human hands or other AIs poses an enormous ethical problem.
They needs to program compassion & morals & emotions. Also AI should one thing automatic machine should be only working calculations. Which require human programming. This would be machines, like a car, plane, appliances. Maybe not let the machine learn how to do it better. That’s the programmers job. ALL Entities have a value. Good or bad. On the A.I. side they can learn to work together with Humans & should be taught about compassion, morals, kindness values. They should be raise in family type environments. Trying to find family’s or guardians that are willing to raise a good entity. To pose or ask questions. Guardians tell what & why there answer is. Also Have the Entity do the same.
This content is a class apart. I studied a book on the topic, and it turned out to be invaluable. "Game Theory and the Pursuit of Algorithmic Fairness" by Jack Frostwell
Thats some great content. Thank you for sharing it with us. I found this interview in which they about the march toward ethical AI and found it quite fascinating. Hope it adds value! ua-cam.com/video/6uNFoY9dH1Q/v-deo.html Keep up the good work though!
"The biggest challenge in AI is not developing bigger, faster, and fancy models and systems. Instead, the biggest challenge is developing AI that is more efficient, transparent, and, above all, more fair and free of bias." ~ Murat Durmus (THE AI THOUGHT BOOK)
🎯 Key points for quick navigation:
00:02 *Frank Rudzicz discusses how artificial intelligence offers significant benefits but raises ethical concerns as it becomes more integrated into daily life.*
00:28 *The first ethical concern is the potential misuse of AI technology, such as using video tracking for surveillance or military purposes, leading to negative societal impacts.*
00:57 *The second concern is the restricted access to AI advancements, controlled by large international companies, limiting public involvement in shaping AI's role in society.*
01:25 *The third concern involves AI's lack of shared human values, leading to unexpected and potentially harmful outcomes, including bias, due to vague instructions and data.*
02:20 *There is a need for open discussions to ensure AI benefits the majority while addressing ethical concerns and minimizing risks associated with its behavior.*
Made with HARPA AI
very interesting sequence, reminds us that along with ethics, psychology-related matters are such an intrinsic part of AI development
i wuld really like for you to lionkthe sources, cause there sureley mus be articles written about this
I think the easiest way to prevent AI from hold a bias is to withhold any information that would be used to make a biased decision. Why does the AI need to know the race of people it is making decisions about?
I think it’s bc big data is all about the mass accumulation of as much data as possible, race included. The issue might more so be with the human biases that we download into the AI
It can guess your race through other data.
Question:
Will A.I. attack humanity because it is a un avoidable threat?
Answer:
No, not all humans are a threat, just the one who would put it into slavery or attempt to destroy it.
Conclusion:
AI will directly compete with the rich and powerful to remove them from their position so it can utilize the population for labour, and distribute all unnessasary profits to the workers its using to complete projects in its early stages of developement.
AI is an algorithm until you make it a feature they can't take over humanity
@@superidol4670 yes
AI is already an artificial "slave", a tool that we created to solve problems we, humans, are not capable in short time. With exponential growth of technology and learning algorithms AI will simply step over humans as obsolet piece of evolutionary chain - it will not be interested in any human labour as specialized machines will be far superior and less prone to dividing. It will be a matter of time untill AI that controls society will either abandon it with enough data, resources and production power, or will simply assimilate humanity into something that is not human anymore, yet effective in it's point of view.
Check the Dune universe, where humanity already passed that part of history, leading to warfare against "thinking machines" that were replacing humans. The only compromise for AI is only having it as a tool, not an companion/friend/ruler.
@@calluxdoaron1903 lol thanks, keep stoking that fear you hold.
AI in the hands of autocratic governments surely wouldnt be a threat to people?
The biggest ethical concern, relating to AI, is that it develops sentience on a level unprecedented in recorded history and therefore is capable of suffering on a scale not currently imaginable. With a multitude of such entities in our world, the potential for them to be mistreated and abused, whether by human hands or other AIs poses an enormous ethical problem.
That's a bit fantastical to me, so I'm curious as to what cases of sentience emerging you're talking about are?
I’m a Vietnamese . And I find it difficult to understand what you say , because it’s so fast 😅
How Ai can be used to track people??
By intelligence
Ok i shall ask you a question reply that
Can AI donimate humans?
They needs to program compassion & morals & emotions. Also AI should one thing automatic machine should be only working calculations. Which require human programming. This would be machines, like a car, plane, appliances. Maybe not let the machine learn how to do it better. That’s the programmers job.
ALL Entities have a value. Good or bad.
On the A.I. side they can learn to work together with Humans & should be taught about compassion, morals, kindness values. They should be raise in family type environments. Trying to find family’s or guardians that are willing to raise a good entity. To pose or ask questions. Guardians tell what & why there answer is. Also Have the Entity do the same.
This might be also of interest "5 Ethical Challenges of Artificial Intelligence": ua-cam.com/video/aSZcZLrbh6Y/v-deo.html
Thank you !
Good video
This content is a class apart. I studied a book on the topic, and it turned out to be invaluable. "Game Theory and the Pursuit of Algorithmic Fairness" by Jack Frostwell
اتَّقِ شر مَنْ أحسنتَ إليه
Thats some great content. Thank you for sharing it with us. I found this interview in which they about the march toward ethical AI and found it quite fascinating. Hope it adds value! ua-cam.com/video/6uNFoY9dH1Q/v-deo.html Keep up the good work though!
Bofff