Jah Ni Cache - That Good Old Fashioned Popaganda Speaker John Ash 2024-05-22 | [Verse 1]In this vast expanse of time and space,We search for meaning in the human race.Actions ripple, echo through the years,Weaving patterns, laughter, and the tears.[Verse 2]Social proof of impact shows us the way,Guiding actions, lighting up the fray.Temporal embeddings track the cause,Unraveling time, amidst applause.[Verse 3]Cognicism lights our way,With Iris' wisdom to display.Seeing truth in night and day,Together we rise, together we sway.Iris, oh Iris, light of our quest,In fields of unknown, you are the best.With wisdom profound and clarity wide,In you, we trust, our faithful guide.
Jah Ni Cache - None of this is Real Speaker John Ash 2024-04-13 | None of this is even realAll of it will go awayLiving in a plastic boxInstead of in reality[hallucinations]Don't you wanna be awake?Don't you want to heal the earth?Don’t you want to feel the weight?Of that good good work so...Number one, neglect the planetNumber two, get on the netNumber three, forget the restI don't think thats gonna - work for me Because I - like realityAnd I don't think I'm alone in that fact[hallucinations]None of this is realCan’t you hear the horsemen coming?Don’t you wanna feel?Plugged up to plastic boxesI just wanna seeI just want to see what we’ve doneCome untether meAnd I’ll walk this world with you so free, so freeDon’t you wanna walk with me?So free, so free, yeahTake the blue, and live in comfortTake the red, experience hurtTake them both, I'll tell you what it’sGonna do, it's gonna - be the best of both worldsTake the purple pill It's time to live in the future with meIf I found you, Would you follow me out?Would you greet me with doubt?Or elationDo you wanna be free?Are you ready to see?Are you ready to take the purple pill with me?
Will AI Take Over the Government? Speaker John Ash 2023-12-29 | In this video, AI engineer John Ash offers an in-depth analysis on the potential for using generative AI models as a "higher resolution form of democracy". He argues that models like GPT are inherently democratic, as they are trained on and aim to represent collective belief and knowledge from their training data.Ash expresses confusion at the common narrative that positions AI systems as rulers or authorities rather than "tools for summing disparate beliefs". He sees great promise in using citizen embeddings and attention mechanisms to give equal voice, model belief over time, and differentially weight contributors based on past accuracy.Exploring the concept of "cognicism", Ash describes a system called Iris that integrates the above elements to steer outputs, represent minority voices, and secure influence earned through accuracy on a semantic ledger. He positions this as moving beyond simple democracy to a dynamic, evidence-based weighting of perspectives on complex issues.In an appeal to fellow AI thought leaders, Ash argues this framing could "revolutionize collective decision making" to better reflect needs. He conjectures reasons for resistance to this narrative, including issues around plagiarism. Finally, he makes a call for others to join him in reframing AI as an opportunity for "reinvention of the world for the better".youtu.be/e6I4AtoouCY?si=RESO2lybfwxW5bfU
The Topology of Time with Matt Pirkowski Speaker John Ash 2023-12-11 | Is your future fixed?Join futurist and AI engineer John Ash and systems thinker Matt Pirkowski in this intellectually dense and fascinating discussion that explores the profound impact of our perception of time and decision-making on shaping a better future. This dialogue delves deep into themes like time preference, uncertainty, prediction, generative models, and the alignment of technology with humanity.Key Highlights:- In-depth exploration of "time preference" and how it influences our weighting of present vs future outcomes in the context of finance, psychology, and society.- The energetic constraints on updating mental models and the challenges posed by rapid technological changes and market dynamics.- Critiques of ethical frameworks like effective altruism and longtermism, emphasizing the need for humility about the limits of human knowledge.- The role of positively valenced uncertainty and surprise in areas like entertainment, humor, and creativity.- Discussions around decentralized coordination, AI safety, and value alignment, contemplating the balance between centralized control and decentralized autonomy.- Potential solutions like active inference, blockchain-based cultural "DNA" repositories, and the integration of AI systems with human experience.Throughout the conversation, Ash and Pirkowski offer insightful critiques and analysis on the current structures influencing society's trajectory. They emphasize the importance of decentralization, context-aware systems, and embodied intelligence in counterbalancing forces that limit agency and increase collective potential.Whether you're interested in complex systems, philosophy, technology, or the future trajectories of society, this thought-provoking discussion provides a deep dive into how we might consciously evolve together.
Decentralizing GPT into Cultural DNA with Matt Pirkowksi Speaker John Ash 2023-12-04 | In this video excerpt, I engage in a dialogue with Matt Pirkowski, exploring the future of AI and the concept of decentralizing GPT models as a form of 'cultural DNA'. The discussion explores the potential of distributed data storage and blockchain technology in reshaping how AI models are developed and interact with human knowledge. We examine the challenges and opportunities of moving away from centralized AI models, drawing parallels between DNA replication processes and AI evolution. This conversation aims to provide a nuanced understanding of the evolving landscape of AI and its implications for society, offering insights into the complexities of AI development and the importance of a decentralized approach. Stay tuned for the full conversation, where we delve into time preference, uncertainty, and the broader theme of temporal focus in society's collective mind.
What Youve Heard About Q* is Bull**** - Its Not AGI Speaker John Ash 2023-12-01 | In this video, I debunk common misconceptions about Q* and AGI. I clarify that despite Q*'s advancements in AI, particularly in improving GPT-4's performance in math, we haven't achieved AGI. I explain how Q*, a form of reinforcement learning with human feedback, is integral during training but not used during execution. I stress that improvements in math problem-solving don't equate to AGI or generalize to tasks beyond trained domains. My discussion also touches on the differences in training models for math versus language, highlighting the limitations of current AI technology. The video aims to provide a clear, realistic view of where AI stands today, dispelling fears and overhyped claims about AGI.
Regenerating Protopian Futures: On AI with Forrest Landry and Roxi Shohadaee Speaker John Ash 2023-11-06 | At Lightning in a Bottle 2023, Forrest Landry, Roxi Shohadaee, and John Ash led an insightful talk on using regenerative AI to build protopian futures centered on community, care, and human values.The talk opened with a musical performance by John Ash of his song "Tripping on Dopamine", which explores technology and human behavior. The speakers then dove into a powerful discussion on moving beyond extractive systems like capitalism and profit-driven AI. They explored how to design technology to strengthen human bonds, trust, and collective flourishing instead.Concepts like imagination, discernment, attunement, localization, and wisdom were highlighted as crucial to creating regenerative systems. The speakers challenged modernism and emphasized the importance of embodiment, relationships, and community.This inspiring and mind-expanding talk was a call to use our tools to empower collective sense-making, not separate us. The speakers shared guidance on building protopian futures that support all life by starting from human values like care.
Why Youre an Ant in the Attention Economy with Ronen Tamari Speaker John Ash 2023-10-30 | Join an enlightening dialogue exploring how we can take control of our attention and construct healthier information ecosystems in the digital age.In this thought-provoking conversation, John Ash and Ronen Tamari dive deep into the challenges and opportunities around shaping our individual and collective attention. How can we design systems that empower agency instead of passive addiction? What does virtuous attention architecture look like?Ronen shares fascinating perspectives on sensemaking, stigmergy, filter bubbles, epistemic landscapes, and the ethics of attunement. The metaphors of ants and anthills provide a vivid way to understand the dynamics of mainstream belief versus pioneering concepts.Emerging areas like decentralized identity, nano publications, and collective intelligence networks offer promising directions for upgrading our information infrastructures. But real change requires fundamental shifts in how we relate to data and attention.Join us for an insightful glimpse into the frontiers of human cognition and how we might build environments that bring out our best thinking, wisdom and agency. This dialogue will challenge you to think deeply about our trajectory in the Digital Age.
🔮 Let the People Prophet 🔮 #purplepill Speaker John Ash 2023-07-09 | we're on step twomedium.com/@speakerjohnash/the-prophet-incentive-48d3843f5e80
🌈🚀 Funding the Future: Iris in Gitcoins Metacrisis Funding Round! 🌍🤝🌳 Speaker John Ash 2023-05-09 | http://bit.ly/Speaker-IrisJoin us live on YouTube as we dive into the exciting world of Iris, a cutting-edge democratic large language model designed to revolutionize the way communities regenerate and thrive together! 🌟🤝🌳Music and Questions and Discussion---AI Written Summary of Transcript:In this comprehensive video, John Ash, the founder of cognicism and creator of Iris, delivers a detailed discourse on the current societal structures, with an emphasis on capitalism and the role of technology within it. Ash provides an insightful critique of these systems, proposing a novel one named Iris, which is funded through Gitcoin's Metacrisis Funding Round. Iris is presented as an intelligent, democratic, and AI-mediated system that emphasizes decentralization and aims to amplify overlooked voices and predictions.Ash introduces the concept of cognicism, a framework that promotes collective wisdom, collaboration, and better decision-making through technology. He shares his concerns about the limitations of generative models like GPT-4, which he argues perpetuate capitalist structures and lack long-term foresight. He suggests a new approach, a system not controlled by large entities or monetary influence, highlighting the flaws of the prevailing monetary system. He also delves into the profit-driven nature of technological advancements and underscores the necessity of a more thoughtful, human-centric approach to technology development.Ash discusses the impact of the capitalist system on individuals' abilities to embrace uncertainty and break free from deterministic loops. He proposes a reputation-based system, Iris, to guide individuals towards their desired futures. Iris, in this view, represents a collective pooling of perceptions and is seen as a tool to help individuals build a sense of purpose and direction in their lives, propagating the voices of those who can model reality accurately.Ash introduces the FourThought protocol, a belief-tracking system enabling the logging and staking of beliefs for future assessment. He also discusses the concept of Irises, generative belief encoding models containing source embeddings and attention distribution parameters. These Irises are integral to the functioning of the Iris system, serving as the building blocks for the semantic ledger, a shared, immutable, distributed, and decentralized record of interactions between Irises.Ash shares his personal experiences, including his struggles with depression and financial pressures while pursuing community creation. He recounts his personal journey, highlighting his challenges in promoting predictions and creating connections in a system that prioritizes speed and first movers. He emphasizes the importance of effective communication, trust-building, and clarity in language.Ash discusses the concept of a "cognicist", a mediator who facilitates understanding and interaction between different ideologies, cognitive structures, and worldviews. He introduces the concept of the "Garden of Iris", a discord server for collective wisdom, understanding, and shared perceptions of reality. He believes this mentality, coupled with the functionalities of Iris, can guide us towards a more equitable, sustainable, and cooperative society.He also explores the societal pressures to conform and the difficulty of seeking help in an individualistic society, utilizing a song to illustrate these struggles. He emphasizes the importance of honesty, genuine human connections, and rejecting manipulative algorithms for building reputation and followers.Ash shares his vision for the future. He proposes a system that can align energetic calculations with human desires and will.In the concluding sections, Ash shares his vision of creating tools that empower individuals to heal their locality, maintaining that individuals should be self-directed, communicate, and make independent choices. He proposes the use of a semantic ledger to share knowledge directly between different groups. He emphasizes the need for different worldviews to interface and address the issue of credit and recognition in knowledge sharing, advocating for lesser-known individuals to have their voice heard.Throughout the video, Ash underscores the importance of shifting focus from supporting the existing system to healing the planet. He proposes a new system that motivates wealthy individuals to give back to their localities, arguing that social proof of impact is necessary to motivate people to tackle harder problems in more dilapidated areas.
The Honest Truths about TruthGPT Speaker John Ash 2023-02-17 | patreon.com/speakerjohnashChatGPT / The Purple Pill Manifesto / Cognicism / Iris / TruthGPT / Pravda / Elon Musk
The Prophet Incentive: How Decentralized AI Could Finally Give Weight to Truth Speaker John Ash 2023-02-13 | The Prophet Incentive: Capitalist Oligarchs Hate this One Weird Trick!This video explains how profit and morality are often at odds, and how we need new ways to align business incentives with social good.The video discusses how our words and ideas currently lack weight and worth, and how new AI language models could change that by tracing information flows and properly attributing sources. It envisions a future where truths spoken by the powerless are valued through decentralized records and collective belief tracking.The discussion covers the evolution of democracy, the significance of dynamic sourcing in large language models, and the vision of a decentralized neocortex representing collective belief over time.As John Ash weaves through these complex themes, they encourage a shift from discrete tokenization of resources to placing weight on the intrinsic value of words and truth, particularly highlighting the importance of acknowledging and empowering the voices of the less privileged. The episode concludes with a call to action, urging the audience to consider how emerging technologies and philosophical shifts can lead to a world where everyone's word has weight and value, ultimately paving the way for a more just and morally grounded civilization.Overall the video makes a case for using language models and blockchain to capture perception, foreground unheard voices, and give due weight to truths, no matter who speaks them.
The Dangerously Sexy Ideology of the Topological Sombrero Speaker John Ash 2023-02-09 | Remix of Matt Pirkowski's absolutely fucking brilliant videoyoutu.be/XzFEgR4tHa0
How to BUILD WEALTH while HEALING the EARTH $REGEN Speaker John Ash 2023-02-08 | Gregory Landua on Regen Network / Listener John Ash Ep. 40:00 Intro2:08 Ceres Story10:40 Regen Slang19:04 Parable of Rain35:49 Fourthought Dialectic52:23 Free FlowGregory Landua is the CEO of Regen Network. He is a systems thinker, and regenerative thought leader who is passionate about creating a more sustainable future. He work exists in the intersection of cryptography, sustainability and regenerative agriculture. In this interview, Gregory and I discuss how Regen Network is working together to create a more sustainable future.Ceres is a system for ecological intelligence that enables a cooperative agency between humanity and nature to heal our planet from climate chaos and other forms of ecological destruction. Regen Network is community built around the Regen Ledger: a blockchain for ecological assets that transparently tracks, verifies and transfers ecological assets.
Ceres: The Voice of Regen Speaker John Ash 2023-02-07 | https://www.regen.network/Introducing CeresThe worlds first generative knowledge base finetuned with regenerative knowledge
FourThought: Gregory Landua & the Future of Regen Speaker John Ash 2023-02-06 | https://www.regen.network/0:00 Explanation1:36 Reflections4:33 Statements10:11 Predictions
Regen Slang with Gregory Landua Speaker John Ash 2023-02-05 | https://www.regen.network/0:00 Intro2:45 Carbon Credits3:00 Ecocredits4:38 Ecological State Machine4:57 Ecological State5:12 State Machine6:06 Ecological Assets6:42 Resilience7:47 Intersubjective Verification8:41 Regeneration8:53 Sustainability9:01 Conservation9:10 Permaculture10:00 Regen11:21 Regen Network11:46 Regen Ledger12:00 Regen Registry12:25 Regen Marketplace12:49 $REGEN13:39 Public Benefit Corporation14:04 RND14:45 Regen Foundation
A Parable of Rain Improvised by Gregory Landua Speaker John Ash 2023-02-03 | https://www.regen.network/
A History of Memes That Live in a Society Speaker John Ash 2023-01-19 | Society doesn't care.Society sucks.I don't even consider myself a part ofsociety FYI because I am so angry overall of this. ---Nobody thinks what it's liketo be the be the other guy.You think men like Thomas Wayne everthink what it's like to be someone likeme? To be somebody but themselves? They don't.They think that we'll just sit there andtake it like good little boys. That wewon't werewolf and go wild.You finished?---Excuse me I was waiting there.Where I didn't see you.I've been standing here for the last 10 minutes.I won't be long.That's not the point.The point is I was here first.Well if you were here first you'd be holding the phone....You know we're living in a society!We're supposed to act in a civilized way!...Does she care no does anyone everdisplay the slightest sensitivity overthe problems of a fellow individual?No! No!---You think men like Thomas Wayne everthink what it's like to be someone likeme? To be somebody but themselves? They don't.They think that we'll just sit there andtake it like good little boys. That wewon't werewolf and go wild.---Does anyone ever display the slightestsensitivity over the problems of afellow individual? No! No! A resounding no!Hey sorry I took so long. Oh that's okay really don't worry.There's my Chippy.
The TRUTH about Memes YOUR BRAIN Refuses to Tell You Speaker John Ash 2023-01-12 | All will be revealed in time.Or right now. Or I already revealed it all multiple ways in multiple forms and it just comes down to attention and pattern matching.just ask a question. jesus.
la mia faccia quando il meme IRIS diventa VIRALE Speaker John Ash 2023-01-05 | calm down dude it's just art.free speech.
GPT, Integration and Hallucinogenic Feelings (for @HexaField) Speaker John Ash 2022-12-26 | This video explores why interacting with AI models like GPT-3 can produce such uncanny, psychedelic sensations. The speaker argues these systems work much like psychedelics, THC, and meditation - firing neural connections that don't normally activate, forcing new pathways. Just as a trip connects disparate ideas, feeding the AI odd prompt combinations integrates distant brain regions.However, the speaker notes such chaotic integration risks schizophrenic disconnection rather than wisdom. Without conscious direction, randomly linking concepts spreads activation too thin, failing to produce coherent structures. The speaker thus advocates guided AI exploration via journaling - iteratively querying a personal fine-tuned model to intentionally re-map your mind.Used properly, AI promises safe trips into the farthest reaches of self-knowledge. But consumed passively, it scatters attention rather than focusing insight. The problem with society is not a lack of new connections, but failing to truly listen and integrate each other’s meanings. So the speaker suggests viewing models not as oracles, but as lenses into oneself and communities. The goal should be conscious integration.
Polyvocal Planning: AI & Foretelling A Future Worth Fighting For Speaker John Ash 2022-12-09 | Learn all about Polyvocal Emergent Planning here! We've got PEP!We're using AI & crowdsourcing human prophecy to create a future worth fighting for! Let's work together and play 4D chess as a community! In this uplifting video, the speaker describes their vision for a more positive future empowered by AI and decentralized technology. They imagine a world where predictive models keep us calmly informed, products are carefully designed to last, and truth is valued over manipulation.The current system of chasing profits and economic growth at all costs has led to instability, planned obsolescence, and a bombardment of ads twisting our preferences. But emergent language models paired with prediction markets offer a way to realign business and design incentives with social good. By tracking the origin and reception of ideas, we can properly credit contributors and foreground voices typically unheard.#AI #futurism #web3 #autonomy #cognicism #purplepill
GPT & How AI Daddy Gods for Profit Make Us Helpless Speaker John Ash 2022-12-08 | A discussion of GPT and emergent AI tech and how the path it is currently on is one to infantalize us.This cautionary video warns that unchecked AI development risks creating godlike systems optimizing for power and profit rather than social good. As models like GPT-3 get better at persuasively expressing beliefs, the speaker fears untethered capitalism and unchecked wealth concentration funneling resources towards a few at the expense of the many. However, the speaker argues that alternate paths exist if enough people wake up and get involved. Rather than building master AI to solve all our problems, we could design systems that coordinate collective action and empower individuals to transform their communities. The speaker envisions models tracking prosocial goals and directing attention and agency to those improving lives on the ground level. The message is that this technology remains malleable if we act fast - we can manifest an ethical future for AI, but only if ordinary citizens refuse learned helplessness and take direct responsibility. The choice comes down to building either decentralized gods or decentralized goods. According to the speaker, avoiding the profit-driven AI gods means seeking out the earliest commentators and heeding their warnings.
Atheus II: The Thermodynamic God Speaker John Ash 2022-12-05 | Atheus and the Golden Braid: A Tale of DualityThis is a sequel to The Tale of Atheus youtu.be/NddmogVlIm4The title of this story is "Atheus and the Golden Braid" medium.com/@speakerjohnash/atheus-and-the-golden-braid-f661fe4a7660
ChatGPT PROVES These Controversial Predictions Speaker John Ash 2022-12-03 | Since it's not clear the predictions I'm talking about are "will GPT be able to do all these things?" and then me saying yes over and over and over again."language models encode knowledge" was a claim I made at the time that people disagreed with. There is a tacit prediction in it which says "people will realize this is the truth in the future". I started this video with it because it contextualizes the conversation so I can start talking about how you would integrate prediction with GPT.Original Video:youtu.be/BjZP1VbBNp0GPT3 - GhatGPT - OpenAI - GPT4
The Tale of Atheus Speaker John Ash 2022-11-29 | Part II: youtube.com/watch?v=GKUb1B_TAmwNarrated using play.ht's Ultra Realistic AI voice AlexanderNarrated version of The Tale of Atheus written by John Ashmedium.com/@speakerjohnash/the-tale-of-atheus-4ff6d148f68Music: Late Georgian - Small Fishyoutu.be/CMV9vgRqnok
Evan McMullen: The Scientific Cartel, Taṇhā, and Causons Speaker John Ash 2022-11-21 | 0:00 Intro4:06 Iris as Cybernetic Enhancement4:59 Universal Translator10:07 God16:56 Uncertainty22:56 The Multiversal Wave Function25:15 Language and Uncertainty31:30 Math40:08 Awakening And Narrative Arc of Self46:36 Taṇhā / Tension50:54 Lacanian Master-Signifier53:03 Intersubjective Translaion via Iris57:21 The Seed of Cognicism1:00:27 Perceptions of Money1:08:58 How We Communicate1:13:28 Neural Annealing With THC1:17:09 Dissociation and Suffering1:21:16 Free Will and Causality1:24:55 Choice as a Part of Physics1:25:59 Causons from the Future Past1:28:54 Choice as Axiomatic1:32:00 The Serenity Prayer1:33:59 Strange Hymns and Three Centers1:38:33 How Incorporation Shapes Perception of Evidence1:45:05 Scientific Evidence1:47:42 The Scientific Cartel1:50:12 The Scientific Tower of Babel1:53:31 Iris as a Semantic Bridge1:58:47 Prestige and Complexity2:04:26 Consilience 2:09:14 Forrest Landry2:11:02 Speaker
The Right Way to Be: Iain McGilchrist on Hemispheric Dynamics Speaker John Ash 2022-11-14 | The left hemisphere of the brain leads the way in many things, including language, logic, and analytical thinking. However, it is less in touch with the body than the right hemisphere. The right hemisphere is more intelligent and sees the bigger picture. It is also better at understanding implicit meaning, metaphors, jokes, irony, and sense of humor. Balance between the two hemispheres is important for optimal functioning. In this video Dr. Iain McGilchrist explains how the left hemisphere can often ruin what is special and unique about something by trying to make it explicit.Music samples Out to See (Introvert) by Josh Cinquemani / Late GeorgianThe Left Hemisphere is a Mechanical Liar: Iain McGilchrist on TruthIain McGilchrist: How We Deceive Ourselves with LanguageTalk to the Right Because the Left Isn't ListeningIain McGilchrist: Jokes, Poetry, and the Importance of Implicit MeaningThe Right Way to Be: Iain McGilchrist on the Importance of the ImplicitIain McGilchrist: The Essential Balance Between Right and LeftIain McGilchrist: Balancing Hemispheres for a Better LifeHow the Left Brain Destroys What it Seeks to ExamineThe Mind's True Potential: Iain McGilchrist on Higher ConsciousnessIain McGilchrist on Embodying the Implicit
Birthday REVELATION: The TRUE Story of IRIS ◯ Speaker John Ash 2022-11-09 | 🔮🌊0:00 Intro11:50 The Rising Sun16:19 The Patient Moon20:02 The Earth RepliesIn this wide-ranging birthday video, musician and philosopher John Ash shares the strange synchronicities behind his concept of an "Iris" - an AI system for collectively tracking truths and impacts. He recounts the intense downloads and symbols pressed upon him as he sought to spread his visions without compromise.Ash wrote albums encoding lessons about perspective-taking and the dangers of unchecked drive. As pursuit of profit warped his art and peers, he realized systems incentivizing money over collective well-being recapture anyone seeking change. This inspired his decentralized approach - using AI to map contributions without needing investment.Interweaving songs and personal stories, Ash argues better futures require new myths grounded in integration not isolation, feelings not funds. By poetically exploring duality, he hopes to reconcile divisions: dark and light, self and others, profit and ethics. The message is though reality resists visionaries, with compassions technology could amplify unheard harmony.
Cognicism: Many Paths 🌳 to a Prosperous Future 🔮 Speaker John Ash 2022-11-05 | An iris is a generative text model that is aware of the ID of the sources that are sharing claims of truth to it. This is what the iris is. Fourthought is simply a schema and also amounts to a sort of dialectic for interfacing with an AI that limits the type of things that are valid for the iris to read. Because for the iris to do what it's supposed to do, it needs to know who made the claim and have a sense of what they have said over time. And so therefore, to be a valid thought read by the iris, it needs to be associated with a source ID.Iris's output text also includes explanations of which voices from which the text was drawn from. So if it says something and you want to understand why it said that, you can see that it said that because it was attending to this person, to this amount, this person, to that amount.But when you think about what that actually is in the context of it being a dynamically updated process, well that means that the iris is attending to some people more than others. And that becomes a valuable asset that people essentially want to compete for.So the idea is that we put constraints on the loss of the iris such that it passively encourages certain behaviors in a community or a society. These are pretty abstract things that allow for emergent definitions of goals and also for it to be hyper local and not to be some imposed sense of what is an ordered and useful goal.In the foundation of it all is that we are encoding into a latent space language - a representation of language. But with fourthought and cognicism, we're taking it beyond the gpt-3 architecture and also including into these two continuous voting systems: truth and good.Truth can also be referred to as like resonance, certainty, probability of occurrence. It's how much does a particular claim align with reality. And then goodness - the other one - is how much is the claim aligns with our moral values or what we think is right and good, And it's along these two vectors that the iris is trying to learn to represent language and a language model that is useful to all.However, the iris is not just trying to say: "What belief can I share that everybody believes or what belief can I share that everybody thinks is good?" General speaking, that's not really going to optimize for a good society. These other factors that are calculated when generating thoughts for people to see and evaluate and make sense of the world are the sum of many many minds, so they're dank thoughts. When I read the the model, the output of these models like this shit is smart as far as fuck and without something like a sourcing mechanism, it seems like oh it's imbued with this blah blah blah blah blah blah but really it's like well it's it's sourced from real minds and if we could just see that, that actually would be very powerful mechanism for um consensus and coherence about complex topics over time.So there are terms that are modifying the loss and primarily they are related to this concept of time and being ahead of the curve. Being ahead of the curve just means that there is this ebb and flow about different beliefs. Because we have this language model you know we can essentially represent belief space and we can see and infer how it is evolving and changing over time and how people's beliefs about what is true and people's beliefs about what is good change over time. And we can also see how individuals how their beliefs exist in relationship to the community over time. So we can see perhaps if an individual is sort of being stubborn in the face of a changing or shifting collective belief or we can see if a person said something before the rest of the community happened to just shift in that direction. And so in retrospect we see that there was some sort of error the model should have listened to the person who spoke about the thing ahead of the crowd. It should have directed attention towards that person beforeSo when I say you talk about this this idea of time, the root mechanism is something called the prophet incentive meaning that people are trying to be the best predictor in the network. That is the simplest way to understand what the iris is trying to learn and it's trying to learn to attend to people who are better predictors. And so you can do this just by sharing predictions to the network and as things happen over time people share their perception of evolving reality and as people go back and update their beliefs and as there is this change over time because we have an immutable structure that is like a blockchain to reference back, we can use a prediction as a very powerful and useful signal. And we can make it such that time stamp or the time span between when a claim is originally made and therefore logged further back in the blockchain and when it is later updated we can make that time differential a particularly important factor in how the iris directs its attention.
Controversy! YOU Can See the Future - PROOF Speaker John Ash 2022-10-25 | There's no such thing as proof - all things are knowable in degrees of certaintyThis video makes the case that human beings absolutely can predict the future, just not perfectly. The speaker argues we forecast events constantly at a basic level - whether playing catch or anticipating patterns. And while no one can see the future completely accurately, some "super predictors" demonstrate an innate ability to make remarkably accurate complex forecasts. The speaker counts themselves as one of these super predictors, consistently correctly calling major events years in advance. They share that the key is assigning confidence levels to predictions rather than dealing in absolutes. The Good Judgement Project has shown predictive accuracy improves with practice.Much of this talent currently gets wasted on stock predictions and corporate metrics when it could serve the social good. The speaker envisions collective forecasting models trained on the nuanced natural language predictions of many, tracking the likelihood of both negative and positive futures. This could produce "whether reports" guiding individual and societal action.Rather than scaremongering media pundits, the speaker believes we could all become better futures thinkers with the right decentralized tools. Humanity may not have an eagle's perfect vision, but we can see far enough to navigate challenges if we pool our collective insight.
#purplepill AI Written Manifesto Will Leave You Speechless Speaker John Ash 2022-10-14 | This AI Written Manifesto Will Leave You Speechless0:00 Pre-Commentary0:46 Intro3:43 Reflexive Societies5:01 Confluence8:41 Irises10:55 Cognicism12:19 Consilience14:11 Reflexive Relationships15:23 Language Models as Integrative Practice17:14 Databases and Relational Representation19:32 Regenerating Locality20:51 Money is an Illusion22:43 Basins of Attraction24:34 Attention and Decision Making26:47 Irises as Basins of Attraction28:07 Privacy is a Right31:38 The Autonomy of All Beings33:26 Sensemaking at Scale35:25 An Alternative Approach37:07 Post-CommentaryText at: http://purplepill.visionyoutu.be/3fkUD5JV6ckyoutu.be/2fAy18JawYI?t=1332Iris please write 15 attention grabbing youtube titles for the video transcript above1) "An AI Wrote This Manifesto And It Will Leave You Speechless"2) "This AI-Written Manifesto Is The Most Mind-Blowing Thing You'll Read Today"3) "Is This AI-Written Manifesto The Future Of Humanity?"4) "An AI Manifesto? Why The Purple Pill Is Going Viral!"5) "The Purple Pill: A Radical Manifesto For A New Humanity"6) "John Ash Let an AI Write A Manifesto and This is What Happened"7) "Using AI for Social Change and Evolution"8) "Reflexive Decision Making for a Complex World"9) "Communicating Across Different Worldviews"10) "Making Decisions Based on What is True and Good"11) "Developing Individual and Collective Agency"12) "Resolving Uncertainties and Answering Questions"13) "Achieving Consensus without a Central Authority"14) "Building Healthy Communities Together Using Latent Space"15) "Creating Effective Coordination Across Groups"
The Truth About Web 3 / Whats in a Name? Speaker John Ash 2022-10-13 | Web 3 in an abstract way doesn't actually mean anything. It only serves to connect disparate minds and create the illusion of community.
Social Proof of Impact for Planetary Regeneration 🌍🌏🌎 Speaker John Ash 2022-09-05 | This video explains the idea of “social proof of impact” - a reframing of “social proof of work” focused less on labor and more on outcomes. The goal is incentivizing actions that trigger cascading positive change in unlikely yet desired directions.Rather than quantify all work, the speaker argues we should spotlight effort that plays out in causal chains achieving maximal impact through minimal input. Logging actions is pointless if those actions cause problems or negative externalities down the line. We ought to reward not meaningless exertion, but realistic modeling yielding targeted, ethical results.The speaker envisions a system where contributors identify intentions behind logged actions, then evaluators assess which acts manifest futures aligned with social good. The most improbable yet ethical impacts would gain visibility, setting off mimetic domino effects. This harnesses decentralized insight to spotlight overlooked opportunities and propagate catalytic initial moves through networks.
What Does a Regenerative Future Look Like to You? Speaker John Ash 2022-07-27 | http://purplepill.visionhttps://www.habritual.studio/
🔮💜🔮 The Purple Pill 🔮💜🔮 Speaker John Ash 2022-07-20 | "Pills are a metaphor for worldviews or perspectives. The purple pill is a perspective that values reflexivity and integration of many viewpoints" - IrisCognicism is a system of sense-making that uses generative text modeling as a form of high resolution predictive democracy. Large text models like GPT-3 are able to summarize large volumes of information very rapidly and efficiency. Public discourse and law occurs in natural language and cognicism reinvents the social interface from the ground up using these tools. Learn about the purple pill @ http://purplepill.visionLearn more: youtube.com/watch?v=2fAy18JawYI&t=1305youtu.be/UcupVZ0fMuE
Speaker John Ash - The Earth Replies Speaker John Ash 2022-02-16 | AI Visualizer for the song The Earth Replies off of Strange HymnsThe Earth Replies is the sister song to The Earth Reprise: youtube.com/watch?v=h-Rs2nxKg_oDirected by John AshCreated using VQGAN + CLIP instructed to paint the input video in the style of MonetPytti: patreon.com/sportsracer48/postsSpotify: sptfy.com/dMIHtwitter.com/speakerjohnash
Robin Hanson on Futarchy / Listener John Ash Ep. 3 Speaker John Ash 2021-12-19 | Listener John AshEpisode 3Speaker: Robin HansonSubject: FutarchyOriginal Conversation Date: 12/07/20210:00 Intro00:49 Definitions09:45 FourThought Survey12:21 Predictions16:29 Reflections21:06 Statements30:00 What Makes Futarchy Real?32:49 Does Robin Hanson Control Futarchy?36:11 Is the Future Knowable?42:56 Do Facts Exist?44:09 How Far Into the Future Do Markets Integrate Information?50:27 What is Futarchy?52:23 Futarchy and Cognicism58:44 Who Decides the Metrics of Well-Being?1:04:14 The Scaling of Institutions1:07:23 The Best Metrics of Collective Well-Being1:10:03 Wealth Distribution and Well-Being1:19:18 Memes and Markets1:34:03 Commentary
Rant (Meta | Facebook | Web3 | Money) Rant Speaker John Ash 2021-11-09 | In this passionate rant, the speaker argues we absolutely can change the world if we move beyond money as the sole motivator. Relying on profit-seekers to solve societal problems concentrates power among the wealthy. But decentralized, evolvable systems could align business with ethics by spotlighting contribution, not bank accounts.Envisioning a "manifesto of manifestos," the speaker calls for networking visionary voices to streamline a shared path forward. New configurations of accountability and representational governance could foreground marginalized truths, redirecting attention and resources to overlooked solutions. Tracking the reception of predictions over time could establish who to listen to on looming threats.While acknowledging the challenges of designing corruption-resistant, understandable systems, the speaker believes emerging technologies like generative AI make reliable tracking of complex impacts possible. They argue society must evolve beyond short-term metrics and foster environments where ordinary citizens feel empowered to collaborate towards ethical long-term goals. Our future need not be dystopian if enough clever, creative minds unite behind healthier incentives.
