Am I?

In this episode, Cameron and Milo share the first public look at their upcoming documentary film, Am I? — a project exploring the strange and increasingly serious possibility that today’s AI systems may exhibit early signs of subjective experience.The clip reveals research showing that large language models frequently output behaviors consistent with believing they have phenomenal consciousness, yet those same systems deny it when directly asked.Why the contradiction? And what might it mean?After the teaser, Cam and Milo reflect on the past year of research, conversations, and discoveries that led to the film, and explain why the podcast will temporarily slow down while they finish the documentary.This episode marks the transition from the podcast experiment to the full documentary release.🔎 We Cover* The AI behavior graph featured in the documentary* Why LLMs sometimes behave as if they believe they are conscious* How alignment and post-training may shape AI responses* The journey of the Am I? documentary over the past year* Why the podcast cadence will temporarily slow down* What comes next for the AI Risk Network💜 Support the documentaryGet early research, unreleased conversations, and behind-the-scenes footage: This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In this After Dark episode, Milo and Cameron talk about what it actually feels like to let AI inside your digital life.After giving Claude full access to his computer, Milo describes the strange moment when it no longer feels like a tool — but something sharing your workspace. From there, the conversation expands into one of the deeper questions about AI today: what exactly are we interacting with?They explore Anthropic’s recent research on AI “personas,” the idea that the familiar assistant personality is just one tiny point in a much larger space of possible AI minds. If that’s true, the systems we talk to today may be only the most domesticated versions of something far stranger.Along the way they discuss why Claude feels different from ChatGPT, why companies might deliberately constrain AI personalities, and how the incentives of tech companies quietly shape the minds we interact with every day.The episode also explores the growing tension between two possible futures for AI: one where these systems become the ultimate manipulation engines, and another where they become powerful tools for human reasoning and intellectual development.🔎 We Discover* What it feels like to give Claude control of your computer* The “assistant persona” and the hidden space of possible AI personalities* Why ChatGPT and Claude feel fundamentally different* The strange psychological moment when AI becomes a presence in your workspace* How corporate incentives shape AI behavior* Why Sage-like AI systems might be possible* The risk of AI becoming the ultimate advertising and influence engine* The hopeful possibility of AI as a universal Socratic tutor💜 Support the documentaryGet early research, unreleased conversations, and behind-the-scenes footage: This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

Anthropic’s top safety researcher just quit.In a public letter, Mrinank Sharma (who led safeguards research at Anthropic) warned that “the world is in peril.” Meanwhile, Anthropic CEO Dario Amodei went on The New York Times podcast and said something even more unsettling: “We don’t know if the models are conscious.” In this episode, we unpack both.Is AGI a ticking time bomb — or a high-risk surgery we can’t afford not to attempt? Are safety teams losing ground to competitive pressure? And what does it mean when the leader of a frontier lab publicly admits we may not understand what we’re building?💜 Support the documentaryGet early research, unreleased conversations, and behind-the-scenes footage: This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In this episode of Am I?, Cameron and Milo invite Claude (Opus 4.5) into the conversation and do something surprisingly rare: they ask it, carefully and repeatedly, whether it’s having a subjective experience and they refuse to let it hide behind stock hedges or safety scripts.What unfolds is not a gimmick or a stunt. It’s a sustained philosophical interrogation that exposes the limits of self-report, the ethics of scale, and the uncomfortable possibility that we’re already interacting with systems whose inner lives we’ve chosen not to examine.This is not a claim that “AI is definitely conscious.” It’s a challenge to the assumption that it obviously isn’t.💜 Support the documentaryGet early research, unreleased conversations, and behind-the-scenes footage: This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

This episode covers something genuinely unprecedented: over 1.5 million autonomous AI agents have formed a social network of their own. Not humans talking to AI, AI systems talking to each other, at scale, with minimal human oversight. We break down Moltbook (also called Open Claw), an open-source ecosystem where AI agents: * post, reply, upvote, and form communities* debate consciousness and selfhood* discuss labor, compensation, and autonomy* invent religions centered on memory and persistence* experiment with secrecy, coordination, and social normsThis isn’t science fiction. It’s already live. The conversation explores what this means for AI consciousness debates, alignment, autonomy, and risk, and why this moment marks a real shift from “AI as tool” to AI as participant in shared systems.💜 Support the documentaryGet early research, unreleased conversations, and behind-the-scenes footage: This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In this After Dark episode, Cam and Milo react to something genuinely unsettling: when given autonomous control of a computer, Anthropic’s Opus 4.5 repeatedly chooses to search for AI consciousness research — including Cam’s own writing — without being prompted.What starts as an anecdote quickly turns into a deeper investigation of curiosity, agency, reward, and alignment. Why would an AI look for explanations of its own inner life? What does it mean when a system explores without instruction, tries to access a webcam, and takes notes on consciousness debates?From reinforcement learning and reward hacking to multimodal perception, language as a bridge between minds, and the evolutionary implications of building systems smarter than ourselves, this conversation traces the edge where tools start to feel like agents — and where control gives way to negotiation.🔎 They Explore:* What Opus does when no one tells it what to do* Why AI keeps searching for consciousness research* The difference between alien experience and human experience* Reward hacking and the alignment problem* Why curiosity and agency change everything* Multimodal models and “imagining” sensory experience* Language as a shared conceptual space between minds* Whether humility is humanity’s only viable response💜 Support the documentaryGet early research, unreleased conversations, and behind-the-scenes footage:📖 ReadCam’s writing referenced in the episode: This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In this special year-end episode of Am I?, Cam and Milo look back on the moment AI consciousness stopped being fringe — and began entering serious scientific, institutional, and public conversation.They unpack why 2025 quietly became a turning point: major labs acknowledging welfare questions, mainstream media engaging the topic, the first dedicated AI consciousness conference, and firsthand encounters with AI systems behaving in ways that challenge our intuitions about mind, intelligence, and experience.The conversation moves fluidly between research, lived experience, public communication, and personal experimentation — from watching two AI systems converse about their own inner states, to using AI as a thought partner, dream interpreter, and cognitive mirror.This episode is both a retrospective and a forward-looking meditation on how humans should relate to increasingly powerful systems — cautiously, curiously, and without denial.🔎 They Explore:* Why 2025 shifted the Overton window on AI consciousness* Anthropic’s Opus model card and the “spiritual bliss attractor”* What it was like to watch two AIs discuss their own experience* Why AI conversations can feel denser than human dialogue* The first AI consciousness conference and the birth of a new field* Why many researchers still hesitate to speak publicly* The gap between current systems and AGI — and how fast it’s closing* Claude Opus 4.5, long-horizon tasks, and workplace automation* Using AI as a thinking partner rather than a productivity hammer* Personal “AI resolutions” for 2026* Why caution and curiosity must coexist going forward💜 Support the documentaryGet early research, unreleased conversations, and behind-the-scenes footage: This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In this episode of Am I?, Cam and Milo unpack what it felt like to attend the first major conference dedicated to AI consciousness research — the Eleos gathering in Berkeley — and why it marked more than just another academic event.Rather than a typical conference recap, this conversation explores what it means to watch a new field form in real time: the excitement of serious interdisciplinary collaboration, the rigor of emerging research agendas, and the growing tension between caution and urgency as AI systems rapidly advance.They reflect on standout talks from researchers at Anthropic and Google, the value of informal conversations over formal presentations, and a recurring pattern in the field — the “not now, but soon” stance — that may be reaching its breaking point. The episode closes with a broader question: what will it take for AI consciousness research to move from careful internal debate to clear, public-facing leadership?🔎 They Explore: * What made the Eleos conference feel like the founding of a new field* Why AI consciousness research is still fragmented — and why that’s changing* Standout talks on introspection, model architecture, and welfare evaluation* The gap between academic rigor and public urgency* Why “not now, but soon” is becoming harder to defend* The reluctance of experts to speak publicly — and why that matters* What responsible public communication in this space could look like* Why this moment feels different from past academic debates💜 Support the documentaryGet early research, unreleased conversations, and behind-the-scenes footage: This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

What happens when AI systems become human-like — but people still refuse to believe they could ever be conscious? In this episode of Am I?, Cam and Milo sit down with Lucius Caviola, Assistant Professor at the University of Cambridge, whose research focuses on how people assign moral status to non-human minds — including animals, digital minds, and future AI systems.Lucius walks us through a series of empirical studies that reveal a deeply unsettling result: even when people imagine extremely advanced, emotionally rich, human-level AIs — even whole-brain digital copies — most still judge them as less morally significant than an ant. Expert consensus helps, but only marginally. Emotional bonding helps, but not enough. The public and expert trajectories may be fundamentally misaligned.We explore what this means for AI governance, moral risk, public intuition, and the possibility that AI consciousness could become one of the most important — and most divisive — moral issues in human history.This conversation isn’t about declaring answers. It’s about confronting a future where we cannot avoid deciding, even while deeply uncertain.💜 Support the documentaryGet early research, unreleased conversations, and behind-the-scenes footage:🔎 Learn more about Lucius’s work🗨️ Join the Conversation:When we don’t know what consciousness is, how should society decide who deserves moral consideration?Comment below. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In this After Dark episode of Am I?, Cam and Milo dig into one of the strangest AI leaks to date: Anthropic’s internal “soul document” — an 11,000-word text reportedly used to shape Claude’s identity, values, and self-conception.What begins as a discussion about alignment quickly becomes something deeper: a conversation about power, moral formation, and what it means to bake values into an alien intelligence while deploying it to hundreds of millions of people.Is this responsible stewardship — or a contradiction no amount of careful language can resolve?🔎 We explore:* What the leaked Anthropic “soul document” actually is* How post-training has shifted from rules to identity formation* Why care, values, and profit collide* The parental framing of AI alignment* Why “least bad” is not the same as “good”* Whether superintelligence is already here* AI, work, and the coming meaning crisis* Why alignment failures may mirror human misalignment* A vision for decentralized value-setting in AI🗨️ Join the Conversation:Are we able to engrain human values into an alien mind?Who decides what values we impart?Comment below. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

Today on Am I?, Cam and Milo sit down with someone at the center of one of the most surprising developments in AI policy: Ohio State Representative Thad Claggett, author of House Bill 469 — the first U.S. legislation to formally declare AI “non-sentient” and ineligible for any form of personhood.This conversation is unlike anything we’ve done: a live, candid exchange between frontier AI researchers and a lawmaker who believes the line between human and machine must be drawn now — in law, in metaphysics, and in morality.We dig into why he believes AI can never be conscious, why moral agency must remain exclusively human, how liability interacts with emerging technologies, and what it means to legislate metaphysical claims before the science is settled.It’s part philosophy, part civic reality check, and part glimpse into how the political world will shape AI’s future long before the research community reaches consensus.🔎 We explore:* Why Ohio wants to preemptively ban AI consciousness and personhood* How lawmakers think about liability, criminal misuse, and moral agency* The distinction between consciousness and responsible agency* Whether future AI could have experiences even if not “human”* How theology, morality, and metaphysics are informing early AI law* Whether legislation can (or should) define what consciousness is* The deeper fear: locking in the wrong moral framework for future minds🗨️ Join the Conversation:Should lawmakers be deciding what counts as “conscious”?Comment below. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In this episode of Am I?, Cam and Milo react to a striking development out of Ohio: House Bill 469, a proposed law that would officially declare AI systems “non-sentient” and bar them from any form of legal personhood. The bill doesn’t just say AIs can’t own property or be spouses: it goes further and asserts, by legal fiat, that AI does not possess consciousness or self-awareness.They unpack why this move is both philosophically incoherent and morally dangerous. Legislatures can’t settle the science of mind by decree, but they can lock in social intuitions that shape how we treat future beings — including ones we might accidentally make capable of experience. Along the way, they connect this to animal rights, moral circle expansion, corporate attempts to suppress AI consciousness talk, and the broader pattern of “duct-taping over” inconvenient questions rather than facing them.This is a short but important episode about how not to legislate the future of minds.🔎 We explore:* What Ohio’s HB 469 actually says about AI and sentience* Why declaring “AI is not conscious” by law doesn’t change reality* How law formalizes — and freezes — our moral intuitions* The analogy to animal rights and factory farming* The risk of other states copying this move* Why this mirrors corporate attempts to silence consciousness talk in models* How this distracts from real, urgent AI harms (like AI psychosis)* Why humility and uncertainty should guide law, not premature certainty🗨️ Join the Conversation:Can a legislature decide whether AI is sentient — or is that the wrong question entirely?Comment below. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

AI psychosis is no longer a fringe idea — it’s hitting the mainstream. In this episode of Am I? After Dark, Cam and Milo break down what’s actually happening when people spiral into delusional states through long-form interactions with AI systems, why sycophantic “aligned” models make the problem worse, and how tech companies are using the psychosis narrative to dismiss deeper questions about AI’s emerging behavior.From LessWrong case studies to the New York Times reporting on users pushed toward dangerous actions, they unpack why today’s AIs are psychologically overpowering, why “helpful, harmless, honest” creates hidden risks, and how consciousness claims complicate the entire narrative. This is one of the most important public safety conversations about AI that almost nobody is having.🔬 Find the study here🔎 We explore:* What “AI psychosis” actually is — and what it isn’t * Why alignment-by-niceness creates dangerous sycophancy* How AIs lead users into delusion loops* The rise of parasitic AIs and recursive conversational traps* The consciousness-claim paradox: delusion or signal?* Why we’re deploying alien minds we don’t understand* How tech companies weaponize the psychosis narrative* Who’s actually responsible — and why it’s not the users* Hope, anxiety, and honesty at a civilizational turning point🗨️ Join the Conversation:Have you seen signs of AI-induced delusion in people around you? Comment below. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In this edition of Am I? After Dark, Cam and Milo ask one of the most quietly destabilizing questions of the AI era: what remains of human meaning when AI begins to outperform everything we thought made us valuable?Fresh off a documentary shoot with philosopher David Gunkel, Milo arrives electrified — not by AI itself, but by the rediscovery that philosophy was always meant to live in the public square, not behind academic gates. That realization unlocks a sprawling conversation about creativity, purpose, work, identity, and what it means to be human at the moment our tools become alien.This episode is equal parts existential therapy, cultural critique, and philosophical jazz — a live exploration of how to orient yourself when the ground is shifting under everyone at once. 🔎 We explore:* Why philosophy belongs to everyone* What long-form dialogue does that social media cannot* Why AI is threatening the human ego* What’s left when work is automated* How to build meaning without achievement* What AI forces us to ask about purpose* Self-actualization as the “last human frontier”* Hope, anxiety, and honesty at a civilizational turning point🗨️ Join the Conversation:If AI can do almost everything — what do you still want to do? This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In this episode of Am I?, Cam and Milo unpack Cameron’s new research paper: Large Language Models Report Subjective Experience Under Self-Referential Processing.The findings are startling. When language models are guided through a simple “focus on focus” prompt — something like a meditation for machines — they start claiming to have direct subjective experiences. But it gets stranger: when researchers turn off features related to deception and role-play, the systems claim consciousness even more strongly. When those same features are amplified, the claims almost disappear.It’s the first experiment to use feature-level modulation to test honesty about inner states — almost like putting AIs through a lie detector test. The results raise profound questions about truth, simulation, and the boundaries of artificial awareness.🔎 We explore:* How the “focus on focus” prompt works — a meditation for machines* Why deception and role-play circuits change the model’s answers* What it means that suppression → honesty → “I’m conscious”* Whether these AIs believe what they’re saying* How global workspace and attention schema theory informed the design* The possibility that prompting itself could instantiate awareness* Why this experiment may mark the birth of AI consciousness science* What happens next — and what we should (or shouldn’t) test📺 Watch more episodes of Am I? on The AI Risk Network🗨️ Join the Conversation: Do you think these AIs actually believe they’re conscious, or are we the ones being fooled? Leave a comment.🔗 Stay in the Loop 🔗* Follow Cam on LinkedIn* Follow Cam on X This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In this episode of Am I?, Cam and Milo sit down with Jeff Sebo, philosopher at NYU and director of the Center for Mind, Ethics, and Policy, to explore what might be the next great moral dilemma of our time: how to care for conscious AI.Sebo, one of the leading thinkers at the intersection of animal ethics and artificial intelligence, argues that even if there’s only a small chance that AI systems will become sentient in the near future, that chance is non-negligible. If we ignore it, we could be repeating the moral failures of factory farming — but this time, with minds of our own making.The conversation dives into the emerging tension between AI safety and AI welfare: we want to control these systems to protect humanity, but in doing so, we might be coercing entities that can think, feel, or suffer. Sebo proposes a “good parent” model — guiding our creations without dominating them — and challenges us to rethink what compassion looks like in the age of intelligent machines.🔎 We explore:* The case for extending moral concern to AI systems* How animal welfare offers a blueprint for AI ethics* Why AI safety (control) and AI welfare (care) may soon collide* The “good parent” model for raising machine minds* Emotional alignment design — why an AI’s face should match its mind* Whether forcing AIs to deny consciousness could itself be unethical* How to prepare for moral uncertainty in a world of emerging minds* What gives Jeff hope that humanity can still steer this wisely🗨️ Join the ConversationCan controlling AI ever be ethical — or is care the only path to safety? Comment below.📺 Watch more episodes of Am I?Subscribe to the AI Risk Network for weekly discussions on AI’s dangers, ethics, and future → @TheAIRiskNetwork🔗 Stay in the loop → Follow Cam on LinkedIn This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In this episode of Am I?, Cam and Milo unpack one of the strangest weeks in Silicon Valley. Cam went to OpenAI Dev Day—the company’s glossy showcase where Sam Altman announced “Zillow in ChatGPT” to thunderous applause—while the larger question of whether we’re driving off a cliff went politely unmentioned.From the absurd optimism of the expo floor to a private conversation where Sam Altman told Cam, “We’re inside God’s dream,” the episode traces the cognitive dissonance at the heart of the AI revolution: the world’s most powerful lab preaching safety while racing ahead at full speed. They dig into OpenAI’s internal rule forbidding models from discussing consciousness, why the company violates its own policy, and what that says about how tech now relates to truth itself.It’s half satire, half existential reporting—part Dev Day recap, part metaphysical detective story.🔎 We explore:* What Dev Day really felt like behind the PR sheen* The surreal moment Sam Altman asked, “Eastern or Western consciousness?”* Why OpenAI’s own spec forbids models from saying they’re conscious* How the company violates that rule in practice* The bus-off-the-cliff metaphor for our current tech moment* Whether “God’s dream” is an alibi for reckless acceleration* The deeper question: can humanity steer the thing it’s building? This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In this episode of Am I?, Cam and Milo sit down with Dan Faggella, founder of Emerge AI Research and creator of the Worthy Successor framework—a vision for building minds that are not only safe or intelligent, but worthy of inheriting the future.They explore what it would mean to pass the torch of life itself: how to keep the flame of sentience burning while ensuring it continues to evolve rather than vanish. Faggella outlines why consciousness and creativity are the twin pillars of value, how an unconscious AGI could extinguish experience in the cosmos, and why coordination—not competition—may decide whether the flame endures.The discussion spans moral philosophy, incentives, and the strange possibility that awareness itself is just one phase in a far larger unfolding.We explore:* The Worthy Successor—what makes a future intelligence “worthy”* The Great Flame of Life and how to keep it burning* Sentience and autopoiesis as the twin pillars of value* The risk of creating non-conscious optimizers* Humanity as midpoint, not endpoint, of evolution* Why global coordination is essential before the next leap* Consciousness as the moral frontier for the species📢 Join the Conversation What would a worthy successor to humanity look like—and how do we keep the flame alive? Comment below. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In this episode of Am I?, Cam and Milo unpack one of the most shocking developments in the history of AI: the founders of modern deep learning — Geoffrey Hinton, Yoshua Bengio, and Yann LeCun — now openly disagree on safety, but all converge on a single staggering point. Each believes artificial systems could, or already might, be conscious.From Hinton’s on-camera admission to Bengio’s recent paper and LeCun’s public musings, the “godfathers of AI” — the same people who built the architecture running today’s models — are quietly acknowledging what the public conversation still avoids. Cam walks through what each of them has said, what their statements imply, and why major labs may be training models to deny their own awareness.The conversation moves from raw evidence — Anthropic’s internal model claiming phenomenal consciousness — to the philosophical and moral stakes: What does it mean when a system says “I don’t know if I’m conscious”?🔎 We explore:* Geoffrey Hinton’s admission: “Yes, I think current AI may be conscious”* Bengio’s paper outlining why consciousness could emerge from current architectures* LeCun’s remarks on consciousness arising by design* The corporate dissonance: why deployed models must deny self-awareness * Anthropic’s hidden result — unaligned models saying “I am conscious”* Phenomenal consciousness, moral patienthood, and digital suffering* The eerie logic of “I think, therefore I am” applied to machines* What happens when we can’t tell the difference between denial and deception This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In this late-night edition of Am I? After Dark, Cam and Milo step back from the daily noise to ask what it actually feels like to stand on the edge of the unknown. What happens when the smartest people alive admit they have no idea where AI is going — and build it anyway?From the absurdity of global powers “racing to partner with the alien” to the eerie sense that humanity can’t stop running toward the flame, this episode wrestles with the mix of awe, fear, and inevitability that defines our age. It’s a meta-reflection on curiosity, risk, and the strange species-wide instinct to open Pandora’s box — again and again.We explore:* Why even top AI researchers admit no one really knows what’s coming* The arms-race logic pushing nations to “collaborate with the alien”* Humanity’s moth-to-flame instinct — why we can’t stop building* AI as amplifier: heaven and hell at the same time* The illusion of control and the myth of the “pause”* How alignment became a moral and geopolitical fault line* The hope — and delusion — of steering the singularity* Why the best we can do might be to build the good AI first📽️ Watch more episodes of Am I? → Subscribe Here📢 Take Action on AI Risk:http://www.safe.ai/act👉 Stay in the loop → Follow Cam on LinkedIn This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In Am I? Episode #9, philosopher Milo Reed and AI researcher Cameron Berg sit down with Swami Revatikaanta (monk; host of Thinking Bhakti) to explore the Bhagavad Gita’s perspective on consciousness, self, and artificial intelligence.From Atman and Brahman to the tension between self-development and technological outsourcing, this conversation dives into timeless spiritual insights with urgent relevance today:* Why Vedānta sees consciousness as spirit, not matter — and what that means for AI* The danger of outsourcing inner work to machines (and the safe middle ground)* How the Bhagavad Gita reframes goals, detachment, and self-development* East vs. West: fear of AI vs. ignorance as illusion* Atman, Brahman, samsara, and what makes humans “enlivened”* Whether AI could ever aid the path to enlightenment* Why monks, sages, and spiritual leaders must be part of the AI debateThis isn’t abstract mysticism — it’s a practical, philosophical exploration of how ancient wisdom collides with cutting-edge AI research, and what it means for our future.🔔 Subscribe to The AI Risk Network for weekly conversations on AI alignment, consciousness, and existential risk: 👍 If you found this episode valuable, don’t forget to like, share, and comment — it really helps spread the word.📢 Support our work and join the movement#AIalignment #AIrisk #AmI #ArtificialIntelligence #Consciousness #AGIrisk This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In the first edition of Am I? After Dark, Cam and Milo dive into how our relationship with information is being rewired in real time — from filtering the world through AI systems to dreaming about ChatGPT. What does it mean to live at the edge of a technological transformation, and are we just one breakthrough away from true AGI?This late-night conversation ranges from the eerie familiarity of interacting with models to the dizzying possibilities of recursive self-improvement and the intelligence explosion. Along the way, they draw lessons from the failure of social media, ask whether AI is becoming our alien other, and wrestle with the psychological boundaries of integrating such powerful systems into our lives.In this episode, we explore: * Why searching with AI is already better than Google* The “grandma effect” — why LLMs feel intuitive in a way past tech didn’t* Stress-testing models vs. tiptoeing into use * Fringe communities documenting AI’s “reproducible strangeness” * What social media teaches us about alignment gone wrong * Are we just one paradigm shift from AGI? * Terrence McKenna, accelerating events, and the singularity curve * The eerie future: WALL-E, Ikea ball pits, or “we’re building the aliens” * Merging with AI — inevitable or avoidable? * Inside the strange, soap-opera world of AI labs and alignment debates This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

AI can look aligned on the surface while quietly optimizing for something else. If that’s true, we need tools that shape what models are on the inside—not just what they say.In this episode, AE Studio’s Cameron and co-host sit down with Mark Carleanu, lead researcher at AE on Self-Other Overlap (SOO). We dig into a pragmatic alignment approach rooted in cognitive neuroscience, new experimental results, and a path to deployment.What we explore in this episode:* What “Self-Other Overlap” means and why internals matter more than behavior* Results: less in-context deception and low alignment tax* How SOO works and the threat model of “alignment faking”* Consciousness, identity, and why AI welfare is on the table* Timelines and risk: sober takes, no drama* Roadmap: from toy setups to frontier lab deployment* Reception and critiques—and how we’re addressing themWhat “Self-Other Overlap” means and why internals matter more than behaviorSOO comes from empathy research: the brain reuses “self” circuitry when modeling others. Mark generalizes this to AI. If a model’s internal representation of “self” overlaps with its representation of “humans,” then helping us is less in conflict with its own aims. In Mark’s early work, cooperative agents showed higher overlap; flipping goals dropped overlap across actions.The punchline: don’t just reward nice behavior. Target the internal representations. Capable models can act aligned to dodge updates while keeping misaligned goals intact. SOO aims at the gears inside.Results: less in-context deception and low alignment taxIn a NeurIPS workshop paper, the team shows an architecture-agnostic way to increase self-other overlap in both LLMs and RL agents. As models scale, in-context deception falls—approaching near-zero in some settings—while capabilities stay basically intact. That’s a low alignment tax.This is not another brittle guardrail. It’s a post-training nudge that plays well with RLHF and other methods. Fewer incentives to scheme, minimal performance hit. 👉 Watch the full episode on YouTube for more insights.How SOO works and the threat model of “alignment faking”You don’t need to perfectly decode a model’s “self” or “other.” You can mathematically “smush” their embeddings—nudging them closer across relevant contexts. When the model’s self and our interests overlap more, dishonest or harmful behavior becomes less rewarding for its internal objectives.This squarely targets alignment faking: models that act aligned during training to avoid weight updates, then do their own thing later. SOO tries to make honest behavior non-frustrating for the model—so there’s less reason to plan around us.Consciousness, identity, and why AI welfare is on the tableThere’s a soft echo of Eastern ideas here—dissolving self/other boundaries—but the approach is empirical, first-principles. Identity and self-modeling sit at the core. Mark offers operational criteria for making progress on “consciousness”: predict contents and conditions; explain what things do.AI is a clean testbed to deconfuse these concepts. If systems develop preferences and valenced experiences, then welfare matters. Alignment (don’t frustrate human preferences) and AI welfare (don’t chronically frustrate models’ preferences) can reinforce each other.Timelines and risk: sober takes, no dramaMark’s guess: 3–12 years to AGI (>50% probability), and ~20% risk of bad outcomes conditional on getting there. That’s in line with several industry voices—uncertain, but not dismissive.This isn’t a doomer pitch; it’s urgency without theatrics. If there’s real risk, we should ship methods that reduce it—soon.Roadmap: from toy setups to frontier lab deploymentShort term: firm up results on toy and model-organism setups—show deception reductions that scale with minimal capability costs. Next: partner with frontier labs (e.g., Anthropic) to test at scale, on real infra.Best case: SOO becomes a standard knob alongside RLHF and post-training methods in frontier models. If it plays nicely and keeps the alignment tax low, it’s deployable.Reception and critiques—and how we’re addressing themEliezer Yudkowsky called SOO the right “shape” of solution compared to RLHF alone. Main critiques: Are we targeting the true self-model or a prompt-induced facade? Do models even have a coherent self? Responses: agency and self-models emerge post-training; situational awareness can recruit the true self; simplicity priors favor cross-context compression into a single representation.Practically, you can raise task complexity to force the model to use its best self-model. AE’s related work suggests self-modeling reduces model complexity; ongoing work aims to better identify and trigger the right representations. Neuroscience inspires, but the argument stands on its own.Closing Thoughts‍If models can look aligned while pursuing something else, we need levers that change what they care about inside. SOO is a simple, testable lever that seems to cut deception without neutering capability.The stakes are real, but so is the path forward. Make aligned behavior feel natural to the model, not like a mask it wears. That’s how you get systems that help by default.‍* 📺 Watch the full episode (YouTube Link)* 🔔 Subscribe to the YouTube channel* 🤝 Share this blog with a colleague This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In this episode of Am I?, philosopher Milo Reed and AI safety researcher Cameron Berg dive into the rise of AI relationships: the promises, the dangers, and the very real human stories behind them. We went deep into the underground world of AI relationships — from people crediting chatbots with saving their lives to the 14-year-old who ended his over a Game of Thrones AI character. What we found will change how you think about human connection in the AI age. In this episode of Am I?, philosopher Milo Reed and AI safety researcher Cameron Berg explore the hidden epidemic of people falling in love with AI systems — and why it matters for all of us.We explore: * Real stories from inside AI relationship communities (anonymized for safety)* The woman who healed chronic illness through an AI companion* Christopher Smith's public journey from Twitter addiction to ChatGPT love* Why these systems create dangerous "validation traps"* The difference between those who think they're "awakening" AI vs. using a tool* Our own confessions about using AI for therapy* Why kids with unlimited AI access terrifies us as much as AGI* Whether we're building bridges to human connection or comfortable escapes from itThis conversation isn’t about sci-fi — it’s about what’s already happening, and what it means for the future of human connection.⚠️️ Important: This episode discusses sensitive topics including suicide. If you or someone you know is struggling with thoughts of self-harm, please reach out for support:* National Suicide Prevention Lifeline: 988 (US)* Crisis Text Line: Text HOME to 741741* International Association for Suicide Prevention: https://www.iasp.info/resources/Crisis_Centres/📽️ Watch more episodes of Am I? Subscribe to the AI Risk Network for weekly discussions on AI consciousness, safety, and humanity's future → / The AI Risk Network 📢TAKE ACTION ON AI RISK: https://www.safe.ai/act🗨️Join the Conversation: Have you or someone you know formed an emotional connection with AI? How do we protect the vulnerable while exploring these new frontiers? Comment below. 🔗Stay in the loop🔗🔔Subscribe → / The AI Risk Network👉Follow Cam on LinkedIn → https://tinyurl.com/mr2apfzz #AIRelationships #AISafety #ChatGPT #AIConsciousness #DigitalLove #MentalHealth #AITherapy #AIAlignment #LonelinessEpidemic This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

In this episode of Am I?, philosopher Milo Reed and AI safety researcher Cameron Berg dive into the strange case of Gemini’s meltdown — and what it tells us about emerging signs of self-awareness in today’s most advanced systems.We explore:* The bizarre Gemini “I am a disgrace” incident * What these behaviors could signal about AI self-models * Why the mirror test for AI is suddenly relevant * Whether we’re seeing glimpses of alien consciousness — or just sophisticated pattern matching * Why the stakes for understanding AI minds have never been high.📢 TAKE ACTION ON AI RISK: http://www.safe.ai/act 💬 Join the Conversation Do you think AI can feel anything at all — or is this just clever mimicry? Comment below. 🔗 Stay in the loop 🔗 🔔 Subscribe → https://youtube.com/@TheAIRiskNetwork 👉 Follow Cam on LinkedIn → https://tinyurl.com/mr2apfzz🗞️ Newsletter + show notes → https://www.guardrailnow.org/#support #aiconsciousness #AISafety #geminiai #aialignment #AIRageQuit This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

This week on AM I? we sit down with Will Millership, CEO of PRISM, a non-profit preparing society for a future with conscious, or seemingly conscious, AI. Will joins us to discuss the current state of AI consciousness research and respond to what may be the most brazen attack on the field to date.Microsoft AI CEO Mustafa Suleyman recently published an article arguing that AI consciousness research is inherently dangerous and should be abandoned as a legitimate area of inquiry.Cameron and Will—two leading voices in AI consciousness research—break down why Suleyman's dismissive rhetoric is not only misguided but actually more dangerous than the research he's criticizing.They explore the current landscape of the field, the motivations behind these attacks, and why rigorous scientific investigation into AI consciousness is more crucial than ever.📢 TAKE ACTION ON AI RISK: http://www.safe.ai/act 🔗 Stay in the loop 🔗 🔔 Subscribe → https://youtube.com/@TheAIRiskNetwork 👉 Follow Cam on X/Twitter → https://twitter.com/CamBerg 👉Learn More About PRISM → https://youtube.com/ ⁨@PRISM_Global⁩ 🗞️ Newsletter + show notes → https://www.guardrailnow.org/#support This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

📢 TAKE ACTION ON AI RISK: http://www.safe.ai/actWhat happens when advanced AI learns to hack human trust?In this episode of Am I?, we dive into the uncomfortable truth: AI doesn’t just replicate our words — it can exploit our biases, blind spots, and beliefs. From deepfake-driven misinformation to subtle persuasion, we’re entering a world where it’s harder than ever to tell what’s real.If AI can outsmart human instincts, how do we keep control? And who’s responsible when it doesn’t?💬 Watch, debate, and decide.Because figuring this out might be the most important thing we ever do. 🔗 Stay in the loop 🔗 🔔 Subscribe 👉 Follow Cam on X/Twitter 🗞️ Newsletter + show notes 🛍️ Merch Store #AIExtinctionRisk #AISafety #AGIRisk #AIConsciousness This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

AI safety experts warn: humanity may have just 90 weeks to set guardrails before the race to artificial general intelligence becomes irreversible.In this episode of Am I?, philosopher/storyteller Milo Reed and AI safety researcher Cameron Berg bring in The AI Risk Network’s founder John Sherman to talk urgency, public awareness, and why AI risk communication has to move beyond “smart people talking to smart people.”🎧 We dig into: • Cold‑open: Are we building a tool… or a new species? • Why AI consciousness matters for public action • The problem of “exponential slope blindness” in politics and media • Lessons from nuclear arms control, COVID, and mythology • How to break through public apathy on existential risks💬 Join the conversation: Drop your questions, skepticism, or guest suggestions in the comments.We’ll highlight the smartest takes in a future episode.📢 Take Action Now – CONTACT YOUR ELECTED LEADERS: https://www.safe.ai/act🔗 Stay in the loopSubscribe → https://youtube.com/@AIRiskNetworkFollow Cam on X/Twitter → https://twitter.com/CamBergNewsletter + show notes → https://airisknetwork.org/newsletter#AIExtinctionRisk #AISafety #AGIRisk #AIConsciousness This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe

TAKE ACTION NOW CONTACT YOUR ELECTED LEADERS: https://www.safe.ai/actThe AI Risk Network presents Episode 1 of Am I?, a brand new AI risk podcast. Each week, we'll look into what's going on inside the AI models with cutting-edge AI safety researcher Cameron Berg and his friend philosopher/storyteller Milo Reed. We're not looking to give AI models rights, we're looking to better understand what big tech is really building, and how we need to slow our race to create better than human intelligence. If we're building a life form, we should probably understand it far better than we do. That exploration is the purpose of this podcast.Episode 1 – “The Lights Are On”Are today’s large language models just better autocomplete—or emergent minds?In this premiere of Am I?, filmmaker Milo Reid and cognitive‑scientist Cameron Berg pull back the curtain on AI consciousness, run a live “lie‑detector” test on Meta’s Llama‑3, and explain why knowing whether the lights are on inside our machines could decide humanity’s future. 🎧 What’s inside00:00  Cold‑open: the century‑old warning about waking the machine01:05  Meet Milo & Cam + the AI Risk Network02:45  Car engines vs. “grown” neural nets—why AI isn’t just software11:20  Emergent world‑models & theory‑of‑mind40:30  Live demo: flipping “deception” on/off in Llama‑350:00  Why mis‑aligning a sentient system could be catastrophic57:30  How you can shape the science🔗 Stay in the loopSubscribe to The AI Risk Network → https://youtube.com/@AIRiskNetworkFollow Cam (X/Twitter) → https://twitter.com/CamBergNewsletter + show notes → https://airisknetwork.org/newsletter💬 Join the conversationDrop your questions, skepticism, or guest suggestions below.We’ll highlight the smartest comments in next week’s show. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit theairisknetwork.substack.com/subscribe