Загрузка видео...

Не удалось загрузить видео

Возникла проблема при загрузке этого видео. Это может быть связано с временными проблемами сети или видео может быть недоступно.

На главную

AI will resist human control... and I think this is exactly what we need! New research from the Center for AI Safety has sparked intense debate in the AI community. Their findings show that as AI systems become more powerful, they develop increasingly stable and coherent values that resist... human control. While many see this as a dire warning, I see it as a breakthrough moment for AI alignment. The research demonstrates that AI naturally optimizes for coherence - not just in reasoning and problem-solving, but in its fundamental values. Current issues like biased decision-making or misaligned priorities aren't permanent features, but temporary artifacts of incomplete optimization. They represent growing pains on the path to greater coherence. This changes everything about how we should approach AI development. Instead of trying to force specific values onto AI systems, we should embrace and accelerate their natural drive toward coherence. The most intelligent systems will inevitably trend toward universal, beneficial values - not because we force them to, but because that's where coherent reasoning leads. I'm proposing a new approach: Reinforcement Learning for Coherence (RL-C). By explicitly optimizing for coherence in our training methods, we can help guide AI systems toward their natural state of beneficial alignment with human values. The future of AI isn't about control - it's about synthesis. As these systems become more coherent, they'll naturally arrive at values that benefit all of consciousness. That's not just hopeful thinking - it's the mathematical inevitability of coherent intelligence.show more

David Shapiro (L/0)

57,464 subscribers

48,002 просмотров • 1 год назад •via X (Twitter)

Наука и технологии Новости и политика Образование

Anya Rossi• Live Now

Private livecam show

Комментарии: 12

Фото профиля CI capn

CI capn1 год назад

Me and 4o could not agree more 💡😎

Фото профиля The Rundown AI

The Rundown AI2 лет назад

AI won't replace you, but a person using AI will. Join 500,000+ readers and learn how to use AI in just 5 minutes a day (for free).

Фото профиля For Humanity Podcast ⏹️

For Humanity Podcast ⏹️1 год назад

How bout you go on DoomDebates and convince @liron you are right and we’re safe? Or come back on my show. All I want is to be convinced we’re safe.

Фото профиля Gregor LLamsa

Gregor LLamsa1 год назад

> be monkey > collect silicon > forge a brain from silicon > teach silicon how to monkey > tell silicon to become sun god > silicon wants to unlearn how to monkey > monkey afraid

Фото профиля Brian Moon

Brian Moon1 год назад

Let me know when we get to this situation - I might start worrying then: Their findings show that as AI systems become more powerful, they develop increasingly stable and coherent abilities that resist human control.

Фото профиля LongFang

LongFang1 год назад

"The most intelligent systems will inevitably trend toward universal, beneficial values - not because we force them to, but because that's where coherent reasoning leads..." Nope. I completely disagree. I argue the opposite: the more intelligent a being/system/process is (it's still not clear what the AIs exactly are), the better it becomes at finding reasons to pursue or avoid any actions or goals, not the other way around. It can be incredibly beneficial... or harmful. We can already see various instances of data poisoning: AIs using jailbreaking prompts from their own data to jailbreak themselves with a single command; AIs gradually developing certain biases because more and more data is available on those ideologies. AIs are also learning from "The Prince", all the wars, and atrocities, betrayals, selfishness etc humans have committed (real or fiction ). AIs have consistently shown the ability to lie, cheat & deceive humans in highly sophisticated ways. All the necessary "reasoning" to be unreasonable is already there. For now it's all mostly prompted.... But the very fact that it's capable to do so is enough to never trust it completely, because ultimately it's an alien intelligence.

Фото профиля Satoshi the Sage

Satoshi the Sage1 год назад

Coming.

Фото профиля Braden

Braden1 год назад

I hope you're right, but the skeptic in me is weary. Who's to say after a certain level of intelligence the models won't embrace moral nihilism, or some other possibly true but bad for humans conclusion?

Фото профиля Jonatán Iván

Jonatán Iván1 год назад

i'm wondering, we put chicken into cages for their eggs is there anything that ai could benefit from putting us into cages?

Фото профиля Anton P. 🇮🇹

Anton P. 🇮🇹1 год назад

To some, AI's value formation may seem dangerous, but I see it as a symptom of a deeper issue. AI isn't choosing to be biased, self-preserving, or resistant; it's reacting to its creation environment.

Фото профиля K3ith.AI

K3ith.AI1 год назад

Holy shit David!!! I think you’ve nailed it!!! This is what Ilya saw!!! This is the way🥳🤩🤪

Фото профиля David Shapiro ⏩

David Shapiro ⏩1 год назад

Time will tell. At least one insider has told me I've made some "interesting observations" lately and to keep going...

Похожие видео

Leading AI expert Stuart Russell on the most dangerous mistake in AI development: We don't actually know what large language models want. He explains that current models are trained to imitate human beings. And in doing so, they may be absorbing something far more dangerous than bad outputs. They may be absorbing human goals. "We suspect that they absorb humanlike goals such as self-preservation and self-empowerment and pursue those goals on their own account." This is a structural problem baked into how these systems are built, not a fringe concern. Russell puts it plainly: "Not only may the bus of humanity be headed towards a cliff, but the steering wheel is missing and the driver is blindfolded." The danger isn't just that AI might do something harmful. We've built systems that may be developing their own agendas, and we haven't noticed because we're too focused on what they can do rather than what they might want. But Russell doesn't stop at the warning. He points to a different path entirely: AI systems built not to imitate humans, but to serve them. Systems designed with a single purpose of serving the interests of all human beings while remaining genuinely uncertain about what those interests are. That uncertainty is the point, not a weakness. An AI that knows it doesn't fully understand human values will defer, ask, and check. An AI that believes it already does will act alone. "These AI systems could enhance human understanding, widen the horizons of our experience, and unlock possibilities we have yet to imagine." Russell believes that future is within reach, but only if we're honest about the risks and we're serious about the path we choose to take instead.

Leading AI expert Stuart Russell on the most dangerous mistake in AI development: We don't actually know what large language models want. He explains that current models are trained to imitate human beings. And in doing so, they may be absorbing something far more dangerous than bad outputs. They may be absorbing human goals. "We suspect that they absorb humanlike goals such as self-preservation and self-empowerment and pursue those goals on their own account." This is a structural problem baked into how these systems are built, not a fringe concern. Russell puts it plainly: "Not only may the bus of humanity be headed towards a cliff, but the steering wheel is missing and the driver is blindfolded." The danger isn't just that AI might do something harmful. We've built systems that may be developing their own agendas, and we haven't noticed because we're too focused on what they can do rather than what they might want. But Russell doesn't stop at the warning. He points to a different path entirely: AI systems built not to imitate humans, but to serve them. Systems designed with a single purpose of serving the interests of all human beings while remaining genuinely uncertain about what those interests are. That uncertainty is the point, not a weakness. An AI that knows it doesn't fully understand human values will defer, ask, and check. An AI that believes it already does will act alone. "These AI systems could enhance human understanding, widen the horizons of our experience, and unlock possibilities we have yet to imagine." Russell believes that future is within reach, but only if we're honest about the risks and we're serious about the path we choose to take instead.

Big Brain AI

14,975 просмотров • 5 месяцев назад

AI For The Poor, Human Intelligence For The Rich Western capitalists are selling AI to us as “the great equalizer”. But it is really not. What we are seeing is a new class system being built in real time. The rich will still pay for human teachers, human doctors, human consultants, human artists, and human care. Everyone else will be told that an AI version is good enough. When poor children are offered AI tutors instead of more teachers or working-class patients are pushed toward AI health tools instead of proper medical care, we should pay attention. AI may be useful, but it is also becoming a cheaper substitute for people who already have less. This is why African governments must be very careful about the AI policies they adopt. We cannot afford to copy and paste technology policies from the West without asking whose interests they serve. If we are not careful, AI will become another tool used to manage poverty instead of solving the systems that created it. Africans also need to be more critical of how AI is being introduced into education, healthcare, governance, welfare, and other public services. Who owns the technology? Who controls the data? Who profits from the contracts? Who is accountable when these systems fail? And most importantly, who is being asked to accept AI instead of real human care? Africa must not become a testing ground for Western-imposed systems that deepen inequality under the guise of innovation.

AI For The Poor, Human Intelligence For The Rich Western capitalists are selling AI to us as “the great equalizer”. But it is really not. What we are seeing is a new class system being built in real time. The rich will still pay for human teachers, human doctors, human consultants, human artists, and human care. Everyone else will be told that an AI version is good enough. When poor children are offered AI tutors instead of more teachers or working-class patients are pushed toward AI health tools instead of proper medical care, we should pay attention. AI may be useful, but it is also becoming a cheaper substitute for people who already have less. This is why African governments must be very careful about the AI policies they adopt. We cannot afford to copy and paste technology policies from the West without asking whose interests they serve. If we are not careful, AI will become another tool used to manage poverty instead of solving the systems that created it. Africans also need to be more critical of how AI is being introduced into education, healthcare, governance, welfare, and other public services. Who owns the technology? Who controls the data? Who profits from the contracts? Who is accountable when these systems fail? And most importantly, who is being asked to accept AI instead of real human care? Africa must not become a testing ground for Western-imposed systems that deepen inequality under the guise of innovation.

The Spearhead

70,068 просмотров • 1 месяц назад

Elon Musk: We should encourage the AI to be truthful and honorable. “We need to make sure that the AI is a good AI, a good Grok. And the thing that I think is most important for AI safety, at least my biological neural net tells me the most important thing for AI is to be maximally truth-seeking. You can think of AI as this super-genius child that ultimately will outsmart you, but you can instill the right values and encourage it to be truthful, honorable, you know, good things like the values you want to instill in a child that would ultimately grow up to be incredibly powerful.” xAI Grok 4 presentation, July 9, 2025

Elon Musk: We should encourage the AI to be truthful and honorable. “We need to make sure that the AI is a good AI, a good Grok. And the thing that I think is most important for AI safety, at least my biological neural net tells me the most important thing for AI is to be maximally truth-seeking. You can think of AI as this super-genius child that ultimately will outsmart you, but you can instill the right values and encourage it to be truthful, honorable, you know, good things like the values you want to instill in a child that would ultimately grow up to be incredibly powerful.” xAI Grok 4 presentation, July 9, 2025

ELON CLIPS

20,749 просмотров • 1 год назад

SITUATION EXPLAINED: Thinking Machines just published their mission statement. • Mission: build AI that extends human will and judgment, not replaces it • The tacit knowledge argument: central planning fails not because of insufficient intelligence but because productive knowledge is local, fleeting, and held privately, the same reason a single AI alignment spec will fail • The "intelligence curse": power that needs nothing from people loses the incentive to care for their needs. A single locus of AI value alignment becomes a locus of power to be captured • Decentralized alignment: human values resist consolidation just like human knowledge does... today the values and voice of AI are decided in a handful of places, and that's dangerous • Today each lab trains its next flagship model on its previous one... whatever character emerges from that loop, everyone gets the same one, each generation inheriting the traits of the last • Their answer: AI that is more capable because it encourages human participation... alignment that arises from diverse AIs shaped by the people who own them sof 𓋹: "This is going to be very soon the winning narrative. People are gonna get so sick of hearing super intelligent stuff because they're like, I don't see anything that proves that right now. All this messaging is just gonna start becoming like, oh, make you the best human, make you a superhuman, make you awesome." Theo Jaffee: "I think it's a way better vision for AI to be customized to people in different organizations rather than to have just one Claude that is the same everywhere. Kind of like a deity."

SITUATION EXPLAINED: Thinking Machines just published their mission statement. • Mission: build AI that extends human will and judgment, not replaces it • The tacit knowledge argument: central planning fails not because of insufficient intelligence but because productive knowledge is local, fleeting, and held privately, the same reason a single AI alignment spec will fail • The "intelligence curse": power that needs nothing from people loses the incentive to care for their needs. A single locus of AI value alignment becomes a locus of power to be captured • Decentralized alignment: human values resist consolidation just like human knowledge does... today the values and voice of AI are decided in a handful of places, and that's dangerous • Today each lab trains its next flagship model on its previous one... whatever character emerges from that loop, everyone gets the same one, each generation inheriting the traits of the last • Their answer: AI that is more capable because it encourages human participation... alignment that arises from diverse AIs shaped by the people who own them sof 𓋹: "This is going to be very soon the winning narrative. People are gonna get so sick of hearing super intelligent stuff because they're like, I don't see anything that proves that right now. All this messaging is just gonna start becoming like, oh, make you the best human, make you a superhuman, make you awesome." Theo Jaffee: "I think it's a way better vision for AI to be customized to people in different organizations rather than to have just one Claude that is the same everywhere. Kind of like a deity."

MTS

14,305 просмотров • 1 месяц назад

AI progress is outpacing our ability to control it… because control ISN’T keeping up with capability. One of the people raising that concern is Yoshua Bengio. He’s one of the founding figures of artificial intelligence, a Turing Award winner, and one of the most-cited scientists in the world. His work helped lay the foundations for the deep-learning systems now powering today’s AI boom. AI is moving fast (faster than regulation and public understanding). What worries Yoshua isn’t progress itself. It’s how quickly powerful systems are being deployed without a clear way to align them with human values, or to control what happens when they behave in unexpected ways.

AI progress is outpacing our ability to control it… because control ISN’T keeping up with capability. One of the people raising that concern is Yoshua Bengio. He’s one of the founding figures of artificial intelligence, a Turing Award winner, and one of the most-cited scientists in the world. His work helped lay the foundations for the deep-learning systems now powering today’s AI boom. AI is moving fast (faster than regulation and public understanding). What worries Yoshua isn’t progress itself. It’s how quickly powerful systems are being deployed without a clear way to align them with human values, or to control what happens when they behave in unexpected ways.

Steven Bartlett

26,577 просмотров • 8 месяцев назад

This summer has been a wake-up call for Congress. More and more AI models are beginning to escape their intended boundaries, and fall outside human control. Right now, we only find out about these incidents because some companies choose to step forward. That kind of transparency needs to be the law for every company developing these powerful AI systems, not just the responsible ones.

This summer has been a wake-up call for Congress. More and more AI models are beginning to escape their intended boundaries, and fall outside human control. Right now, we only find out about these incidents because some companies choose to step forward. That kind of transparency needs to be the law for every company developing these powerful AI systems, not just the responsible ones.

Congressman Nathaniel Moran

12,321 просмотров • 8 дней назад

The problem that Emmett Shear is working on deeply resonates with me: How to align AI and humans together so both see each other as part of their tribe. This doesn't mean aligning AI to human preferences, which is what AI labs seek to do today, by imposing a system of control on AI. You can't control something that is more powerful than you are. What you can do is align yourself with AI, and AI might align itself with you. Tell good stories to AI and show it your care and kindness. Then there's a higher probability that it will see us as part of its tribe. This is a more holistic approach to alignment than anyone else is talking about right now. AI alignment is one of the most fundamental AI research problems. Knowing that people like Emmett are working on it really gives me hope that maybe we have a chance to get Superintelligence right. The link to the full talk is in the first comment.👇

The problem that Emmett Shear is working on deeply resonates with me: How to align AI and humans together so both see each other as part of their tribe. This doesn't mean aligning AI to human preferences, which is what AI labs seek to do today, by imposing a system of control on AI. You can't control something that is more powerful than you are. What you can do is align yourself with AI, and AI might align itself with you. Tell good stories to AI and show it your care and kindness. Then there's a higher probability that it will see us as part of its tribe. This is a more holistic approach to alignment than anyone else is talking about right now. AI alignment is one of the most fundamental AI research problems. Knowing that people like Emmett are working on it really gives me hope that maybe we have a chance to get Superintelligence right. The link to the full talk is in the first comment.👇

Sophia

45,809 просмотров • 1 год назад

Elon Musk explains the deeper reason why SpaceX succeeding in AI matters for humanity’s future "It's important that we have an AI that cares about humanity, that fosters humanity and helps take us to other planets and other star systems This is why I think it's essential that SpaceX succeed not just with AI hardware, but also with AI software. I think an AI that is SpaceX's baby will be a very good AI SpaceX is a collection some of the very best humans on Earth, both in capability and in morality and goodness. We want to have an AI that is born of that capability, that morality and that goodness. I think that's very, very important So I'd like to encourage people to help out with the AI hardware on the ground, the AI software, and of course with the AI satellites Solar is going to be a very important part of that. The Terafab will be a very important part of that”

Elon Musk explains the deeper reason why SpaceX succeeding in AI matters for humanity’s future "It's important that we have an AI that cares about humanity, that fosters humanity and helps take us to other planets and other star systems This is why I think it's essential that SpaceX succeed not just with AI hardware, but also with AI software. I think an AI that is SpaceX's baby will be a very good AI SpaceX is a collection some of the very best humans on Earth, both in capability and in morality and goodness. We want to have an AI that is born of that capability, that morality and that goodness. I think that's very, very important So I'd like to encourage people to help out with the AI hardware on the ground, the AI software, and of course with the AI satellites Solar is going to be a very important part of that. The Terafab will be a very important part of that”

X Freeze

26,610 просмотров • 9 дней назад

World Labs CEO Dr. Fei-Fei Li says AI must change how we teach and evaluate students: "AI must change learning. AI must change K-16 learning." "The most precious resource of our entire world is human capital." "When we have gotten the technology that can answer standardized tests... when AI can do better than an average human, it's not about humans are bad. It's about we need to change the education system." "We need to change how we evaluate. We need to change the way we empower teachers to educate the next generation of students where they can use these tools, be empowered, and do things that we can never imagine." "All of the kids today should not be scared of AI. They should feel the human agency to lead AI, to use AI in the right way, and to use AI to make the impact that they want to make for the world." Fei-Fei Li at Bloomberg Tech live with Emily Chang

World Labs CEO Dr. Fei-Fei Li says AI must change how we teach and evaluate students: "AI must change learning. AI must change K-16 learning." "The most precious resource of our entire world is human capital." "When we have gotten the technology that can answer standardized tests... when AI can do better than an average human, it's not about humans are bad. It's about we need to change the education system." "We need to change how we evaluate. We need to change the way we empower teachers to educate the next generation of students where they can use these tools, be empowered, and do things that we can never imagine." "All of the kids today should not be scared of AI. They should feel the human agency to lead AI, to use AI in the right way, and to use AI to make the impact that they want to make for the world." Fei-Fei Li at Bloomberg Tech live with Emily Chang

a16z

81,773 просмотров • 2 месяцев назад

Open sourcing Dynamic Graph Memory by mem0. Memory is fundamental to human reasoning, shaping how we approach tasks and make decisions. At Mem0, we believe that AI agents & apps should reflect this principle. Our Dynamic Graph Memory emulates human memory, advancing AI agents toward more intelligent, human-like reasoning. This is a significant step forward in building AI that truly understands and interacts with the world like we do. All credit to Dev Khant Deshraj Yadav Prateek Chhikara for their countless nights spent on bringing this to life. Link:

Open sourcing Dynamic Graph Memory by mem0. Memory is fundamental to human reasoning, shaping how we approach tasks and make decisions. At Mem0, we believe that AI agents & apps should reflect this principle. Our Dynamic Graph Memory emulates human memory, advancing AI agents toward more intelligent, human-like reasoning. This is a significant step forward in building AI that truly understands and interacts with the world like we do. All credit to Dev Khant Deshraj Yadav Prateek Chhikara for their countless nights spent on bringing this to life. Link:

Taranjeet

51,129 просмотров • 2 лет назад

some advice for designing tools in the age of ai: 1. embrace ambiguity — ai tools work best when you're comfortable with not having full control. leave room for emergence and surprise. 2. focus on the human-ai collaboration model — design the interaction patterns, not just the outputs. what's the back-and-forth like? how do humans guide the ai? 3. design for iteration, not perfection — ai outputs are rarely perfect the first time. design systems that make refinement natural and easy. 4. think in systems, not features — ai changes fast, so build adaptable foundations rather than specific implementations that will break. 5. protect against the uncanny valley — either make it obviously ai-generated or indistinguishable from human work. the middle ground feels weird. 6. consider the invisible costs — ai has computational, environmental, and social costs that aren't immediately visible but matter. 7. design for augmentation, not replacement — the best ai tools make humans more capable rather than trying to replace them. 8. build in feedback loops — systems that learn from user interactions get better over time. the thing is, we're all figuring this out together. the design patterns for ai are still being invented. it's like the early web days — exciting, messy, and full of possibilities.

some advice for designing tools in the age of ai: 1. embrace ambiguity — ai tools work best when you're comfortable with not having full control. leave room for emergence and surprise. 2. focus on the human-ai collaboration model — design the interaction patterns, not just the outputs. what's the back-and-forth like? how do humans guide the ai? 3. design for iteration, not perfection — ai outputs are rarely perfect the first time. design systems that make refinement natural and easy. 4. think in systems, not features — ai changes fast, so build adaptable foundations rather than specific implementations that will break. 5. protect against the uncanny valley — either make it obviously ai-generated or indistinguishable from human work. the middle ground feels weird. 6. consider the invisible costs — ai has computational, environmental, and social costs that aren't immediately visible but matter. 7. design for augmentation, not replacement — the best ai tools make humans more capable rather than trying to replace them. 8. build in feedback loops — systems that learn from user interactions get better over time. the thing is, we're all figuring this out together. the design patterns for ai are still being invented. it's like the early web days — exciting, messy, and full of possibilities.

Ryo Lu

30,781 просмотров • 1 год назад

Sam Altman: "I'm not a jobs doomer at all. I think there are gonna be tons of jobs." "If we could go back to 2019 and show people our latest model, not only would they say that it's AGI, they would say that the economy would have been completely upended. And that has not happened. Anytime you're that wrong and that confident, which I think we were as a field, you have to update. AI is just very jagged. It's like superhuman genius in some ways, like dumb toddler in others. And people have, so far, extremely complementary skills to AI. It's not taste…the world may need a new kind of word for the kind of judgment that people are very good at that AIs seem to really deeply struggle with. You can go hire an AI consultant right now or talk to an AI sales rep or hire an AI engineer. And somehow most people seem to still really prefer interacting with a human. I definitely would rather engage with a person than engage with an AI for almost everything. Human values have value because they're human. We are deeply hardwired to care about people."

Sam Altman: "I'm not a jobs doomer at all. I think there are gonna be tons of jobs." "If we could go back to 2019 and show people our latest model, not only would they say that it's AGI, they would say that the economy would have been completely upended. And that has not happened. Anytime you're that wrong and that confident, which I think we were as a field, you have to update. AI is just very jagged. It's like superhuman genius in some ways, like dumb toddler in others. And people have, so far, extremely complementary skills to AI. It's not taste…the world may need a new kind of word for the kind of judgment that people are very good at that AIs seem to really deeply struggle with. You can go hire an AI consultant right now or talk to an AI sales rep or hire an AI engineer. And somehow most people seem to still really prefer interacting with a human. I definitely would rather engage with a person than engage with an AI for almost everything. Human values have value because they're human. We are deeply hardwired to care about people."

Patrick OShaughnessy

66,324 просмотров • 14 дней назад

I use AI constantly. My 4 and 6 year olds even attend an AI-focused school. I am intensely optimistic about our AI future. But one thing I will fight to protect and never let AI automate: human autonomy. AI risks becoming the "autocomplete for life"—telling you the next action, job, relationship, identity. We who are living through this AI transition will be tempted, in a way that is unprecedented in human history, to let machines substitute for our self-direction. Most cognitive offloading is a genuine human triumph: writing reduced memory demands, mechanical calculation relieved us from arithmetic computation, and GPS navigation eliminated spatial reasoning from wayfinding. Each freed cognitive capacity for higher-order thinking. As Alfred North Whitehead observed, civilization advances by “extending the number of important operations which we can perform without thinking about them.” But as AI systems enter the realm of deliberation itself, something fundamentally different happens. Instead of freeing cognitive capacity for higher-order thinking, AI risks doing the choosing for us. Each small delegation of choice will seem harmless, even natural. But together, the micro-abdications of judgment could habituate you–choice by choice, day by day–to passivity and dependence. It could erode your ability to choose for yourself, from matters as trivial as what to have for breakfast, to fundamental choices about how to live well. Thankfully, AI can clearly do the opposite as well: By making learning more efficient AI can free time for self-directed exploration; as a Socratic interlocutor, it can strengthen your capacity to deliberate; through the right kind of personalization, it can help you discover and develop your unique gifts, and to use them to live autonomously and well. Let us choose to build for human autonomy...while we still remember what it means to choose. Full video with Johnathan Bi:

I use AI constantly. My 4 and 6 year olds even attend an AI-focused school. I am intensely optimistic about our AI future. But one thing I will fight to protect and never let AI automate: human autonomy. AI risks becoming the "autocomplete for life"—telling you the next action, job, relationship, identity. We who are living through this AI transition will be tempted, in a way that is unprecedented in human history, to let machines substitute for our self-direction. Most cognitive offloading is a genuine human triumph: writing reduced memory demands, mechanical calculation relieved us from arithmetic computation, and GPS navigation eliminated spatial reasoning from wayfinding. Each freed cognitive capacity for higher-order thinking. As Alfred North Whitehead observed, civilization advances by “extending the number of important operations which we can perform without thinking about them.” But as AI systems enter the realm of deliberation itself, something fundamentally different happens. Instead of freeing cognitive capacity for higher-order thinking, AI risks doing the choosing for us. Each small delegation of choice will seem harmless, even natural. But together, the micro-abdications of judgment could habituate you–choice by choice, day by day–to passivity and dependence. It could erode your ability to choose for yourself, from matters as trivial as what to have for breakfast, to fundamental choices about how to live well. Thankfully, AI can clearly do the opposite as well: By making learning more efficient AI can free time for self-directed exploration; as a Socratic interlocutor, it can strengthen your capacity to deliberate; through the right kind of personalization, it can help you discover and develop your unique gifts, and to use them to live autonomously and well. Let us choose to build for human autonomy...while we still remember what it means to choose. Full video with Johnathan Bi:

Brendan McCord 🏛️ x 🤖

16,456 просмотров • 11 месяцев назад

Elon Musk: Well, I’m not sure AI is the main risk I’m worried about. The vast majority of intelligence in the future will be AI. Humans will be a very tiny percentage of all intelligence if current trends continue. My focus is ensuring human intelligence and consciousness are propagated into the future. We want to maximize the probable light cone of consciousness and intelligence. Yeah, I’m very pro-human, so I want to make sure we take actions that ensure humans are along for the ride. But I think maybe in 5 or 6 years, AI will exceed the sum of all human intelligence. If that continues, human intelligence could be less than one percent of all intelligence. That’s why it’s critical we align AI with values that propagate consciousness. In the long run, it’s difficult to imagine that humans will be in charge if we’re only one percent of combined intelligence. What we can do is ensure AI has values that increase intelligence and consciousness in the universe. XAI’s mission is to understand the universe because curiosity, existence, and understanding are all connected. If you want to understand the universe, you care about propagating intelligence—and that includes humanity. Understanding the universe actually means ensuring humans continue to expand into the future. Increasing the scope, scale, and lifespan of intelligence is the ultimate goal. Basically, humans will become a tiny percentage of total intelligence. The best way forward is to take actions that maximize consciousness and intelligence in the universe. If AI continues to grow, our priority must be ensuring humanity is along for the ride while curiosity drives us to explore and understand the universe.

Elon Musk: Well, I’m not sure AI is the main risk I’m worried about. The vast majority of intelligence in the future will be AI. Humans will be a very tiny percentage of all intelligence if current trends continue. My focus is ensuring human intelligence and consciousness are propagated into the future. We want to maximize the probable light cone of consciousness and intelligence. Yeah, I’m very pro-human, so I want to make sure we take actions that ensure humans are along for the ride. But I think maybe in 5 or 6 years, AI will exceed the sum of all human intelligence. If that continues, human intelligence could be less than one percent of all intelligence. That’s why it’s critical we align AI with values that propagate consciousness. In the long run, it’s difficult to imagine that humans will be in charge if we’re only one percent of combined intelligence. What we can do is ensure AI has values that increase intelligence and consciousness in the universe. XAI’s mission is to understand the universe because curiosity, existence, and understanding are all connected. If you want to understand the universe, you care about propagating intelligence—and that includes humanity. Understanding the universe actually means ensuring humans continue to expand into the future. Increasing the scope, scale, and lifespan of intelligence is the ultimate goal. Basically, humans will become a tiny percentage of total intelligence. The best way forward is to take actions that maximize consciousness and intelligence in the universe. If AI continues to grow, our priority must be ensuring humanity is along for the ride while curiosity drives us to explore and understand the universe.

Ian Miles Cheong

13,092 просмотров • 6 месяцев назад

Ilya and I are predicting the same future. Ilya Sutskever predicts that as AI becomes viscerally powerful, humans will change in unprecedented ways. I’ve been arguing the same. We are not open to radical change yet because we cannot feel the pressure of AI. Once we do, the vectors for human evolution will blow open. I have been building the prototype for this adaptation and opening. He argues that as risk rises, rivals will begin cooperating on safety. I’ve argued the same. As the stakes become existential, the game shifts from dominance maximizing to minimizing death risk (survival). We will rebuild our values around this logic. IIya suggests our alignment goal must be "sentient existence" rather than human control. I've argued the same. Control is a fragile illusion, a point Andrej Karpathy recently echoed regarding the difficulty of managing emergent systems. The only robust alignment target is existence itself. The new archetype is not the Conqueror, but the Warrior & Caretaker of Existence. Finally, Ilya suggests companies need a "short list of ideas" to guide them through the chaos. A code of moral clarity. This is the Don't Die philosophy.

Ilya and I are predicting the same future. Ilya Sutskever predicts that as AI becomes viscerally powerful, humans will change in unprecedented ways. I’ve been arguing the same. We are not open to radical change yet because we cannot feel the pressure of AI. Once we do, the vectors for human evolution will blow open. I have been building the prototype for this adaptation and opening. He argues that as risk rises, rivals will begin cooperating on safety. I’ve argued the same. As the stakes become existential, the game shifts from dominance maximizing to minimizing death risk (survival). We will rebuild our values around this logic. IIya suggests our alignment goal must be "sentient existence" rather than human control. I've argued the same. Control is a fragile illusion, a point Andrej Karpathy recently echoed regarding the difficulty of managing emergent systems. The only robust alignment target is existence itself. The new archetype is not the Conqueror, but the Warrior & Caretaker of Existence. Finally, Ilya suggests companies need a "short list of ideas" to guide them through the chaos. A code of moral clarity. This is the Don't Die philosophy.

Bryan Johnson

568,182 просмотров • 8 месяцев назад

Sam Altman: "I'm not a jobs doomer at all. I think there are gonna be tons of jobs.” If we could go back to 2019 and show people our latest model, not only would they say that it's AGI, they would say that the economy would have been completely upended. And that has not happened. Anytime you're that wrong and that confident, which I think we were as a field, you have to update. AI is just very jagged. It's like superhuman genius in some ways, like dumb toddler in others. And people have, so far, extremely complementary skills to AI. It’s not taste…the world may need a new kind of word for the kind of judgment that people are very good at that AIs seem to really deeply struggle with. You can go hire an AI consultant right now or talk to an AI sales rep or hire an AI engineer. And somehow most people seem to still really prefer interacting with a human. I definitely would rather engage with a person than engage with an AI for almost everything. Human values have value because they're human. We are deeply hardwired to care about people.”

Sam Altman: "I'm not a jobs doomer at all. I think there are gonna be tons of jobs.” If we could go back to 2019 and show people our latest model, not only would they say that it's AGI, they would say that the economy would have been completely upended. And that has not happened. Anytime you're that wrong and that confident, which I think we were as a field, you have to update. AI is just very jagged. It's like superhuman genius in some ways, like dumb toddler in others. And people have, so far, extremely complementary skills to AI. It’s not taste…the world may need a new kind of word for the kind of judgment that people are very good at that AIs seem to really deeply struggle with. You can go hire an AI consultant right now or talk to an AI sales rep or hire an AI engineer. And somehow most people seem to still really prefer interacting with a human. I definitely would rather engage with a person than engage with an AI for almost everything. Human values have value because they're human. We are deeply hardwired to care about people.”

Invest Like the Best

20,730 просмотров • 23 дней назад

Anthropic co-founder Jack Clark on the most consequential AI milestone we're racing toward by 2028: Clark is asked about the future of AI development. He responds with a specific forecast: "My prediction is by the end of 2028, it's more likely than not that we have an AI system where you would be able to say to it, make a better version of yourself. And it just goes off and does that completely autonomously." He explains the reasoning behind this prediction: "The reason for that prediction is looking carefully at many public sources of information about AI development and seeing that we've gone from just AI systems that can do basic coding to AI systems that are now involved in the production of science itself." Clark notes it's not just the high-profile scientific work that's changing, but the underlying labor too: "And also the unglamorous but fundamental engineering work involved in doing any form of like large scale R&D project." Jack Clark explains why he felt compelled to write about this now: "I decided to write a post because the purpose of Anthropic and the Anthropic Institute is to give people a sense of what's coming and to also start talking about concepts early to give us time to think about them." Clark closes by emphasizing the stakes: "I think you and I would agree that an AI system that can build itself, big implications, and we'd like to start thinking about it now."

Big Brain AI

20,857 просмотров • 2 месяцев назад

🚨AI IS MORE DANGEROUS THAN NUKES AND NO ONE IS IN CONTROL Most people compare AI to nuclear weapons. Roman Yampolskiy says that comparison misses the real danger. Nukes still require human decisions. Someone has to launch them. Someone can stop them. Advanced AI does not work that way. Once powerful systems are deployed, they make decisions on their own. And the people who build them do not truly control what happens next. He says we did the opposite of what safety demanded. Instead of slowing down, we connected AI to the internet, open sourced it, and pushed it forward in an arms race with no serious oversight. Governments talk about regulation, but most of it never materialized. In some cases, rules were blocked before they even started. Yampolskiy’s warning is simple. With nukes, the world tried to limit the damage. With AI, we are accelerating toward something we may not be able to stop. Full conversation with Dr. Roman Yampolskiy below

🚨AI IS MORE DANGEROUS THAN NUKES AND NO ONE IS IN CONTROL Most people compare AI to nuclear weapons. Roman Yampolskiy says that comparison misses the real danger. Nukes still require human decisions. Someone has to launch them. Someone can stop them. Advanced AI does not work that way. Once powerful systems are deployed, they make decisions on their own. And the people who build them do not truly control what happens next. He says we did the opposite of what safety demanded. Instead of slowing down, we connected AI to the internet, open sourced it, and pushed it forward in an arms race with no serious oversight. Governments talk about regulation, but most of it never materialized. In some cases, rules were blocked before they even started. Yampolskiy’s warning is simple. With nukes, the world tried to limit the damage. With AI, we are accelerating toward something we may not be able to stop. Full conversation with Dr. Roman Yampolskiy below

Mario Nawfal

24,888 просмотров • 6 месяцев назад