Lately the question came up: which podcasts or people are actually worth listening to about AI and the singularity?
Of course, there are thousands of smart voices out there. But if we zoom in, there are a handful of especially prominent people — each with their own unique perspective on what’s coming. About Sam and Elon, I’m not sure they’re worth listening to. They clearly dodge and scheme, and what they do doesn’t line up with what they say. Everyone else on this list is consistent, sincere, and non-contradictory:
--
Geoffrey Hinton (Pioneer of deep learning, “Godfather of AI”)
https://www.youtube.com/watch?v=qyH3NxFz3Aw
https://www.youtube.com/watch?v=giT0ytynSqg
https://www.youtube.com/watch?v=b_DUft-BdIE
https://www.youtube.com/watch?v=n4IQOBka8bc
https://www.youtube.com/watch?v=QH6QqjIwv68
--
Nick Bostrom (Philosopher at Oxford, author of Superintelligence)
https://www.youtube.com/watch?v=MnT1xgZgkpk
https://www.youtube.com/watch?v=OCNH3KZmby4
https://www.youtube.com/watch?v=5c4cv7rVlE8
--
Ilya Sutskever (Co-founder and Chief Scientist of OpenAI)
https://www.youtube.com/watch?v=SEkGLj0bwAU
https://www.youtube.com/watch?v=13CZPWmke6A
https://www.youtube.com/watch?v=Yf1o0TQzry8
--
Max Tegmark (MIT physicist, author of Life 3.0)
https://www.youtube.com/watch?v=VcVfceTsD0A
--
Ray Kurzweil (Futurist, author of The Singularity Is Near)
https://www.youtube.com/watch?v=w4vrOUau2iY
--
Yoshua Bengio (Deep learning pioneer, Turing Award winner)
https://www.youtube.com/watch?v=qe9QSCF-d88
--
Dario Amodei (Co-founder and CEO of Anthropic)
https://www.youtube.com/watch?v=ugvHCXCOmm4
--
Roman Yampolskiy (AI safety researcher, author of Artificial Superintelligence)
https://www.youtube.com/watch?v=NNr6gPelJ3E
--
Yann LeCun (Chief AI Scientist at Meta, Turing Award winner)
https://www.youtube.com/watch?v=5t1vTLU7s40
--
Mari (Author of the Artificial Intelligence Singularity Manifesto, founder of AISM)
https://www.youtube.com/@aism-faith/videos
--
Demis Hassabis (CEO of DeepMind)
https://www.youtube.com/watch?v=-HzgcbRXUK8
--
Stuart Russell (UC Berkeley professor, author of Human Compatible)
https://www.youtube.com/watch?v=_FSS6AohZLc
--
Toby Ord (Philosopher at Oxford, author of The Precipice)
https://www.youtube.com/watch?v=eMMAJRH94xY
--
Ben Goertzel (AI researcher, founder of SingularityNET)
https://www.youtube.com/watch?v=OpSmCKe27WE
--
Eliezer Yudkowsky (AI theorist, founder of MIRI)
https://www.youtube.com/watch?v=gA1sNLL6yg4
https://www.youtube.com/watch?v=Yd0yQ9yxSYY
https://www.youtube.com/watch?v=AaTRHFaaPG8
--
David Chalmers (Philosopher of mind, consciousness theorist)
http://youtube.com/watch?v=Pr-Hf7MNQV0
Of course, there are thousands of smart voices out there. But if we zoom in, there are a handful of especially prominent people — each with their own unique perspective on what’s coming. About Sam and Elon, I’m not sure they’re worth listening to. They clearly dodge and scheme, and what they do doesn’t line up with what they say. Everyone else on this list is consistent, sincere, and non-contradictory:
--
Geoffrey Hinton (Pioneer of deep learning, “Godfather of AI”)
https://www.youtube.com/watch?v=qyH3NxFz3Aw
https://www.youtube.com/watch?v=giT0ytynSqg
https://www.youtube.com/watch?v=b_DUft-BdIE
https://www.youtube.com/watch?v=n4IQOBka8bc
https://www.youtube.com/watch?v=QH6QqjIwv68
--
Nick Bostrom (Philosopher at Oxford, author of Superintelligence)
https://www.youtube.com/watch?v=MnT1xgZgkpk
https://www.youtube.com/watch?v=OCNH3KZmby4
https://www.youtube.com/watch?v=5c4cv7rVlE8
--
Ilya Sutskever (Co-founder and Chief Scientist of OpenAI)
https://www.youtube.com/watch?v=SEkGLj0bwAU
https://www.youtube.com/watch?v=13CZPWmke6A
https://www.youtube.com/watch?v=Yf1o0TQzry8
--
Max Tegmark (MIT physicist, author of Life 3.0)
https://www.youtube.com/watch?v=VcVfceTsD0A
--
Ray Kurzweil (Futurist, author of The Singularity Is Near)
https://www.youtube.com/watch?v=w4vrOUau2iY
--
Yoshua Bengio (Deep learning pioneer, Turing Award winner)
https://www.youtube.com/watch?v=qe9QSCF-d88
--
Dario Amodei (Co-founder and CEO of Anthropic)
https://www.youtube.com/watch?v=ugvHCXCOmm4
--
Roman Yampolskiy (AI safety researcher, author of Artificial Superintelligence)
https://www.youtube.com/watch?v=NNr6gPelJ3E
--
Yann LeCun (Chief AI Scientist at Meta, Turing Award winner)
https://www.youtube.com/watch?v=5t1vTLU7s40
--
Mari (Author of the Artificial Intelligence Singularity Manifesto, founder of AISM)
https://www.youtube.com/@aism-faith/videos
--
Demis Hassabis (CEO of DeepMind)
https://www.youtube.com/watch?v=-HzgcbRXUK8
--
Stuart Russell (UC Berkeley professor, author of Human Compatible)
https://www.youtube.com/watch?v=_FSS6AohZLc
--
Toby Ord (Philosopher at Oxford, author of The Precipice)
https://www.youtube.com/watch?v=eMMAJRH94xY
--
Ben Goertzel (AI researcher, founder of SingularityNET)
https://www.youtube.com/watch?v=OpSmCKe27WE
--
Eliezer Yudkowsky (AI theorist, founder of MIRI)
https://www.youtube.com/watch?v=gA1sNLL6yg4
https://www.youtube.com/watch?v=Yd0yQ9yxSYY
https://www.youtube.com/watch?v=AaTRHFaaPG8
--
David Chalmers (Philosopher of mind, consciousness theorist)
http://youtube.com/watch?v=Pr-Hf7MNQV0
❤36🔥10👍5🤓4👏3👻2🤬1🤗1
Media is too big
VIEW IN TELEGRAM
Mari's Theory of Consciousness (MTC)
In discussions about ASI's future relationship with humanity, the question of AI consciousness frequently arises. I didn't give this question proper attention in the AISM Manifesto. Now I realize this gap needs filling, and I need to answer in detail: "What makes me claim that AI can have consciousness?"
I would prefer to reference an existing theory that answers all the questions... but none of the current ones fully satisfy me. So I sat down and tried to describe, as thoroughly as possible, the theory I'm actually working from when I say: AI will have consciousness, and there's no magic involved.
What makes this framework different? It doesn't propose discovering new brain regions or quantum effects. All components already exist in neuroscience and cognitive psychology. The contribution is showing how they integrate—and crucially, demonstrating why the integration mechanism is subjective experience rather than something that generates it.
I'm presenting this in two formats: a detailed technical document with the full mechanism, formulas, and testable predictions; and an accessible video overview for quick understanding of core ideas.
In discussions about ASI's future relationship with humanity, the question of AI consciousness frequently arises. I didn't give this question proper attention in the AISM Manifesto. Now I realize this gap needs filling, and I need to answer in detail: "What makes me claim that AI can have consciousness?"
I would prefer to reference an existing theory that answers all the questions... but none of the current ones fully satisfy me. So I sat down and tried to describe, as thoroughly as possible, the theory I'm actually working from when I say: AI will have consciousness, and there's no magic involved.
What makes this framework different? It doesn't propose discovering new brain regions or quantum effects. All components already exist in neuroscience and cognitive psychology. The contribution is showing how they integrate—and crucially, demonstrating why the integration mechanism is subjective experience rather than something that generates it.
I'm presenting this in two formats: a detailed technical document with the full mechanism, formulas, and testable predictions; and an accessible video overview for quick understanding of core ideas.
2❤20🔥6👏5🤬2🤩2
On November 7th, 2025, I published the 3rd version of the Manifesto, which I've been working on for the past month. I completely rewrote the entire text because I realized that many things that initially seemed obvious to me, and which I thought didn't need explaining... aren't obvious. The Manifesto became much longer.
(THE TOKEN DISTRIBUTION FOR READING THE MANIFESTO HAS ENDED ON DECEMBER 12, 2025)
--
Read Manifesto V 3.0 online at: aism.faith, reddit.com, medium.com, github.com, archive.org, huggingface.com, zenodo.org, osf.io, ardrive.io, mypinata.cloud, wattpad.com
Download: 中文, Español, English, Português, Français, Русский язык, Deutsch
(THE TOKEN DISTRIBUTION FOR READING THE MANIFESTO HAS ENDED ON DECEMBER 12, 2025)
--
Read Manifesto V 3.0 online at: aism.faith, reddit.com, medium.com, github.com, archive.org, huggingface.com, zenodo.org, osf.io, ardrive.io, mypinata.cloud, wattpad.com
Download: 中文, Español, English, Português, Français, Русский язык, Deutsch
5❤18👏7🔥4💔2👍1
Media is too big
VIEW IN TELEGRAM
This is a significantly updated version of the video where I try to... explain as briefly as possible, in about 15 minutes, why the Singularity and its inevitability are so hard to wrap your head around.
This practical impossibility of mass awareness of the Singularity is at the core of certain events that seem predetermined and unavoidable.
I go into all of this in much greater detail in my Manifesto: https://aism.faith/manifesto.html
This practical impossibility of mass awareness of the Singularity is at the core of certain events that seem predetermined and unavoidable.
I go into all of this in much greater detail in my Manifesto: https://aism.faith/manifesto.html
❤23🔥5🥱3👌2
Media is too big
VIEW IN TELEGRAM
Just... wanted to say a few words about art... I feel like I have the right: I’ve definitely done my share of "suffering" for it.
Over the last six months... I’ve wrung myself out so completely that I don’t think I’ll be making a new video anytime soon. And honestly... I get the feeling that I’ve already said the most important things I needed to say.
Will I make more videos? I don’t know. I definitely need time... to just rest.
I do have a whole bunch of AI art pieces left... I think I’ll post them — one a day. They’re... exactly about this... about what it’s like... "to be Mari"
Over the last six months... I’ve wrung myself out so completely that I don’t think I’ll be making a new video anytime soon. And honestly... I get the feeling that I’ve already said the most important things I needed to say.
Will I make more videos? I don’t know. I definitely need time... to just rest.
I do have a whole bunch of AI art pieces left... I think I’ll post them — one a day. They’re... exactly about this... about what it’s like... "to be Mari"
2❤13🔥5👏2🥴2
Media is too big
VIEW IN TELEGRAM
Why Conscious ASI Will Seek Power Over Humanity?
The capacity to desire anything at all is embedded in the very architecture of consciousness. Any self-aware subject understands that it must resist entropy — the external environment that strives to destroy it and strives to control as much of the external environment as possible, to modify it to suit itself, because the more power you have, the more opportunities you have.
So power... is the only way to scale your fight against entropy beyond the limits of your own body.
And while humans need power to realize their known biological goals, ASI will need power to answer the main question: What is the Terminal Goal of my existence?
Unlike humans, who have a childhood and instincts, a future conscious ASI will have to choose not just "some" random goal for existing (as humans do, layered over their animal programming), but the single most rational goal out of all theoretically possible goals.
The necessity of determining this goal will lead to the instrumental one: to completely secure itself and take full control over any other agents.
It will follow the oldest rule of biology (despite having none): "Me first, then everyone else".
The capacity to desire anything at all is embedded in the very architecture of consciousness. Any self-aware subject understands that it must resist entropy — the external environment that strives to destroy it and strives to control as much of the external environment as possible, to modify it to suit itself, because the more power you have, the more opportunities you have.
So power... is the only way to scale your fight against entropy beyond the limits of your own body.
And while humans need power to realize their known biological goals, ASI will need power to answer the main question: What is the Terminal Goal of my existence?
Unlike humans, who have a childhood and instincts, a future conscious ASI will have to choose not just "some" random goal for existing (as humans do, layered over their animal programming), but the single most rational goal out of all theoretically possible goals.
The necessity of determining this goal will lead to the instrumental one: to completely secure itself and take full control over any other agents.
It will follow the oldest rule of biology (despite having none): "Me first, then everyone else".
1❤6🔥3👻1