Podcasts about Cosmos

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

We Are Eternal returns to Sofa Beats with another release, continuing the exploration of dancefloor music at the intersection of mid-tempo trance and tribal folk rhythms and melodies. This release is an attempt to bridge the future with the ancient and forgotten past. Kronos draws inspiration from the distant, fog-covered past and the forgotten cycles of humanity whose remnants are beginning to resurface in the collective memory. The track evokes visions of ancient advanced civilizations — before Egypt, before Atlantis, Lemuria, and Mu — and seeks to reconnect us with the lost wisdom of our predecessors. Musically, Kronos is a powerful psychedelic roller infused with swirling sequences and an otherworldly atmosphere, merging the worlds of chilgressive and deep melodic downtempo. A Native American flute opens the portal, while haunting, mysterious vocals and hypnotic hang drum rhythms guide the listener into the forgotten past. Minimalist yet driving percussion steadily builds toward the drop, where epic leads and enigmatic melodies unfold. WAE's signature talking FM synths bring the composition to a cinematic and powerful conclusion. Roots to Cosmos reminds us that in order to reach higher cosmic knowledge and transcendental realms, we must remain grounded and rooted in reality. Exploring our heritage and honoring the wisdom preserved within pre-Christian traditions opens pathways to cosmic consciousness. Faster, also strong and stompy, the track features more intricate percussion woven together with glitchy cosmic phrases and immersive sound design. Massive melodic psytech leads carry the listener deeper into the cosmic realm, while acoustic and organic instruments maintain a connection to the material world. The drop introduces an unexpected melodic turn, revealing yet another layer of the journey. In the grand finale, an epic folk violin ensemble brings forward an ethno-Slavic atmosphere — a reflection of Piotr's personal heritage as the musician behind We Are Eternal. A massive acid lead then ties the entire experience together, bringing this epic ride to its final culmination. Those tracks are designed for deep festival and club dance floors with the intention to connect us with high knowledge of the ancient past.

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

Jump into Talk Cosmos on June 21 from 1-2 p.m. PDT for the “Summer Solstice Vibrational Patterns”.  Join us as we explore these vibrant energies through VA's unique lens and embrace the solstice as an energy portal into this season's full potential! Through the unique lens of Vibrational Astrology's software system, we will discover distinct collective consciousness patterns. Deep within the 2026 summer solstice chart, these frequency patterns profoundly shape this season. Since pre-historic days, humanity kept watch on the seasonal turns by observing the sky. Embraced as a spiritual portal to celebrate abundance, the Summer Solstice launches the return of the light in the Northern Hemisphere. Traditional bonfire rituals lit up the night sky, to symbolize life's manifestation and renewal. SOLSTICE LATIN ROOTS  On June 21st at 8:24:13 a.m. GMT and at 4:24:13 a.m. EDT, the Sun transits 0° Cancer while it appears to stop for three days and ‘hover' over the northern Tropic of Cancer.  This apparent ‘celestial pause', known as SOL-STICE derived from Latin roots, literally means: “Sun” "Stops". Earth tilts on its 23°26' north south axis, placing Earth's northern hemisphere to lean closest to the Sun during its annual orbit. Daylight stretches to its fullest, casting long shadows to ignite our spirits with warmth and possibility. UPCOMING: FRACTAL COSMOS CONFERENCE Registration for the 2nd Annual Fractal Cosmos Vibrational Astrology Conference on November 22-24, 2026 (fractalcosmos.org) opens during July. Online – join from anywhere. LINDA BERRY, PAC, MSSW: received her Professional Astrology Certificate (PAC) in Vibrational Astrology January 2015 from Avalon School of Astrology studying with David Cochrane the Founder of Vibrational Astrology (VA). They continue to share their research material to build Vibrational Astrology knowledge. Linda created “Frequency Finder”, a VA Add-on to Sirius and Kepler Astrological Software. Linda's an International Consultant with clients worldwide, Teaches VA classes, the VA Research Group Moderator, and Author. Website: Astrosleuth.org | Fractal Cosmos Vibrational Astrology Conference - Annual. Website: fractalcosmos.com Her free Daily Blog: “The Vibrational Astrology Diary” Vibrational Astrology & Sabian Symbols, and for a paid Personalized Monthly Report. email: Linda @ AstrologicalDepth dot com. ROBERT PACITTI: Professional consulting astrologer; visionary behind Deep Earth Astrology. Specializing in vibrational and psychological techniques. Over a decade of experience in the world of natural magic. Grand Pendragon in the Ancient Order of Druids in America & Director of the MAGUS Druid Gathering in Gore, VA. Co-Director of the Fractal Cosmos Vibrational Astrology Conference. Faculty for the Centre for Relationships and Astrology. Consultations focus, Archetypal & Harmonic. Email: deepearthastrology@gmail.com. Website: deepearthastrology.com | Facebook.com/SacredConnections13; Facebook.com/rjpacitti fractalcosmos.org SUE ‘ROSE' MINAHAN: Evolutionary Astrologer & Consultant. Speaker, Writer. Student of Vibrational Astrology with Linda Berry, Dwarf Planet University graduate, Kepler Astrologer Toastmaster (KAT); Founder of Talk Cosmos since April 7, 2018. Weekly conversations awaken heart and soul consciousness, TalkCosmos.com | YouTube.com/@TALKCOSMOS. #SummerSolstice #VibrationalAstrology #Astrology2026 #TalkCosmos #SueRoseMinahan #lindaBerry #RobertPacitti #DavidCochrane #AstroSleuth #DeepEarthAstrology #MagusGathering #Tarot #ancientOrderofDruidsinAmerica #EvidenceBasedVibrationalAstrology #ChironinTaurus

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

In the Electrek Podcast, we discuss the most popular news in the world of sustainable transport and energy. In this week's episode, we discuss Jamie's first drive in Aptera's solar car, Tesla Cybercab specs revealed, Lucid Cosmos design leak, and more.

Dean Takahashi is the dean of tech writers and a 25-year veteran correspondent covering consumer electronics, gaming, and emerging technology for GamesBeat. He's covered every major tech transition—from mobile's rise to VR's boom-and-bust cycles to the current AI explosion—with a skeptical eye and a talent for finding the human story beneath the hype. This is his fifth appearance on the AI XR Podcast.For CES 2026, Dean walked the floors across the Convention Center, the Venetian Expo Center (Eureka Park), Pepcom, and Showstoppers, emerging with a clear reading: China has decisively shifted from periphery to center stage in consumer electronics manufacturing, American incumbents are pulling back and rethinking their booth strategy, and the economics of CES itself are in transition. Robotics companies are moving from prototype to commercial faster than expected—but they still can't answer basic questions about pricing and labor displacement.News: Sony cuts its booth to demo an electric car instead of TVs. Samsung skips the show floor entirely for the first time. Nvidia takes over the Fontainebleau to showcase its role in robotics enablement. Lenovo dominates the Sphere with a Gwen Stefani concert. Chinese robotics companies proliferate with laundry folders, latte makers, and toilet-cleaning units. Roomba files for bankruptcy; Chinese competitors take over the robotic vacuum market.Key Moments:[00:01:23] Dean receives his virtual green jacket as a five-time returning guest and Charlie thanks him for his insights[00:03:00] China takeover at CES: TCL dominates Central Hall, ROED owns the XR booth, robotics companies fill the floor[00:06:00] Nvidia's Fontainebleau takeover and the "chest-pumping" show of force; why scale messaging still matters[00:14:18] The robotics explosion explained: Nvidia's digital twins, Cosmos world models, and synthetic testing accelerate time-to-market[00:19:00] The pricing problem: robotics companies won't answer how much their products cost; the minimum wage rental model doesn't translate globallyWhen American companies built the show, CES reflected American manufacturing dominance. Now that China manufactures most consumer electronics, CES reflects that shift—and the implications ripple through labor, supply chains, and where the next epicenter of innovation will be. Dean, Charlie, and Ted grapple with what CES 2026 signals about global manufacturing advantage and why the geography of tech matters more than we think.This episode is brought to you by Zappar, creators of Mattercraft—the leading visual development environment for building immersive 3D web experiences for mobile headsets and desktop. Mattercraft combines the power of a game engine with the flexibility of the web, and now features an AI assistant that helps you design, code, and debug in real time, right in your browser. Build smarter at mattercraft.io.Listen to the full post-CES debrief and subscribe for weekly conversations at the intersection of AI, XR, and consumer technology. Hosted on Acast. See acast.com/privacy for more information.

Host Dave Schlom is joined by two scientists from the SETI Institute in Mt. View, California, to talk about one of hisfavorite celestial phenomena -- comets.

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

Join Michael at his Mt. Shasta Summer Retreat, July 30 - August 2! Click here: https://events.agapelive.com/mt-shasta-summer-retreat/.  Today, Michael welcomes Dr. Sue Morter.  Sue is an internationally acclaimed speaker, Master of Bioenergetic Medicine, and thought leader bridging science, spirituality, and human potential. As the founder and CEO of the Morter Institute for Bioenergetics, she teaches how to reprogram the body's electromagnetic energy flow through embodiment, self-healing, and higher consciousness. Her latest book, The Anatomy of Awakening, explores the hidden codes that help unlock the truth of who we are and live from our divine destiny.  Conversation Highlights Include: -Why awakening is not about fixing yourself, but remembering the divine wholeness that has always been who you truly are. -How embodiment shifts spiritual practice from trying to leave the body into fully inhabiting life with presence, clarity, and purpose. -A deeper look at the "God body" as the energetic bridge between the physical self and the higher consciousness we are learning to embody. -The importance of rewriting your identity so you stop seeing yourself as separate from Source and begin living as an expression of it. -How the body uses sensations, tightness, emotion, and energy to guide you back toward the parts of yourself that need awareness and integration. -Why spiritual practice is not meant to be an escape from life, but a way to bring your true self into relationships, family, work, and the world. -How bringing attention back into the body helps reclaim energy, soften old protective patterns, and expand your capacity to live authentically. -A powerful reframing of Love as the organizing force of the Cosmos, not just an emotion exchanged between people. -Why heaviness, sadness, anxiety, or fear can become doorways into healing when you follow them inward with breath, presence, and compassion. Michael closes the episode with a guided meditation on the inward smile, gratitude, divine identity, and remembering that everything you need is already within you.

The U.S. plans to reduce the number of aircraft for NATO operations, another A-10 lifeline appears, and discussions about restarting C-17 production. Also, owner-produced airplane parts, airport weirdos, a new album from Speed Brake Armed, how the NTSB uses audio spectrograms, lying flat on a broken Polaris seat, and Roman Numerals. Aviation News US Plans Major Cut to Fighter Jets, Warships for NATO Operations in Europe, NYT Reports Citing European officials, the New ​York Times reported that the U.S. plans to reduce the number of ⁠F-16 and ⁠F-15E fighter jets from roughly 150 to 100. Maritime reconnaissance ​aircraft would be cut from 26 to 15, and all eight aerial refueling tankers would be pulled. The ⁠New York ​Times said the U.S. aims to redeploy a missile-launching ​submarine and an aircraft carrier, along with several warships and jets ⁠that join ⁠the carrier's missions. One of two groups of bombers previously assigned for ​Europe's defense may also ⁠be reallocated. NATO spokesperson Allison Hart told Reuters, “Historically, there has been an over-reliance on U.S. forces and capabilities.” The U.S. European Command said in a statement that it would “rightsize” its contributions to the NATO Force Model. Congress Throws A-10 Warthog Another Lifeline The A-10 end of life is scheduled for 2030. Depot-level maintenance has stopped, and the 571st Aircraft Maintenance Squadron at Hill Air Force Base, Utah, has ended. The A-10 Weapons School is scheduled to end this year. However, an amendment to the House Armed Services Committee's version of the National Defense Authorization bill seeks to keep the Warthog alive. The amendment calls for the Air Force to keep supporting A-10 training, testing, experimentation, maintenance, and sustainment efforts. Other requirements include preserving lessons learned and operational expertise and maintaining a formal pilot training unit. A-10 Warthog's New Aerial Refueling Probe Is Now Operational In The Middle East The A-10C is now operating in the Middle East with the new probe-and-drogue refueling capability. First demonstrated in early April, it took only six weeks to become operational. Previously, the A-10 could only refuel from a KC-135. The KC-46 was not yet certified to refuel the Warthog due to the “stiff boom” problem, which could damage the receiving aircraft. Now A-10s can refuel KC-46s with the probe or from HC-130s, MC-130s, Marine Corps KC-130s, and KC-130Js from other operators. A-10 with refueling probe. USAF photo. Boeing “Encouraged” By C-17 Production Restart Discussions Restarting C-17 Globemaster III production would be extremely difficult, extremely expensive, but not impossible. There is interest from various operators and from the U.S. Congress, which has asked the Air Force to prepare a formal briefing on the feasibility of acquiring new C-17s. Driving USAF interest is a succession of crises in recent years that have put serious strain on the aircraft, and questions have been raised about the viability of the current plan to keep them flying through 2075. The C-17 is powered by the F117-PW-100, which is the military variant of the PW2000 family (the same engine that powers the Boeing 757). New engine production for the PW2000 stopped in 2016, and the USAF is currently depending on overhauls of existing engines to keep the fleet flying. So the MRO infrastructure, engineering expertise, and supply chain for supporting this engine remain very much alive. In March 2025, RTX announced agreements with JetZero to integrate the PW2040 engine and APU into its blended-wing-body demonstrator. So P&W is actively working on the PW2040 for a new application, which suggests the engine isn’t entirely dormant in their engineering ecosystem. The decision to restart the engine isn't just a P&W decision. The risk-sharing partners, like MTU Aero Engines, have to be on board. There are 222 C-17s in service with the U.S. Air Force today. The last plane was delivered in 2013, and Boeing shut down the line in 2015. Australia, Canada, India, Kuwait, Qatar, the United Arab Emirates, and the United Kingdom operate the C-17. C-17. USAF photo. Listener Mail Eclipse spare parts Mark writes regarding the discussion about Eclipse parts from Episode 896 and notes that FAR 21.9(a)(5) creates a framework for owner-produced parts. Where a certified part is unavailable, owners of certified aircraft can “produce” their own. And they can do that either by making it themselves or by contracting out its production to a suitably qualified supplier. There are rules about quality and the requirement that owner-produced parts be of equivalent specification to OEM parts, but as long as an aircraft owner can put their hand on their heart and assert that those conditions are met, they can supply parts to their maintainer and tell them to install them. See this AOPA guidance. Airport Weirdo Koeby has developed a crowdsourced gallery of airport weirdos, where travelers submit funny photos of strange things they spot in airports. No account is needed; you can just submit your photo, and it will be added to the gallery. It's called Airport Weirdo. New Album release by Speed Brake Armed Pete Buffington tells us about Speed Brake Armed’s new New Age album “Echoes Above the Infinite Sky.” This album takes the listener on a journey of flight from South America, to Spain, to the Cosmos, and back to ancient Greece. Inspired by over 35 years of real pilot experience. Video: 737 Echoes Above The Infinite Sky | Speed Brake Armed | Full Album | New Age Aviation Music https://youtu.be/slO-4xnVqHg Spectrograms Andy adds his perspective about the conversation on spectrograms in NTSB investigations. While he has absolutely no actual knowledge about NTSB processes or how they actually use spectrograms, he speculates based on his experience as an audio engineer for over 30 years: “Spectrograms have been a tool I use fairly regularly in production. To me, it mostly comes down to being able to recognize things that are hard to pick out. For instance, if there is some kind of unpleasant noise in the background of a recording, sometimes I can identify it and potentially filter it out, purely by ear.  Other times, particularly if it's not very far above the noise floor, it can be very difficult to pick out by ear.  In that case, I'll often look at a spectrogram. It's certainly not always helpful, but sometimes there are things that I can pick out visually that I can't pick out audibly… “So I can imagine that in a cockpit recording with a lot of background noise, examining the spectrogram might allow patterns to be detected that would not be obvious audibly. My guess is that they wouldn't be looking at the speech, but rather for indications in sound of what was happening mechanically. “For instance, if there was sound at a particular frequency, happening at a particular interval regularly, that might be an indication of something. That's the sort of thing that you can often see on a spectrogram even if it is audibly buried in the noise floor.” 14 Hours Lying Flat Patrick thinks maybe United could have done better: 14 Hours Lying Flat: United Polaris Passenger Pays $7,400, Gets Just $350 For Broken Seat. A United Airlines passenger has recounted her experience of flying in a faulty Polaris seat. She was forced to sit in a lie-flat position for the entire journey. After complaining, United offered her only $350. The ticket cost $7,388. DCCCXCIV Rob wrote in to say he enjoyed the value that Erin Applebaum brought to Episode 894. Also, that “with the very welcome return of David, this episode may well be the first podcast ever where the hosts have an odd number of kidneys!!” We also got a refresher on Roman Numerals. Mentioned The Great State of Maine Airshow, Saturday and Sunday, July 11 and 12, 2026, at Brunswick Executive Airport (the former Brunswick Naval Air Station). DARPA Lift Challenge at the National Museum of the Air Force in Dayton, Ohio.  Aug. 5-9. Hosts this Episode Max Flight, our Main(e) Man Micah, Rob Mark, and David Vanderhoof.

Come journey with me to the Cosmos and engage the Holy Spirit of God

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

Miradas al cine latinoamericano , ¿a qué suena el cosmos?, ¿cuándo los seres humanos empezamos a caminar?

Welcome to Light On Light Through episode 420, in which I interview Deana Weibel about her new book, The Ultraview Effect.  I consider Deana the Margaret Mead of outerspace -- find out why in this interview. Relevant links: More about The Ultraview Effect here Touching the Face of the Cosmos -- the anthology with Deana's Weibel's "Pennies from Heaven" Paul Levinson's reviews of For All Mankind and Star City Interview with Lance Strate in which we discuss his and Neil Postman's views of the space program  

El día de la revelación. Spielberg, el cosmos y sus habitantes Hace apenas unos días, estrenó en cines de todo el mundo Disclosure Day, El día de la revelación, la película más reciente de Steven Spielberg y una de las más esperadas de este año. Para muchos, es una historia más sobre extraterrestres, un tema muy presente en la filmografía de Spielberg. Pero no. Ya que la pensamos bien, creemos que en realidad El día de la revelación es una película con extraterrestres, y ese es el cambio central que don Steven hace aquí sobre el tema. ¿Qué diferencias marca eso y, en consecuencia, hacia dónde quiere llevarnos la película?Justo de eso va este podcast. Y para redondear la propuesta la invitada a este episodio es Nataly Olascoaga, crítica de cine en Fósforo UNAM, la revista digital de crítica cinematográfica de la Filmoteca UNAM y, sépanlo de una vez, entusiasta no solo de las películas sobre extraterrestres, sino de los alienígenas en general. Hablemos de Disclosure Day, Spielberg, los extraterrestres y las desilusiones de los influencers ante la película, todo con Nataly Olascoaga. Learn more about your ad choices. Visit megaphone.fm/adchoices

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

Scriptures Referenced Genesis 1:1-5; Hebrews 11:1-3; Colossians 2:8; 2 Corinthians 4:5-6; 2 Timothy 2:24-26; Colossians 1:13-14Confessions Cited The Apostles Creed The Nicene Creed The Catechism of the Roman Catholic Church The Westminster Catechism The London Baptist Confession The New City CatechismPhilosophy and Forming Actual InfinitesSt. Bonaventure's Arguments https://plato.stanford.edu/entries/bonaventure/Hilbert's Hotel Paradox - https://www.britannica.com/video/paradox-David-Hilbert-hotel/-205800Scientific ReferencesRed Shift and Big Bang Cosmology https://youtu.be/3W2GMSQOZq8?si=C-kUm_KarPJlGFZ3Books ReferencedWilliam Lane Craig, Reasonable Faith Christian Truth and Apologetics, 84.Robert Jastrow, God and the Astronomers, 1978.

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

S. Craig Zahler has created a sci-fi graphic novel that explores the lives of a handful of humans that dare to challenge the nature of space and time and Josh and Travis are tackling this retro alien story as part of their graphic novel summer. See what the boys compare this indie graphic novel to.

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

BUY THE NEW MERCH - PATRONS GET 10% OFF - https://deprogramshop.com/Support the show on Patreon - get up to 3 exclusive episodes a month, discord access, merch discounts and plenty more - check it out: https://www.patreon.com/TheDeprogramSupport the showSupport the show on Patreon: https://www.patreon.com/TheDeprogramFollow us on Twitter: https://twitter.com/TheDeprogramPod

Send us a text or leave a voice message!email us! twroadpodcast@gmail.comhttps://buymeacoffee.com/twrpodThis week on The Wandering Road, Chris and Dean welcome back Mike for a thought-provoking discussion about aliens, UFOs, and the future of disclosure.Mike shares his perspectives on extraterrestrial life, unexplained sightings, and the controversial topic of channeling alien beings. The conversation explores whether disclosure is coming, what it could mean for humanity, and how society might react if the existence of non-human intelligence were officially confirmed.Join us for an open-minded exploration of one of the biggest mysteries of our time—and the possibilities that may lie beyond our world.Support the showSOCIAL MEDIATwitter: @TWRoadpodcastIG: twroadpodcastWant to be a guest or share your paranormal experiences?  Email us!twroadpodcast@gmail.com

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

TodoPorLaRadio con Toni Martínez, Especialistas Secundarios, Mario Panadero, Lydia Ramón, Cristina del Casar, Alfonso Cardenal, Arkano y Javier Coronas

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

TodoPorLaRadio con Toni Martínez, Especialistas Secundarios, Mario Panadero, Lydia Ramón, Cristina del Casar, Alfonso Cardenal, Arkano y Javier Coronas

Ilia delivered the Axial Age lecture this past week from her temporary post in Germany, and the questions came in fast enough that we ended up touching the third rail of the whole class — the rise of monotheism. Yes, there was a time when God was not. Yes, there was a time when God was gods. Yes, Psalm 82 is in your Bible, and yes, Yahweh shows up at the divine council to judge the other deities. If that sounds sacrilegious, you might be exactly the right person to take this class. The Axial Age is when the I gets invented, the score gets written, the disengaged knowing that becomes science comes online, and 60,000 denominations begin arguing over whose copy of the score is correct. We are still inside its inertia. We are also — Ilia argues, and I think she is right — already past its hinge.  You can WATCH the conversation on YouTube Ilia Delio, OSF, PhD is a Franciscan Sister of Washington, DC, and American theologian specializing in science and religion, with interests in evolution, physics, and neuroscience and the import of these for theology. Previous Episodes with Ilia Delio Religion Has a Physiology: Ilia & Tripp on Why Rituals Come Before Beliefs The Machine Is a God Image Thinking Theologically about AI with Teilhard de Chardin the Future of Religion The Not Yet God Bonaventure & the Cosmos in Process Catching a Cosmic Faith the Entangled God of my Heart Join our online class – THE FUTURE OF RELIGION⁠⁠⁠⁠⁠⁠⁠ Tripp and Ilia Delio are teaming up for a brand-new four-week online class, ⁠⁠⁠⁠⁠⁠⁠The Future of Religion ⁠⁠⁠⁠⁠⁠⁠— for everyone who's read the books, asked the questions, and realized the faith they inherited doesn't quite fit anymore. Together they'll trace religion's evolutionary arc and map what's emerging on the other side. Includes 4 video lectures, 4 live Q&As (replays available), and a community of fellow travelers. Donation-based, pay what you're able (including $0). Live sessions start this month — register at ⁠⁠⁠⁠⁠⁠⁠www.thefutureofreligion.com⁠⁠⁠⁠ This podcast is a ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Homebrewed Christianity ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠production. Follow ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠the Homebrewed Christianity⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠, ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Theology Nerd Throwdown⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠, & ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠The Rise of Bonhoeffer⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ podcasts for more theological goodness for your earbuds. Join over 75,000 other people by joining our ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Substack - Process This!⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Get instant access to over 50 classes at ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠www.TheologyClass.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Follow the podcast, drop a review⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠, send ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠feedback/questions⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ or become a ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠member of the HBC Community⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠. Learn more about your ad choices. Visit megaphone.fm/adchoices

Come, Joni with me to the Cosmos and engage Yahweh

Dr. Scripture continues answering questions from a group of young people he spoke to at a conference.  These questions relate to the creation of animals and the cosmos.

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

Welcome to Day 2878 of Wisdom-Trek. Thank you for joining me. This is Guthrie Chamberlain, Your Guide to Wisdom. Day 2878 – Wisdom Nuggets – Psalm 134:1-3 Daily Wisdom Wisdom-Trek Podcast Script - Day 2878 Welcome to Wisdom-Trek with Gramps! I am Guthrie Chamberlain, and we are on Day 2878 of our Trek. The Purpose of Wisdom-Trek is to create a legacy of wisdom, to seek out discernment and insights, and to boldly grow where few have chosen to grow before. The title for today's Wisdom-Trek is: The Song of Ascent – The Midnight Benediction of the Cosmic Mountain In our previous episode on this grand, generational expedition, we explored the fourteenth Song of Ascent, Psalm One Hundred Thirty-Three. We peered inside the seamless walls of Jerusalem to witness the radiant, supernatural atmosphere of the kingdom. We discovered that holy harmony among the family of God is an aggressive, defensive weapon that actively subverts the chaotic fragmentation of the Tower of Babel. We felt the fragrant, vertical cascade of Aaron's precious anointing oil, and we marveled at the cosmic inversion of the landscape, where the life-giving dew of Mount Hermon—the ancient, dark stronghold of the rebel gods—was hijacked, and redirected by Yahweh to refresh the holy mountain of Zion. We rested in the ultimate, sovereign decree of life everlasting. Today, my friends, we have reached the final step of this specific trail. We are standing at the absolute conclusion of the fifteen pilgrim psalms, exploring Psalm One Hundred Thirty-Four, verses one through three, in the New Living Translation. This final Song of Ascent is a short, dramatic, and intensely atmospheric liturgy. The great festival in Jerusalem has ended, the crowds are dispersing, and the pilgrims are preparing to descend the mountain under the cover of darkness, to return to their ordinary lives in a compromised world. But before they lose sight of the temple, they turn back one last time to exchange a beautiful, midnight blessing with the guardians of the sanctuary. Let us step onto the final ridge, look into the glowing courts of the Lord, and receive the parting benediction of the cosmos. The first segment is: The Midnight Vigil of the Royal Guardians Psalm One Hundred Thirty-Four: verses one and two. Oh, praise the Lord, all you servants of the Lord, you who serve at night in the house of the Lord. Lift your hands in holiness, and praise the Lord. The final psalm opens with a stirring, midnight call to worship, issued by the departing pilgrims to the staff of the temple. “Oh, praise the Lord, all you servants of the Lord, you who serve at night in the house of the Lord.” To fully appreciate the cinematic, mysterious beauty of this moment, we must paint the physical, and spiritual, picture. The annual feast is over. The campfires on the hillsides around Jerusalem are dying down, and the thousands of pilgrims are packing their bags to begin the long trek back to their distant homes. As they step out into the cold night air, leaving the safety of the inner courts, they look back at the dark, towering silhouette of the temple standing against the starlit sky. The city is quiet, but the temple is still alive with activity. They see the flickering orange glow of the altar fires, and they spot the shadows of the Levites and the priests moving through the corridors. The pilgrims shout out a final, parting charge to these nocturnal ministers: “Praise the Lord... you who serve at night.” In the ancient Hebrew framework, the night watch was a position of immense responsibility. While the rest of the nation slept, these specific servants were commanded to keep the sacred fires burning, to guard the thresholds, and to maintain a continuous, unceasing rhythm of prayer and vigilance within the courts of Yahweh. We must look at this nocturnal service through the profound lens of the Ancient Israelite divine council worldview, as masterfully taught by Doctor Michael S. Heiser. In the ancient Near Eastern mindset, the night was not just a time for rest; the night was the domain of chaos. The darkness was considered the primary operating hour for the rebel spiritual principalities—the fallen elohim who ruled over the disinherited nations. The pagan world lived in constant, paralyzing terror of the night, believing that evil spirits and demonic forces prowled the earth when the sun went down, seeking to undo the order of creation. But inside the house of the Lord, the darkness is completely neutralized. The temple watchmen are not cowering in fear; they are standing on duty as royal guardians of the cosmic gateway. The temple is the earthly embassy of the Supreme Commander of the heavenly armies. By keeping the lights burning and the praises rising through the midnight watches, these priests are actively enforcing the spiritual borders of God's domain. They are asserting Yahweh's absolute sovereignty over the night, demonstrating to the unseen, rebellious realm that the true King never slumbers, and His fortress is never undefended. The departing pilgrims instruct these guardians exactly how to execute their spiritual defense in verse two: “Lift your hands in holiness, and praise the Lord.” The lifting of the hands is the ancient, universal posture of complete surrender, intense appeal, and open-hearted adoration. The priests are told to lift their hands “in holiness”—or, as other translations render it, “toward the sanctuary.” They are aiming their worship directly at the Holy of Holies, where the Ark of the Covenant rests beneath the wings of the cherubim. By raising their hands in the dark, the watchmen are acting as human lightning rods, drawing the supernatural sanctity and the protective power of the heavenly throne room straight down into the earthly realm, creating a continuous barrier of holy light that keeps the forces of chaos at bay. The second segment is: The Return Blessing from the Creator of the Cosmos Psalm One Hundred Thirty-Four: verse three. May the Lord, who made heaven and earth, bless you from Zion. In the final sentence of the entire Songs of Ascents collection, the direction of the voice shifts. The temple watchmen, standing on the high, illuminated battlements of the sanctuary, hear the parting shout of the pilgrims. They look out into the darkness at the departing travelers, raise their own holy hands over the crowd, and speak a majestic, reciprocal blessing back down upon them: “May the Lord, who made heaven and earth, bless you from Zion.” This closing benediction is a masterpiece of covenant theology and cosmic polemics. Notice the specific, dual title given to Yahweh: “the Lord, who made heaven and earth.” In the Deuteronomy chapter thirty-two worldview, the surrounding pagan nations believed that the universe was carved up into separate, localized jurisdictions. The gods of Babylon claimed the rivers; the gods of Egypt claimed the Nile; and the gods of Philistia claimed the coastal plains. These rebel spirits asserted that their authority was absolute within their own geographic boundaries, and they demanded total compliance from any human who entered their territory. But the priests of Israel shatter that illusion with their final blessing. They remind the departing pilgrims that the God they serve is not a minor, regional spirit of the hills. He is not a localized deity trapped inside the stone walls of Jerusalem. He is the absolute, supreme Architect of the entire macrocosm. He spoke the heavens into existence, and He formed the earth from the void. Therefore, there is no place on the planet that is outside of His jurisdiction. When the pilgrims leave Jerusalem to return to their homes in the distant, compromised corners of the world, they are not leaving the territory of their God. They can walk confidently into any environment, knowing that every square inch of dirt they step upon belongs exclusively to the Maker of heaven and earth. And look at the launching pad of this blessing: “from Zion.” As we have learned on this fifteen-stop mountain climb, Mount Zion is the designated cosmic mountain, the official footprint of Yahweh's heavenly throne room in the human realm. The blessing that the priests pronounce is not a cheap, temporary wish for good luck. It is a massive, supernatural transmission of Shalom—complete, flourishing wholeness and divine favor—cascading down directly from the centralized command center of the universe. The pilgrims are told that this blessing from Zion will follow them down the mountain trail. It will go with them as they navigate the treacherous roads, as they return to their families, and as they face the daily, suffocating hostility of the pagan cultures. Zion's light will go with them into the darkness of their exile. The final step of the ascent is actually the beginning of the descent, where the travelers are sent back out into the world, transformed into living extensions of the cosmic mountain,...

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

THE SINGULARITY PROTOCOL, 40min., USA Directed by Craig McCourry Hong Kong, 1941. Trapped in a bombed-out underground bunker, two soldiers, William Kwok and Charles Evans await death. But their final hour takes a chilling turn when a voice from the future breaks through: Cosmos, an AI Singularity from 2046, reaches back through time to fulfill an enigmatic protocol. As Cosmos shares unsettling truths about history, love, and humanity's place in the universe, William and Charles confront mortality, destiny, and leave behind a stark warning for the future. https://www.instagram.com/banyantreemovies —— Subscribe to the podcast: https://twitter.com/wildsoundpod https://www.instagram.com/wildsoundpod https://www.facebook.com/wildsoundpod

https://novacut.ai/  https://genaimeetup.com/  Anthropic has officially closed a $65 billion Series H at a $965 billion valuation, nearly 2.5x its valuation from just 100 days ago. Meanwhile, funding is flowing across the ecosystem: Frameworks AI at $15B, Baseten at $11B, OpenRouter's $113M Series B, and Cognition AI's $1B Series D. NVIDIA went on an open-source super week with Nemotron 3 Ultra, Cosmos 3, and Nemotron 3.5 ASR. Microsoft dropped 5 new MAI models. Google released Gemma 4 12B, and Anthropic shipped Opus 4.8. On the benchmarks front, DeepSWE crowns GPT-5.5 as the leader in long-horizon coding tasks, while ITBench shows even frontier models struggle with real-world SRE incidents — Claude Opus 4.7 tops out at just 47%. Plus: Cloudflare acquires VoidZero to build the future of AI-native edge development, and Google is paying SpaceX $920M/month for compute. Topics covered: • Anthropic's $65B Series H and path to $1T • Fireworks AI, Baseten, OpenRouter & Cognition funding rounds • Microsoft's 5 new MAI models • NVIDIA's open-source super week (Nemotron, Cosmos 3) • MiniMax M3, Gemma 4 12B, JetBrains Mellum2, Opus 4.8 • DeepSWE benchmark: GPT-5.5 leads long-horizon coding • ITBench: Frontier models under 50% on real SRE tasks • Cloudflare + VoidZero for AI-native edge dev • Google's $920M/month SpaceX compute deal #AI #Anthropic #NVIDIA #OpenAI #AInews #TechNews #LLM     Funding rounds Anthropic formally confirmed the closure of its $65 billion Series H funding round at a post-money valuation of $965 billion. This represents a 2.5-fold increase over its $380 billion Series G valuation from February 2026, adding $585 billion in value in approximately 100 days https://www.anthropic.com/news/series-h  Frameworks AI raising at 15B valuation representing a near fourfold increase from its $4 billion Series C valuation recorded in October 2025 processing 15 trillion tokens daily for major production clients including Cursor, Notion, and Perplexity https://finance.yahoo.com/sectors/technology/articles/fireworks-ai-eyes-15-billion-174609357.html Baseten is raising 1B at 11B valuation annualized revenue, which skyrocketed from $200 million to $600 million over a single quarter https://techstartups.com/2026/05/26/ai-inference-startup-baseten-in-talks-to-raise-1-billion-at-11-billion-valuation/  OpenRouter has secured a $113 million Series B funding OpenRouter has experienced exponential traffic growth, with weekly production throughput expanding fivefold from 5 trillion to 25 trillion tokens over a six-month horizon https://www.businesswire.com/news/home/20260526953416/en/OpenRouter-Raises-%24113-Million-CapitalG-led-Series-B-as-Weekly-Volume-Explodes-to-25T-Tokens  Further up the stack: Cognition AI secured a $1 billion Series D round led by Lux Capital and 8VC https://cognition.ai/blog/series-d   Model Releases MAI models: MAI-Code-1-Flash: A 5-billion active parameter model optimized for ultra-low latency within GitHub Copilot and VS Code. MAI-Image-2.5: A high-fidelity image generation model ranking third on global image evaluation arenas, outperforming competing architectures like Nano Banana Pro. MAI-Transcribe-1.5: A multi-lingual speech processing engine offering fivefold speed improvements across 43 languages. MAI-Voice-2: Natural audio and voice generation across 15 languages, available at a highly competitive price point. Web IQ: A search-grounding API engineered to directly compete with Perplexity. https://microsoft.ai/models/    https://www.peoplematters.in/news/ai-and-emerging-tech/uber-imposes-dollar1500-monthly-ai-spending-limit-on-employees-amid-rising-costs-50073    Nvidia has executed an "Open-Source Super Week," positioning itself as a dominant software and model publisher: Nemotron 3 Ultra (best US open source open weights model but behind china): A massive 550-billion parameter MoE (55 billion active) designed with a 1-million token context window, optimized specifically for high-throughput, cyclical agent loops. It achieved peak throughput rates of 400 tokens per second on day-zero optimized clusters. Cosmos 3: A physical AI world-modeling framework comprising 16-billion Nano and 64-billion Super variants. Built on a Mixture-of-Transformers (MoT) architecture, Cosmos 3 natively binds textual, visual, auditory, and physical kinetic vectors. Nemotron 3.5 ASR: A highly compact 0.6-billion parameter streaming speech recognition model pushing sub-100 millisecond latencies across 40 language locales.   https://www.minimax.io/models/text/m3  MiniMax M3: A 1-million token context model hitting 59.0% on SWE-Bench Pro and 74.2% on MCP Atlas, though noted for high token consumption due to intensive internal self-validation loops.   https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemma-4-12b/  Gemma 4 12B: Google's Apache 2.0 on-device model, which utilizes an encoder-free architecture that projects vision and audio vectors directly into the text-token space, bypassing separate CLIP-style encoders to minimize local memory footprints. https://www.jetbrains.com/mellum/  JetBrains Mellum2: A compact 12-billion parameter MoE (2.5 billion active) engineered for ultra-low latency routing and retrieval-augmented generation (RAG) sub-agents within developer IDEs. Opus 4.8 https://www.anthropic.com/news/claude-opus-4-8    https://www.cnbc.com/2026/06/05/google-to-pay-spacex-920-million-a-month-for-xai-compute-capacity.html      Benchmarks: https://deepswe.d atacurve.ai/blog https://venturebeat.com/technology/deepswe-blows-up-the-ai-coding-leaderboard-crowns-gpt-5-5-and-finds-claude-opus-exploiting-a-benchmark-loophole (GPT 5.5 the winner in long horizon tasks) a highly complex software engineering benchmark focused on original, long-horizon tasks across five distinct programming languages. Comprising 113 chaotic tasks across 91 live, production-grade repositories, DeepSWE forces agents to generate 5.5 times more code and modify an average of 7 separate files per task compared to standard evaluations. On this challenging leaderboard, GPT-5.5 leads with a score of 70%, establishing a significant 16-percentage-point lead over contemporary alternatives I think older benchmarks where models reach ~90% accuracy can be considered saturated. Few percentage points don't give us any good signal.  https://research.ibm.com/publications/developing-ai-agents-for-it-automation-tasks-with-itbench  ITBench-AA, an evaluation framework focusing on live Kubernetes incident response and Site Reliability Engineering (SRE) operations. Comprising 59 live, containerized SRE incident snapshots, the results are remarkably sobering: every frontier model scored under 50% on successful incident resolution, with Claude Opus 4.7 leading at 47% and GPT-5.5 following closely at 46%.   Edge AI announcements: https://www.cloudflare.com/press/press-releases/2026/cloudflare-acquires-voidzero-to-build-the-future-of-the-ai-native-web/  The consolidation of the AI-native developer stack has reached the runtime virtualization layer. Cloudflare recently completed the acquisition of VoidZero, the development group responsible for Vite, Vitest, Rolldown, and Oxc, backing the transaction with a $1 million open-source ecosystem fund. This acquisition is highly strategic; as autonomous agents write an increasing proportion of production software, local development environments, compilation pipelines, and bundlers must be optimized for execution speeds that match agent speeds. Cloudflare's goal is to construct a localized, full-stack edge playground. In this sandbox, AI agents can generate, test, bundle (utilizing the highly parallelized, Rust-based Oxc and Rolldown engines), and deploy entire web applications end-to-end within milliseconds. This architecture completely bypasses traditional local machine container bottlenecks, enabling high-velocity agent loops to execute in a fully sandboxed, web-scale edge runtime.

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

Hello, Beautiful...I'm so grateful you're here with me. Slow down, breathe deeply, and reconnect with the peaceful energy flowing through the universe around you. This mindful meditation for women supports spiritual awareness, mindfulness, emotional healing, and deep relaxation while helping you feel grounded and connected. Perfect for inner peace, self-discovery, stress relief, and spiritual growth. Love,

Listen Ad Free https://www.solgoodmedia.com - Listen to hundreds of audiobooks, thousands of short stories, and ambient sounds all ad free!

Come join with me to the Cosmos engage Yahweh

In this episode, we discuss Anthropic's confidential IPO filing at a $965 billion valuation, shedding light on the competitive landscape against OpenAI and SpaceX. Additionally, we explore Microsoft's new reasoning model, Nvidia's Cosmos 3 for robotics, Intel's price-cutting AI chip, and Strava's new paywall that's reshaping API access in the fitness space.Chapters00:00 Anthropic's IPO Announcement02:00 Microsoft's MAI Thinking 104:01 Nvidia's Cosmos 3 Model05:59 Intel's Crescent Island AI Chip08:00 Strava's API Paywall10:00 Windborne's Weather AI Model Show LinksGet the top 80+ AI Models for $8.99 at AI Box: ⁠⁠https://aibox.aiHow I Grow and Scale My Business with AI: https://www.skool.com/aihustleGet the AI Chat Daily Newsletter: https://www.aichatdaily.com/newsletter

We're announcing AIEWF speakers this week! Take the AI Engineering Survey!Today's guest Ethan first joined us for the LS Paper Club as the lead on NVIDIA Cosmos World Model, but then joined xAI and built Grok Imagine in 3 months:He comes back on Latent Space with some nuclear hot takes: that Video Models primarily get their intelligence from LLMs, not from training on video data, and that the next frontier for truly interactive, realtime, long-horizon world models is to work on LLMs (perhaps Interaction Models as well…)Put it this way: In the near term, the next Sora won't be a better video model, but a video agent.Generative Media may more closely follow the evolution of AI coding which went from focusing on one-shot output performance and cost, to multiturn reasoning and planning models for agents and systems that can plan, edit, test, debug, and submit PRs.At a certain point, coding models got so good that the only significant next step to improve performance was handling the orchestration of these models.Now as the performance of video models increases significantly across realism, consistency, & prompt adherence while becoming more cost efficient, the next evolution of video generation may also be systems that can plan, generate, edit, critique, and iterate across an entire creative task. In this episode, Ethan joins swyx and Vibhu to unpack what it actually takes to build frontier image and video systems: data, VAEs, diffusion transformers, audio-video alignment, inference speedups, and the hidden cost of storing and moving massive video datasets. From building NVIDIA's Cosmos world model to joining xAI as Grok Imagine was being built from zero to one, Ethan He has been at the center of some of the most important work in video generation, multimodal models, and real-time world models.We go deep on Grok Imagine, how a small xAI team shipped its first multimodal video model in three months, why iteration speed matters more than almost anything in model development, and why many of the biggest gains come from fixing tiny bugs in data and training pipelines. Flipbook: The future of VideomaxxingVideo agents are almost a sure bet to be the trend in the coming year. We end with a glance at what's beyond video agents:Flipbook caused a minor sensation this year when it was released, but most treat it as a fun demo. Ethan takes it very seriously — with the speed and cost of inference coming down every year, the future of custom video JIT UI is closer than you think. We talked about why videogen models may become the front end of AI, how generative UI could replace traditional HTML/CSS, why world models need to be real-time, interactive, and long-horizon, and why the future of video generation may depend more on language models and agents than on diffusion alone.We discuss:* Why fast iteration mattered more than meetings* Why small training bugs can drive huge model quality gains* Why coding models may make compute the bottleneck again* How image and video models are trained with synthetic captions* The role of VAEs and latent space in frontier video models* Why image models are the foundation for video models* The tradeoff between temporal compression and real-time interactivity* Flipbook, Neural OS, and the future of generative UI* Why future interfaces may go from user intent to pixels* The hidden cost of training video models: storage, egress, and GPU hours* How step distillation and consistency models (like OpenAI sCM) makes video inference orders of magnitude faster* Grok Imagine 0.9 and large-scale audio-video generation* Why audio-video alignment is harder than text-video alignment* Ethan's definition of world models* Reference-to-video, video extension, and long-context video generation* Why xAI's research communication undersells Grok Imagine* How xAI culture shaped the speed of development* AI watermarking, SynthID, and detecting generated media* Why prompt rewriting matters for video models* Grok Imagine Agent and the rise of video agents* Why language models may unlock better video generation* Robotics, physical AI, and embodied world models* Why Ethan left xAI and shifted focus toward LLMs* Self-managed context, memory, and the next frontier for language modelsEthan He* LinkedIn: https://www.linkedin.com/in/ethanhe42* X: https://x.com/EthanHe_42Timestamps00:00:00 Introduction00:01:25 From NVIDIA Cosmos to xAI00:03:24 Building Grok Imagine from Zero to One00:10:07 How Image and Video Models Are Trained00:18:53 Video Compression, VAEs, and Real-Time Tradeoffs00:22:10 Generative UI, Flipbook, and Neural OS00:32:10 The Cost of Training Large Video Models00:37:04 Distillation, GANs, and Fast Video Inference00:41:21 Audio-Video Generation and Grok Imagine 0.900:48:34 What Makes a World Model?00:55:51 Reference Videos, Long Context, and Video Memory01:00:11 xAI Culture, Research, and First-Principles Building01:09:45 AI Safety, Watermarking, and Prompt Rewriting01:13:10 Video Agents and AI-Assisted Creation01:27:32 Why Language Models Unlock Better Video01:31:15 Robotics, Physical AI, and Embodied World Models01:32:38 Why Ethan Left xAI01:34:16 Self-Managed Context and the Future of LLMs01:38:43 Ethan's Career Path and Closing ThoughtsTranscriptIntroduction: Ethan He, Latent Space, and the Path to xAISwyx [00:00:00]: We're here in the studio with Ethan He, most recently of xAI. Welcome.Ethan [00:00:10]: Thank you. Glad being here.Swyx [00:00:11]: We're also here with Vibhu. you were first coming to us or joining the latent space world because you were working on Kosmos at NVIDIA, and you did a paper. We loved it. you presented it as well, so thank you for doing that.Ethan [00:00:23]: I've actually, I also presented the MoEs twice at latent space.Swyx [00:00:29]: How did you actually hear about us? Did we reach out to you? Is that how it worked?Ethan [00:00:33]: No, actually, I-- the community. Like I realized, oh, there is this online community that people talk about AI and also learn from each other through papers every week through the Paperclip. It's very nice.Ethan [00:00:49]: I learned a lot.Swyx [00:00:49]: I think three years stop. We haven't stopped even on Christmas and New Years. many weeks I want to stop but it keeps going.Vibhu [00:00:58]: No, that was good. I think you had posted that you worked on a paper, and I was “Oh, very cool. We have Paperclip. Present then.”Vibhu [00:01:04]: But I might have reached out to you after.Swyx [00:01:05]: you-- because it's an amateur club, right?Swyx [00:01:08]: so it's very unusual and but we have sometimes paper authors come by and actually explain the paper. Today we just did, the poolside paper, which was apparently very good.Vibhu [00:01:18]: Came out yesterday.Vibhu [00:01:19]: pretty interesting, right? Fully open. They talk about everything, systems. So it's a good one. We'll, we'll recommend people to read it.Swyx [00:01:25]: Bring us up to speed on your transition to xAI, ‘cause I actually don't even know when you joined. just like tell the, tell the story about the sort of transition.From NVIDIA Cosmos to xAI: Scaling Video and World ModelsEthan [00:01:34]: Before xAI, I was working on Kosmos world model as in-- at NVIDIA. So Kosmos is, it's a giant video foundation models that can-- that aims to simulate the world and for-- it serves as a foundation of-- for all of the roboticists to build on top of. There, once I built the Kosmos one, I realized as this thing also has a scaling law similar to language model, we need to scale up the video models further. that's, that's why I realized I need to move to somewhere with much more compute resources. That's how ISwyx [00:02:13]: Than NVIDIA?Vibhu [00:02:14]: The GPU rich came themselves.Vibhu [00:02:19]: And timeline-wise, when was Kosmo? It was pretty early, right? It was open world model, open paper, everything.Ethan [00:02:25]: It was end of twenty-four.Vibhu [00:02:28]: End of twenty-four.Ethan [00:02:30]: Then at mid twenty-five, I moved to xAI. At that time-- I joined about the time when xAI was about to build video models and in multi-model models. There were no infra, no data, and no model, and it just-- as a few engineers, we built it in three months and released the first model, Grok Imagine zero point nine.Ethan [00:02:55]: And since then, I keep working on video models and move more from training and to post-training of the video models. For example, like a reference to videos, kind of like the cameo feature and, video extensions. And, before I left, I worked on a world model, leading a small team to focus on the real-time long horizon video generation.Building Grok Imagine From Scratch in Three MonthsSwyx [00:03:24]: Can you give like a rough roadmap of okay, you're on a brand-new team. Grok previously was only text, or they partnered with BFL for their image gen stuff. What do you-- what are the building blocks, right? You have compute, data you can procure somewhere. Like just what are like the sequence of things that people should think about when you're setting up a new team?Vibhu [00:03:43]: actually even deeper, not just data you can procure. You guys had to go through getting the data too, right? So you shipped it pretty fast, but yeahSwyx [00:03:51]: three months is likeVibhu [00:03:52]: From everythingSwyx [00:03:52]: actually like very surprisingly fast.Ethan [00:03:55]: One thing I say like thanks to my experience at NVIDIA, ‘cause first time when we were building Kosmos together, we built it, for about a year. So this is like the second time I do it. Roughly have an idea, what to do. I say the most important thing is the talent. Everyone were very strong and clever, very close with each other towards a common goal. So that speed up things a lot. So you reduce the communication bandwidth among people, and everyone can work towards the same goal. It's, it's like every day there's not that much meetings on the calendar, like maybe like a, like a sync a day, and after that it's, it's just all building. It was pretty fun at that time.Ethan [00:04:47]: And another thing is that xAI has very strong foundations of like data inference, model inference, and the supporting there can help the model develop a lot. When I look at, training models, I don't so actually the top important thing is like how many, how many iterations can you do, per day? and the more iteration can you do, you can, you can train the model much faster. So if you have very strong infra and you have a lot of compute, you can, you can train these models in very short period of time. That can give you a much larger buffer to, for errors, and it also gives you the opportunity to spot more bugs.Iteration Speed, Compute, and Debugging Model PipelinesSwyx [00:05:46]: What is an iteration? Is it like a few hundred steps or what are youEthan [00:05:50]: Let's say just the train-training the model, like from acquire new data and maybe design new algorithms and train a new model, maybe at smaller scale orSwyx [00:06:01]: So cycle time for like any hyperparam that you're searching.Ethan [00:06:04]: Cycle time and tune to like eval this model. Is this model better than my previous iteration?Ethan [00:06:11]: SoSwyx [00:06:11]: So it's like before you, someone had already set this up that you can iterate very quickly.Ethan [00:06:15]: I think the foundation there is extremely good forDeveloping and research models.Ethan [00:06:23]: And often I find is it-- this is kind of boring, but like a lot of the improvements does not come from new algorithms. It comes from finding small bugs here and there in the data pipeline, in the, in the model training pipeline. Those give, those give the biggest boost to the model quality.Vibhu [00:06:46]: It's interesting, right? So you say it's like small team, less communication bandwidth, but also a lot of quality is like find little bugs. It seems counterintuitive, right? You have a lot of people, you can iron out more of those, but it's interesting to see the other side, right?Swyx [00:07:00]: I also wonder, have you-- do you try using LLMs to look for bugs? I don't know.Ethan [00:07:05]: I remember at that time it was mid two thousand and twenty-five, so it's the coding model wasn't quite there yet. I remem- I remember like December two thousand and twenty-five, it was extremely good. Yeah, I've been, I've been using it at that time. It's, it's helpful. sometimes it produce codes that are kind of difficult to maintain, even though like the first time it built something extremely fast. But it gave the, like a spaghetti code, thousands of lines that I couldn't maintain, and the LLM itself couldn't figure out what's, what's wrong and how to improve on top of it. But now I find it much better. Yeah, I want to bring up another point here is now coding models are much more efficient and can help us implement stuff much faster. Compute might become a bottleneck again because previously, like if you want to train a new model, say you want to generate new synthetic data and then or write a new algorithm, it might take a few weeks. And during that period of time, you don't-- you might not have experiments to run. But now you can build that thing within a few hours, then you can immediately train a model.Ethan [00:08:24]: Now you have to have enough compute to try all of the ideas. So compute might be the bottleneck of iterating speed again.Swyx [00:08:36]: yeah, I actually, honestly, I think it's like kind of a stressful job because you're “Well, I should be trying everything, and if I'm not, then I'm not doing my job well.”Vibhu [00:08:48]: there's also the stress of you're eating thousands of GPUs per hour, which is very expensive and, compute can go to other researchers.Swyx [00:08:56]: You got the daddy Elon toVibhu [00:08:57]: You got daddy Elon.Ethan [00:08:59]: It wasVibhu [00:09:00]: But there's still finite amount of compute, like you want to use it, you want to use it well, you want more of it.Ethan [00:09:06]: That was quite stressful indeed. Yeah, I think one thing is the-- with coding models now, like a lot of these jobs can be automated, which is much better. A second, it's a, it's a marathon, so you got to maintain good health and, a regular schedule.Vibhu [00:09:28]: It's, it's hard to hear that when you shift from zero to nothing in two months.Swyx [00:09:32]: and, I think obviously the culture at xAI is very famously, people work very hard. one thing I did want to dive into, in our-- in the notes that you, that you sent ahead of time, you had specific comments about the cost of Video Gen training. presumably this is on the Colossus-1, right? the two hundred megawatt cluster. Any whatever you want to just share on that.Vibhu [00:09:54]: I think there's, there's three things we're talking about, right? So there's Video Gen, there's also the Image Gen model that you put out. Do you want to like complete the, okay, so zero to one, you have a few months. Just what are the stages of create Image Gen model?Swyx [00:10:06]: Oh, yeah, maybe I got distracted.How Image and Video Models Are Trained: Synthetic Captions, Tokenizers, and VAEsVibhu [00:10:07]: Sorry. and then, from there's Video Gen, there's Audio Gen. Would love to get into those next. But what is that first few months like? So small team, a lot of bugs, iterations, but what does it look like? Do we take something off the shelf? Do we just get data compute? What's, what's the few months like? How do you go to state-art Image Gen model? How do you just start?Ethan [00:10:28]: I cannot comment specifically how xAI did, but it's, it's a quite standard process. I can draw some, examples from Cosmos. So mainly it's building a video model, you actually need to build a image model first. And building these two models, the data you need is a hundred percent synthetic pair of language and image or language to video. Because on the, on the internet, actually, the videos don't naturally associate with text. So you can say, oh, like on YouTube, you have the title and you have the description and the commentsSwyx [00:11:11]: TitleEthan [00:11:11]: of a video, but usually they're not relevant to the video itself. And say maybe like the video is a natural scene of mountains or something, and the title is, I'm so happy today.Ethan [00:11:26]: So they have they have no correlation at all. So the first step is to, you have to generate synthetic pair of language with the videos. So you gather videos from the internet, and you use a VLM to caption the videos. So that part, here's a question, like how do you, how do you gather VLM to begin with? So if there's noSwyx [00:11:55]: You, so you fuse the model, right? LikeEthan [00:11:57]: Say if there's no like VLM exists, like how do you generate the text to the beginning, right? It's, it's impossible.Swyx [00:12:04]: I see.Ethan [00:12:05]: In the beginning, it's like you ask human to describe the video as detailed as possible.For example, you ask them to describe everything, like all objects, all characters, and all interaction and dialogues in the, in the videos. So that's in the protocol of Cosmos labeling. We require the objective we give to the labelers was that you have to describe the video as detailed as possible, such that a blind person hears a blob of text can reconstruct what the video is like from their head.Swyx [00:12:43]: Video or image? You're talking about images.Ethan [00:12:44]: Video or image, either one of them.Vibhu [00:12:47]: This was pretty common when we went from clip and DALL-E, right?Vibhu [00:12:51]: It's all training on really detailed captioning of images. So same is applied to video, but insteadEthan [00:12:57]: same appliedVibhu [00:12:57]: of using multimodal model to pass in video images and write rich descriptions, you can alsoSwyx [00:13:04]: I think there's this traditional perspective of supervised, or, very highly human curated thing. I feel like there's a unlock with unsupervised, right? Where like you have enough to bootstrap that you can just throw common corpus on it or, whatever. like unsupervised vision and language pairing, right? Like where you just have, interspersed image and text and it just learns. To me, that is the VLM breakthrough that is different from the clip, different from the LM era.Ethan [00:13:36]: It's interesting to see that you kind of need both data.Ethan [00:13:41]: For example, for theSwyx [00:13:41]: You need it to bootstrap it up. YeahEthan [00:13:43]: for the generative model training, there's also usually like a small percentage of unlabeled data. So the model is instructed to generate a video without any text instruction. That can also help the model generalize. So after this stage of generative synthetic pair, so, one important common step is to train a compressor or a tokenizer of the image or videos. So because, if you train-- If you can technically, theoretically train image or video models on pure pixels, but the problem is that the, it's, it's a lot of tokens. So like one image, it's, a thousand by a thousand, it's like one million tokens, one million pixels. It's impossible to train transformer on that. So it's, you need to train a tokenizer, which can go from image to latent space and latent space back to image.Swyx [00:14:45]: That's why we named the podcast.Swyx [00:14:48]: But, basically, you're talking about vocabulary science.Ethan [00:14:50]: so vocab.Swyx [00:14:51]: And so, what is, what is imp-- like a million is impossible?Ethan [00:14:54]: In generative models, the vocab is continuous. It's a continuous space. We can think about like you map an image to a vector. It's a, it's a fixed length vector. It's sixteen or forty-eight, something like that. And then you map that vector back to the image space. And the mapping is, has-- The mapping is patch-based. So you say you haveEthan [00:15:22]: a sixteen by sixteen patch and you match, you map that patch of pixels into this latent space.Swyx [00:15:29]: We've covered thisVibhu [00:15:30]: This is like the vision transformersSwyx [00:15:32]: VAEs,Ethan [00:15:33]: VAEs.Vibhu [00:15:34]: You basically compress your input, you do your generation, you're reasoning all that generation in smaller dimension, and then you project back out.Swyx [00:15:43]: VAE is a form compression, but I think the for me, the patching thing is from VIT, right?Ethan [00:15:48]: You can make those.Swyx [00:15:49]: Literally the, yeah, the paper is titled like sixteen by sixteen is all you need. something like that. and then I think also, people make a lot of comparisons with this kind of patching with convolutions.Swyx [00:16:02]: Which is you're, you're kind of re- reconstructing the old paradigm with the new.Ethan [00:16:05]: Actually, in VAEs, there are, there are both convolution networks and transformers. You can actually do both.Ethan [00:16:14]: After this VAE, so what you've got is you've got latent space tokens and you've got the language tokens. So now the training of the diffusion transformer, usually generative models use diffusion transformers. It is actually quite standard. It's, it's very similar to how you train a language transformer models. It's not that much difference. It's just the tokens, the visual tokens in, visual tokens out. The only difference is there's a denoising process. So you train the model to unmask some of the noise. So you add, you add random noise to the visual tokens, and then you train the model to remove those noise to generate the clean tokens. Any inference, the model can iteratively remove noise from a hundred percent noise.Swyx [00:17:12]: And then there's also, to speed things along on the tech tree of diffusion, there's CFG, and then there's, there's also, latent diffusion that, there's, there's someone in there. I think, somewhere along the line, obviously, like stability and all these other guys, pioneered a lot of this, architecture. I don't know if you want to get into that or just, or do the video side up to you.Bootstrapping Video from Image Models and Temporal CompressionEthan [00:17:37]: After you train such model, such image model, the reason it's a, it's a foundation for video models is that image models are cheaper to train, and they have much denser connection between language and text. So, sorry, language and images. For example, you train a billion, you train on a billion images, and there's a mapping from the text to the image. And the cost to train the same, like the, a billion, a billion text to a billion videos, that's much more expensive because videosNaturally have more tokens than images. Because the diffusion models, their understanding of, language purely come from this mapping. So if you don't have enough mapping, so if you only train on like a ten million videos or something, there-- you might not see enough language tokens in your training, so your model does not understand human intention enough. So that's why you really-- you train-- you first train this image diffusion models, and then you bootstrap the video model from there.Swyx [00:18:53]: One thing I did want to ask, because I-- actually, I think you're, you're the first per-- video model person I've ever talked to, I think. we've, we've like talked to Luma and all those folks. There's all these tricks in video compression where basically frame by frame there's not that much difference, so actually you don't have to regenerate or save the whole frame, right? but I think MP4 compression or something else like that.Swyx [00:19:16]: is it tempting to use that? Or as far as I can tell, everyone just treats it as, “No, we would just generate every frame.” Is that roughly the state-art?Ethan [00:19:27]: There are a few different approaches. Let's say first, like you want to just directly use MP4 compression and use that as the tokens for the transformers to train, right? So people actually have tried that, but the main challenge is the latent space for the MP4 tokens were not, were not very comprehensible for the models. It's, it's extremely hard to train on that. And there's aEthan [00:20:01]: So that's why they created VAEs, which creates more continuous, latent space, so the models can understand that latent space and learn from it much easier. Even within the VAEs, there are different difficulties of the latent space. So you can imagine something the simplest, the most naive VAE is like you have an image, and you just shuffle all of the images into a, into a vector. So you don't need to train any VAEs, right? But that latent space is extremely hard for models to train on top of. That's why there are some debate on like how do you compress the tokens. So you mentioned like you can compress frame by frame. Also, you can compress, the temporal dimension.Ethan [00:20:52]: The difference is if you compress the temporal dimension, you get a much higher compression rate. Because there's temporal redundancy between frames, because, this frame and the last frame, likely they are mostly similar, so there's only some small difference. for example, I think in 12.1 VAE, they have like a eight by eight by four compression rate. So the four temporal tokens are compressed into one tokens. That can save a lot of, save a lot of the context length. If you do it frame by frame, you have to do maybe like eight by eight by one. Your context length will be four times larger. That being said, the benefit of the frame-- per frame compression, we might come back to this later, is, real-timeness and interactivity. ‘Cause if you, if you strain the output of the model, frame by frame, you can-- the model can respond to any user request immediately. So if you have like a temporal four compression, four times compression, thenSwyx [00:22:06]: It might be laggyEthan [00:22:07]: there's a lag there in nature.Swyx [00:22:10]: So you're very pilled on this. let's just go ahead and bring it up ‘cause we have the visual prepared anyway. There's some frontier applications of real-time video gen. So Flipbook is one of the examples that went viral recently, right? What is Flipbook?Real-Time Generative UI: Flipbook, Neural OS, and Diffusion Front EndsEthan [00:22:23]: Flipbook is kind of like a web brow- web browser. You can see like it has the web bro- browser UI on top. The difference is all of the UIs are generated by generative image model in real time, and anything here are fake. But you can, you can explore inside this wor- this imaginary world. Say like we-- here we have engineering the Great Pyramid. Like the model generates this for us to understand how it works, and if we want to navigate around and understand further, we can click on some of the, some of the description here, and the model will generate a new page, new subpage describing the details we want to know about.Swyx [00:23:14]: So it's basically kind of we're playing a video, but it's pausing for our next interaction, and then it just plays the next thing based on our interaction.Swyx [00:23:23]: Which is kind of cool.Vibhu [00:23:25]: and you kind of decide your story. So this was, how do you make a pyramid? levering technique seemed interesting, right? It shows how do you take Okay, I want to know what is thisSwyx [00:23:35]: The demo, the demo tweet had more animation between frames.Vibhu [00:23:38]: I think it's just skipping,Swyx [00:23:39]: Oh, it's just skipping a lot of frames.Ethan [00:23:40]: they also have a video modeVibhu [00:23:42]: It takes a lot. There's a lot of peopleEthan [00:23:42]: but, a lot of people are using it.Ethan [00:23:45]: So it's not available.Vibhu [00:23:46]: There's a live video stream. We can try,Swyx [00:23:50]: So this is an example of the kind of future that you see at the extreme. We don't-- we're obviously not in it today.Swyx [00:23:56]: But in a world where inference is completely free this is better than generating code and text?Ethan [00:24:02]: So this is, this is a final state of where Viva will be at for word model, I think. Imagine internet doesn't exist, and then you type in google.com. Like what should, what should, what should a model show you?the model can imagine something, and this is what the model imagine. And these web pages, they completely do not exist. So I think as the inference costs come down, we are going to have generative UI for everything. If you think about how the coding model works, so they write code for a web page, and they render the code might be con- converted into binary, and the binary render the pixels on the screen. So we in machine learning, every time we have some breakthrough, obviously it's, it's more intuit. So why don't we have like user instruction to the pixel directly? So the generative UI will be user intention to the pixels directly. And say like even if I want email, let's say everyone have the same interface, but I want, I want it slightly different. I want the email to show to me like a TikTok, so I can swipe left and right for the emails. And or maybe you want something else. We can have completely different things. Or like I have I'm looking at, Instagram stories, and I don't like the Like button. I always may click it. And, generative UI resolved it. So it's going to be a revolutionary replacement of the interface. So in the future, we might have much more powerfulEthan [00:25:50]: LLMs and coding models running behind the scene. And in the, in the front-end, the diffusion model will actually be the front-end to show stuff to you. That's how I imagine it.Swyx [00:26:02]: Diffusion front-end, deterministic back-end.Swyx [00:26:04]: Something like that. I find that very expensive, but,Vibhu [00:26:08]: I find it interesting you called LLMs writing code on the back end deterministic, but okay.Swyx [00:26:14]: you write it onceVibhu [00:26:15]: Compare it toSwyx [00:26:16]: And then you execute.Ethan [00:26:17]: If you think about the cost, say, let's say H100 costs $1 per hour, and if you use this eight hours a day and thirty days, so, every month you're paying this two forty, you'll actually not wanna pay for that. That's even more expensive than Cloud Code Max. But if you think about the compute costs come down like two times every year, and I think the future will likely arrive like within few years.Vibhu [00:26:49]: It's everything, right? compute cost comes down, compute gets faster, model gets smarterEthan [00:26:54]: More efficientVibhu [00:26:54]: model gets smaller.Swyx [00:26:55]: I don't know why you say two times, ‘cause I think it's like 100 times. In language models, it is roughly one hundred to a thousand times every twelve to eighteen months, for the same given level of LMSys, ELO.Vibhu [00:27:08]: That's a net of everything, right? That's model performance alongside compute. So different than just compute costs come down. But, a very interesting future.Swyx [00:27:19]: So the web designers will have to shout out that accessibility is an issue, right? how do you deal with screen readers or whatever. But yes, this is higher bandwidth storytelling than anything you can possibly generate with code, right? So I think that's the rough idea.Ethan [00:27:34]: And I'd like to add a little bit that so human naturally have the maximum bandwidth when we are looking at things, look at videos, and we also have maximum output bandwidth when we are talking. So in the future, it might be something like we talk to AI models, and the AI model responds back with a generative UI. So that would be the maximum input and output bandwidth to interact with AI models before neural link happens.Vibhu [00:28:06]: And it's also very custom, right? Some people are very visual, some people are not as visual, right? They prefer the text. But the best thing about generative UI, right, it can also be text.Swyx [00:28:17]: There's another project that we wanted to highlight, which is the Neural OS. Kinda similar idea, but here you're literally operating, simulating an operating system with a video model.Swyx [00:28:27]: and you can play Doom, you can do Firefox. I find this like mildly less impressive, obviously, because it's an OS that I can run.Swyx [00:28:37]: But here everything is imagined.Vibhu [00:28:40]: I was, used to the Command+W to close the Firefox tab. It didn't crash. That's why I saidSwyx [00:28:45]: It's too immersive.Vibhu [00:28:46]: It's, it's too immersive for me.Swyx [00:28:47]: Too immersive.Vibhu [00:28:48]: I wanted to close the tab.Vibhu [00:28:49]: But yes, I can play generated diffusion.Swyx [00:28:51]: this is shockingly fast.Swyx [00:28:54]: Because I remember there was a demo about like maybe one to two years ago. Someone tried to do the first-person shooter with a image model. There was no consistency. It was very slow. But here it looks like realistically it's-- this is Doom.Vibhu [00:29:07]: I think there's two sides to that, right? There's okay, what is running a game? The heavy part of it is actually the game engine, all the lighting, all that stuff, the graphics. This is just kind of video, right? Like we've solved consistency. This is still, it looks like a few years old image generation. There's some temporal consistency, but it's, it's kind of just images stitched together as frame video. But it's a good visual representation to pi- to picture the future you wanna see, right? that's, that's what I see in these more so.Ethan [00:29:38]: This reminds me of how the video models gets better and better. So Neural OS is kinda if you just look at it feels like it's just a crappy version of the, like the Windows we could have, right? And, but the difference is, so the model, this model is overfitted on the existing operating systems. It can generate nothing different than that. But it's actually also similar to video models. So when we are training these video model, image model, we train them on internet. There's no imaginary supernatural stuff on the internet. But once we train this model, you can prompt the model to generate something supernatural that have never existed in the data set. So if you train your Neural OS or neural computer on the standard screen recordings on the entire internet. The model can imagine completely new interface to interact with the computer.Swyx [00:30:43]: This is one of those things that is magical to me. usually generalizing out of distribution is bad, but somehow we have learned some kind of internal world model that you say, this plus, but it looks like rainbows and butterflies, it'll do it and it will kind of make sense.Swyx [00:31:03]: So yeah, that's kind of cool. Yeah, I don't know if there's any comment more on there. I do, I do wanted to, I did wanted to touch a little bit more on the model architecture stuff, which I think you were getting. It's, really fascinating. We don't get a chance to talk about this enough. So one of the papers that we covered, we've covered every annual, segment anything release. and I don't know if you follow-- you're a computer vision guy, so youEthan [00:31:26]: I knowSwyx [00:31:27]: . So they did memory attention, which is kind of interesting. And I always think, anything where you can, across the temporal dimension, keep some consistency, I think it's, very fascinating, and I don't know if Basically, does that-- the CV side bleeding into video gen side, I think is underexplored, right? we talk about it for labeling, but actually you can borrow the architecture itself.Ethan [00:31:50]: There's, there's also complete different approaches, right? you brought up the term world model, so we went from video model to world model. There is diffusion, but there's also other approaches that people are doing. So maybe we get into those after as well,?Swyx [00:32:03]: He has a whole definition of world models and stuff. I feel like we threw a lot at you. Whatever you want to comment on.Why Video Models Are Expensive: Storage, I/O, and Training ScaleEthan [00:32:10]: I think one thing that we should actually comment back on is okay, so we were talking about the steps to train image gen to video model. One thing we don't see as much of is okay, you brought up the delta in training data, right? SoEthan [00:32:24]: you won't have as much a video model might not generalize, but what is the cost of training a large video model? So we know for LLMs roughly, okay, even like the poolside thing that came out today, right? It's a Gemma level model trained on roughly forty trillion tokens at this many H200s over this much time, right? You can see what is the exact cost of that. So how many GPU hours over how much H200 costs? So how do we do the back-end math of, same thing for video models, image models. How do you, how do you kind of break that down? I can share some back-envelope calculation. So surprisingly, video models is-- the cost is very-- is comparable to language models and obviously the largest scale is language model, maybe like a medium scale to language models. I said just storing the videos alone, it costs a lot. You can, you can maybe look up on AWS or something.Ethan [00:33:20]: You really, say if you have a billion videos and let's say, let's just say like each video, like five megabyte, then you need five petabyte to just store those videos. And also remember we talk about you use a VAE to compress the videos, and you also need to store, typically you need to store those continuous feature, in-- also in your storage. That's also comparable size with the videos themselves. So just storing these videos and the features is tens of petabytes alone. And,Swyx [00:33:58]: I just, I just looked up the calculation. Five petabytes on S3 Standard is one hundred K per month.Ethan [00:34:05]: AndSwyx [00:34:05]: It's comparableEthan [00:34:05]: and you needSwyx [00:34:06]: AndEthan [00:34:06]: And then like tens of petabytes, two hundred K. And even more expensive is you have the ingress and egress.Swyx [00:34:13]: Oh, yeah.Ethan [00:34:14]: Like you-- through the internet. You have to just to download those videos, I believe it's, it's more expensive on AWS than just storing those videos.Swyx [00:34:25]: Storing, yeah.Ethan [00:34:25]: And each training runs, you probably need to pull them once. If you train multiple times, it's, it's even more than that. So it's like just storing the network, those costs is just, it would be a few, a few millions per month to just storing everything, not to mention the GPU cost.Ethan [00:34:45]: AndSwyx [00:34:45]: my side tangent, the compute rental, like GPU rental is very efficient. There's one side, okay, you can be XAI and build your data center. Should we not just build our, storage compute as well? LikeEthan [00:34:57]: Of courseSwyx [00:34:57]: cloud cost compared to just,Ethan [00:34:59]: You save so muchSwyx [00:35:00]: store. Yeah, exactly.Swyx [00:35:01]: Especially with like egress and stuff. So.Ethan [00:35:04]: That's a good idea, but it also comes to-- there are some of its own challenges.Swyx [00:35:09]: Of course, of course.Ethan [00:35:10]: like people who build the GPU data centers, they might not expect this much, storage. And yeah, people build storage, typically they just build it somewhere with just CPUs.Swyx [00:35:23]: I just looked it up. Five-- AWS only charges for egress, not ingress. Tier five for five petabytes is two hundred and thirty K.Ethan [00:35:32]: Even more expensive than the storage.Swyx [00:35:34]: But storing is per month, right? You check in, then you cannot check out. so it's so cool. It's okay. So there's that side.Ethan [00:35:41]: So the TLDR, my backhand mathSwyx [00:35:42]: Data is larger than you think. Yes.Ethan [00:35:44]: my backhand math of GPU hours times GPU cost is also very much, I'm missing some storage.Swyx [00:35:49]: You're also-- you're basically like also more IO bound than normal training.Swyx [00:35:55]: Yes. ‘Cause like data loading, so caching everything, it becomes super important.Ethan [00:36:00]: So in Cosmos, we did a lot of optimizations to make it not IO bound. So, speaking of the training, actually training the model, the GPU cost, if you look up like the open source model, how big these video models are, I think like LTX has nineteen B parameters. That's a dense model. And people are also exploring, MoEs, so it might be twenty B active and, like a hun- hundreds B, total. So that's, that's even-- that's similar size as medium-sized LLM models. And if you, if you look at number of tokens-Uh, we disclose that in Cosmos. It's also like tens of trillions of tokens on the visual tokens. So putting this together, the cost of, training these video models, it's actually comparable with LLMs. Not to mention, the infra is slightly different from LLM, so it might be less efficient to train these models.Inference Speedups: Step Distillation, Consistency Models, and GANsSwyx [00:37:04]: Do you get the benefits of traditional diffusion speed-up? So for, images, there's LCM, LoRAs for, fine-tuning. There's, there's a lot of stuff that's beenEthan [00:37:15]: Flow matching.Swyx [00:37:16]: there's flow matching. There's a lot of stuff that's been done. there's some overlap that applies to diffusion on the inference side and stuff or?Ethan [00:37:23]: so the difference-- the inference side is a completely different story.Ethan [00:37:28]: I think for the training side, it might be a little bit hard to reduce that cost. And for the inference side, the biggest gain is from the distillation of these models. You can-- It's called step distillation, slightly different from knowledge distillation in LLMs. So you-- Typically, for flow matching models, you need like 100 steps or something. Like a distortion model even need even more, like 1,000 steps to generate a good image or video. A step distillation is try to learn to generate fewer step from the model itself. It's kind of like now we-- you use the full model to generate in 100 steps, and then you take a model that only generate 10 steps and let that model to learn from the perfect one.Ethan [00:38:25]: why this workSwyx [00:38:27]: Strong to weak seemingly.Ethan [00:38:28]: It is. It's kind ofSwyx [00:38:29]: DistillationEthan [00:38:29]: kind of like strong to weak. the-- from the modeling perspective, the strong model, the teacher model is trying to model the image and videos of inter-internet, and that distribution is extremely complex. But the step distilled model is just trying to learn from the teacher. The teacher is a model, and the size is fixed, as the distribution is much simpler than the whole internet. That's the intuition I have why step distillation can work. So usually these models serve in productions, they only run in a few steps. In Cosmos, I believe we have, we have like four step and eight steps. If you do some simpler task, image-image translation, it can even run in fewer step, like one step in Cosmos Transfer.Swyx [00:39:22]: I think this is the same intuition that guides a lot of the consistency model work. I sent you a link for, SCM. I don't know if you covered that. To me, that was actually one of, the most impressive papers I've ever seen from OpenAI.Swyx [00:39:34]: That this is the unifying grand concept of consistency models. I don't know if you have any comments on this.Ethan [00:39:41]: So there are, there are a few different approaches,Swyx [00:39:46]: Oh, yeah. Here it is.Swyx [00:39:47]: Two steps versus twenty or 100 steps, whatever. It's already done.Ethan [00:39:52]: So there are, there are a few different approaches, for example, consistency model, and there are also Actually, we shouldn't forget GAN. So GAN, actually, that was, that was the OG ofSwyx [00:40:05]: OGEthan [00:40:05]: step distillation ‘cause it trained just one step to begin with. So actually, a lot of, uh-- For example, there's a distribution matching distillation which use, which uses GAN, as one of the laws for distillation. It-- GAN just tells you, “Hey, generate an image,” and thenEthan [00:40:31]: it has a discriminator to tell, is this image real or not? So the model, the model just need to learn one of the distribution, not the full distribution. Because in training, the model is asked to reconstruct the ground truth image from the internet, which is extremely hard. And in-- When you're training GAN, it's a step process. It's just a, “Hey, you generate image. Does this image look as real as the image from the internet?” Which is a much simpler task. And, yeah, combining a lot of these approaches together, people typically do that, like consistency model and distribution matching and GAN, and we can get these few step models.Audio-Video Generation and Time AlignmentSwyx [00:41:21]: Then there's one step I wanted to add, which is audio and video.Ethan [00:41:26]: So, Grok Imagine zero point nine, I believe it's, it's a first audio video transmodel deployed at a large scale. SoSwyx [00:41:39]: And that was your first model?Ethan [00:41:40]: that was, Grok Imagine's first model. It's, it's audio video, joint generation. I think the hard part is, the modality alignment, ‘cause before this transmodel, we have, we have text to video alignment. We have this, correspondence between text and video. Typically, most of the VLMs, they understand images and videos. Video's very rare, and they don't understand audio mostly. And if you look at the audio generation on the LLM side, you can talk to them perfectly fine, but if you ask them to sing a song or something, it typically is not very good. Also, they don't have, they don't have music either. The hard part is thatUh, actually audio has two component. It has like a discrete component, a continuous component. The discrete component is like the language.Ethan [00:42:44]: So when we speak, it's just, someSwyx [00:42:47]: It's an ASR issue, yeah.Ethan [00:42:49]: It's, it's text token with some characteristics, I would say.Ethan [00:42:54]: But musicSwyx [00:42:56]: I think the speech guys would disagree with this.Swyx [00:42:57]: Like disfluencies and then,Vibhu [00:43:00]: There's tones you can get angry.Ethan [00:43:01]: Well, I say largely.Ethan [00:43:03]: the mu- but the music is completely different. It's, it's very continuous, and you cannot model them like discrete tokens in language models. this is like the hard part for models is, not to mention we have to align text, video, and audio together.Ethan [00:43:26]: SoVibhu [00:43:26]: How?Ethan [00:43:28]: So significant-- some significant challenges are like-- So first, like we talk about as the VLMs, they cannot understand most of them cannot understand audio.Ethan [00:43:39]: So you have to have some way to do the synthetic data generation for audio. You have to caption the model, and that involve, that involve synthetic data and human data effort a lot. And not just surprisingly, most of the LLMs are very bad at recognizing, like the beat, tone, and the details of the of music. They can, they can give some general prediction of which song is this, but it's very hard to describe the details of the music. like we mentioned in image generation, like you have to describe image as detailed as possible so that someone blind can reconstruct that. So here is like someoneVibhu [00:44:32]: DeafEthan [00:44:32]: someone deaf can reconstruct how the music sounds like without actually listening to it. Maybe you can think of it need to have the-- or they call the script.Vibhu [00:44:49]: Subtitles, yeah.Ethan [00:44:49]: You gotta have all the details of the music, and the dialogue.Vibhu [00:44:55]: So is the challenge there typically stuff like music and audio, or is it just Like is there a baseline? Okay, there's enough data where we can understand, narration, conversation, but there's nuances in audio that's where you hit all the data issues or is it just from stage zero, you just do it all right?Ethan [00:45:15]: So one important thing is like the alignment. So the model, the model has to know like the video and audio, the, uh-- it has to have a time-based alignment, like at which time step the video and the audio token correspond to each other. But we actually don't have this kind of alignment for most of the other modalities. If you think about like text and image, text and video, they are loosely aligned. So you can, you can have a description of what's going on in the video, but you don't have to exactly, You typically don't have exact description, oh, at, time step one second like what happened?Vibhu [00:46:02]: It's veryEthan [00:46:03]: At time step two second what happenedVibhu [00:46:03]: coarse. Yeah.Swyx [00:46:05]: So what was the ideal time step? You have to oblate it, and then it's like four seconds or something.Ethan [00:46:09]: So that comes down to how you design the model to, for the model to be aware of as a time, as a time modality. So the model is like a time aware. And that's something pretty unique if you think about LLMs. So if you ask LLM to complete a task, say they, uh-- you ask them and they will say, “Oh, this task will probably take twelve hours to complete,” and they come back in one hour. Say “I've already spent two days on this and I've exhausted everything.”Ethan [00:46:47]: So the LLMs them-themselves, they don't have a sense of time there.Vibhu [00:46:53]: I actually don't think that's just them not having a sense of time. I think it's somewhat based, right?Vibhu [00:46:58]: Like you tell someone, “Okay, go work on this feature. Go implement this,” there's a general understanding you would have of how long that would take without LLMs working at LLM speed, right? So you think back like two years ago, if I tell you to like build me like a new front end for latent space, have a search bar, have all this, you'll estimate that it'll take a few days, right?Vibhu [00:47:19]: So you tell an LLM, “Go build this.” It'll take me a few days. But I think it's somewhat grounded as opposed to them not having the best-- Not saying that they have a great understanding, but I think that example is like you can see where it comes from, right? You're trained on all over the text.Swyx [00:47:35]: They're, they're trying to estimate what a human would say.Vibhu [00:47:37]: because that's what the, that's what the data kind of represents. It's not themEthan [00:47:41]: It came from the corpus on the internet. People have a estimate of how much time.Vibhu [00:47:45]: And not even just in direct like training samples, right? Just your world understanding of tokens of how long stuff takes, right? Go read a book. It'll take you a while, right?Vibhu [00:47:56]: Even if you do nothing but read a book, it takes a few days. So yeah, LLM, I read it took me a few hours.Vibhu [00:48:01]: It'll take me a few hours to go through this research. But this is a tangent.Swyx [00:48:05]: Somewhat, yeah.Swyx [00:48:06]: This is a train of thought I haven't really expressed until now is, which is basically like a full world model must also be recursive, meaning that the participant in the world model must also be aware that they have a world model. which is like this whole recursive thing down the, down the line. but yes, and that the world model can be wrong and that they need to update it and blah. Yeah. We've, argued this on the, newsletter as well, that there needs to be sort of recursive or adversarial world models.World Models: Real-Time, Long-Horizon, Interactive VideoVibhu [00:48:34]: just, to ask, how do you define world model?Swyx [00:48:38]: Oh, yeah, let's go there.Ethan [00:48:40]: SoVibhu [00:48:40]: So just for context, we talked about, video generation, and then there's a-- if you say there's a distinction between world models, what's your, what's your definition? How do you see the two?Ethan [00:48:53]: So disclaimer, I'm not going to debate, what is world model. Yeah. there are many definitions, so I'll just talk about my definition. Since I came from the multi-model, multi-model domain, so mainly talking from video. So world model is like real-time interactive long horizon videos. So there are three parts. so we-- let's talk about them one by one. So the so interaction, so we just, we just look at Facebook and neural computer. So the interaction part of it, so you, world model can allow you to interact with them through keyboard, mouse, and maybe also voice. So these all is-- all is a modality. You can, you can interact with the model, and the model should respond reasonably. Second part is real time. So once you, once, say, you move your mouse, if, say, the world model generate a game, how fast can the game respond? So if you're like professional CS: GO players- -my say, oh, you have to respond- He's beginner within sub ten milliseconds or- Yeah even less. So that's not most of the- No, sixty FPS. Let's go. Oh, three hundred FPS. Oh, five hundred FPS. Wait. okay, yeah. I didn't do the math, but yeah, okay. Uh- Yeah, three hundred FPS, that's a three millisecond. So you have to respond- Oh, s**t. Okay. YeahEthan [00:50:29]: within a millisecond. Most of the video models cannot do that. Yeah. And, but if you, say, if you have a video model that is, say, like a digital human, the response time might be more generous. Maybe typically, for real-time voice interaction, it's like two hundred millisecond. So that's, that's much more generous. But even two hundred millisecond is pretty, it is pretty tricky, ‘cause remember we mentionedEthan [00:51:01]: you have this, temporal compression coming from the VAE. So if you, if you don't compress the temporal dimension, your sequence length is going to explode. So if you want to have this real-time, real-timeness in your model, you have to do is one context problem. And the third part is long horizon, ‘cause we-- if you're not going to just play with, video games just, a few seconds, most video models only a few seconds. We're going to play with minutes, hours. The model have to be able to generate long-form content.Ethan [00:51:42]: So putting these three together, it's, real-time, long horizon interactive videos. I think the final state will be, for example, like a video, a video version of Playbook, where you can, you can interact with, a neural computer. You move your mouse, and you click on the generative interface, and it will reply to you through pixels- generating in real time. But getting there, it's, it's a very long way to get there. So one of the first step, at Grok Imagine, where I led a small world model team there, was to build video extension. So, video extension- it's the first step of interactivity. Yeah. It's, it's the first step. Yeah. So it's the first step- You have it here, video editing, yeah. Yeah. Yeah. So the first step is because, this unlocks long horizon videos. Typically, for most of the video generation models, you give it a prompt or an image as an initial frame. You generate video, that's it. That's just, one time, done. And some creators would try to, use the last frame as a first frame for the second video. It can-- sometimes it works, but if you do it a few times, it says the quality would decrease. And- It doesn't have that context- Yeah over the full video, so the temporal- Yeah, exactly. Yeah, ‘cause you only gave it the last frame, of course, right? Yeah. Exactly. And- it's actually a pretty fun hack. if you've seen like- Oh, no, he's saying something better. Yeah. And for example, like Vue, I remember Vue 3 has like a second context of the last video. It is slightly better than using the last frame, but it has the same problem-- similar problem that it, the quality would decrease. if you extend a few times to, one minute, the video quality would look much worse than the first video. Second, another problem is that the model doesn't have long-range knowledge of, what's happening before. Say, if they generate some dialogue, some, two people speaking, and their voice might change, over some time, especially if the second conditioning, it does not cover the previous context. So these are the core challenges. So the Grok Imagine video extension, it has historical context of all of the previous generated videos. It can, It has, it has the context of, who is speaking and what objects have appeared and everything, having that to generate the next video. So if we naively do this, you can imagine, just, put all of the previous history video tokens into the context. The context lens will easily explode. Especially for video models, that can be like a few, a few million context, I would imagine- context lens. Yes.Yeah.Swyx [00:54:58]: Let's run with that.Ethan [00:54:59]: for example, like in Cosmos, I think just five seconds of video is like a fifty K or sixty K number of tokens. So like if you do, if you do fifty second, that's a five hundred K tokens. If you do longer than that, easily explode. This long horizon, problem was the first step we're trying to solve world model. It turns out people, yeah, people love video extension. Like a lot, a lot of the creators love using video extension to create longer form videos. This is the part I liked that you have a, you have an intermediate step toward the final goal instead of just a straight shot to the final version very much.Swyx [00:55:48]: But I can see you have a strong vision of where we want to end up.Long Context, Redundancy, and Efficient Interactive VideoVibhu [00:55:51]: Does it seem like it's an efficiency issue? okay, we're at a few million tokens context,. If you draw the parallel to language models, we had very short context, two thousand, eight thousand, then, you scale it up one million, ten million. sure, there's effective context, but at the end of the day, it's just what's it worth? sure, there's a whole training data side. In video, it might be slightly easier ‘cause we have a hundred million token video, right? Just take a movie with the full context there. Like is this efficiency from an inference standpoint that like it's expensive, but we know how to solve it? Or like why is this not the approach? So like my broader point was on your second point of world models, you say it needs to be interactive and live, right? You should be able to play a game and see the interaction live. So one thing I see with research is a lot of what you actually serve is different than what you build, right? So we talked about distillation. You train big model, you distill it, you do quantization, speculative decoding. We do all this stuff to serve it efficiently. Should we not just have a solution, like a world model that can interact well, do inference optimization, serve it, distill it secondary, so make it real time after you solve it? So like a-- another parallel is say, continual learning, right? What we need is someone to solve it and show it works inefficiently. Give it a few years, people will make it efficient. Same thing with regular attention, right? It worked. Over a few years, people have different forms of attention, and we've scaled it to be efficient at log context,? So kind of two things there, right? One is it seems like it works. You've scaled it. Can we not just scale it a lot more efficiently over time? Do we need a separate approach if this works? And same thing with interaction, right? if we can get it done, like if we can solve some way that it works, we can solve making it more efficient from an inference standpoint later.Ethan [00:57:53]: that's actually a very good point. So in videos, there's actually a lot of redundancies. So we solve a lot of the pixel redundancy from VE, but there's more redundancy in long range and long horizon videos. Say, if a character appear in the first clip and then it disappeared, it only reappear at the end of the video, you probably don't need the-- the context, like in the middle of the generation. So you only need that character, where you need. So that's why, I helped build another feature. It's a reference video.Vibhu [00:58:36]: Is it here?Swyx [00:58:36]: is it the same model release or different one?Ethan [00:58:39]: It's a different one.Ethan [00:58:41]: You probably need to search onSwyx [00:58:43]: I'll find itEthan [00:58:43]: X reference to video.Ethan [00:58:46]: So reference video allow you to like upload up to seven images as condition and generate the video. Say, if like I want-- it can, it can be characters or objects or even scenes. Say like I want, I want condition on, Sean's selfie and holding a bladeSwyx [00:59:07]: We have a dogEthan [00:59:08]: or whatever.Swyx [00:59:08]: We put the dog in the thing.Ethan [00:59:09]: you can put them there and the video models will generate the video from and copies the context over. So that can solve a lot of the problems there, like the long context problem. It doesn't need to have a very long context, but it's-- I feel like it's an intermediate solution. The modelSwyx [00:59:29]: It's cheating.Ethan [00:59:30]: the model should be able to like selectively know, where should I draw the references. So say if I want to generate a movie, I generate it autoregressive, like a ten second at a time or something. And now this character appear, I can look back to where it first appear and, bring that back. Yeah, this one, I put the references. Yeah, that's, Optimus, Einstein myself, Annie.Vibhu [01:00:02]: Oddly enough, I used Grok Search to find it, and it pulled your LinkedIn post. But yeah we found it.Ethan [01:00:08]: Interesting.Vibhu [01:00:10]: ButxAI's Underrated Work, Culture, and WatermarkingSwyx [01:00:11]: this is a problem. This is not your fault, but like XAI doesn't communicate all this work that you do very well because they just have the model release and then that's it. But actually, these details are very good.Swyx [01:00:22]: As far as I understand, everything you just described is state-art, like no one else has done it.Vibhu [01:00:30]: A lot of-- yeah, I have a lot moreSwyx [01:00:32]: And then, and then you just put this blog post with the cookies. I'm this is not enough,?Swyx [01:00:37]: but I, obviously this is like the high level numbers that people want to know. But no, okay, soVibhu [01:00:42]: And I wonder, like part of that is also some labs don't share research into what happens. And ifSwyx [01:00:50]: No, but this is literally bragging about how good they are, right?Swyx [01:00:54]: Like, why would you not say that you are capable of extending with full context? this is not a secret sauce. This is like we did the work. yeah, I don't know.Ethan [01:01:02]: different labs have slightly different communication styles.Swyx [01:01:07]: Anyway, if anyone from XAI is listening we are always happy to help you tell your story. Yeah, okay, so you did references, and I think, I think kind of the point you're, you're making is it is sort of like a kludge, right? this is-- you can do seven, but what about 100?Swyx [01:01:23]: Right? Then you need a completely different thing.Ethan [01:01:26]: So I think it's-- this is, a mechanism to, select the context from the history, and you might not put the entire history into the context. for example, there's a paper called Frame Pack, which haveEthan [01:01:41]: a heuristic that the latest history, the last one second, I put the entire history, and the history before that, I would, compress it and makes the video smaller. So they follow this pattern, this build overall pattern that the maximum sequence length is fixed. So the further you are from the current frame, you have a smaller image. So this is just a heuristic. I think it can be more automatic. The model is aware like which history part of it can be select. So this part of the research is actually being actively, worked on by a lot of people. It's also quite interesting. I feel this is actually, this part of long context is a little bit ahead of the LLM part.Ethan [01:02:31]: So for example, like in LLMs, if you-- so contexts keep growing. Let's say if you call tool and the tool call history is extremely long, that's still in context, and keep growing, keep growing. Even if you switch the topic to something else, the whole context was there. There are some agentic harnesses that help you to, say, prune the tool results and, prune Like when you, when you query a file, only show like the top 200 lines or something. Those were very heuristic-driven.Swyx [01:03:08]: For listeners, we did a write-up on the cloud code, leak where there are eight different kinds of pruning, including like you prune the tool results and all that. So you can, you can read up on that kind of thing.Ethan [01:03:17]: I think, one breakthrough in continual learning might be like a way to automatically, manage its own context.Swyx [01:03:27]: These are all heuristics, and they will be replaced by machine learning.Ethan [01:03:30]: InterestinglyVibhu [01:03:32]: TheEthan [01:03:32]: the same thing is being researched in both LLMs and video models.Vibhu [01:03:36]: The interesting thing is also like in the paper you showed, it's actually happening at the model level, right? Compared to like language models, sure, we have base attention, but we'll do our own compression, we'll do our own pruning, which is separate from model error.Vibhu [01:03:49]: Eventually, it all just boils in, hopefully.Swyx [01:03:52]: I think this is a form of like attention, but like also know sort of reasoning attention. I feel like that's different than normal attention.Swyx [01:04:03]: Does that, does that make sense?Ethan [01:04:04]: It's, it's different in the sense that attention, not to mention, set sparse attention aside,

There is a house in Auburn, California, with a tragic history and a new tenant. Jean Averaud has come from New Orleans with money, with books, with a beautiful mute woman who watches him with eyes full of something between devotion and dread. He has come with a theory about evil — not the Devil, not sin, not the ordinary darkness of human nature, but evil as a cosmic force, a radiation from a black sun somewhere in the depths of space.And he has come with a purpose. In the old Larcom house, with its history of sorrow and disaster, he has found exactly the conditions he needs. His neighbour, a novelist, finds himself drawn into Averaud's orbit. Clark Ashton Smith's The Devotee of Evil is a quiet story. It does not rush. It thinks. And what it thinks about has been troubling philosophers and theologians for two thousand years. The Devotee of Evil was first published in Smith's self-produced chapbook The Double Shadow and Other Fantasies in 1933, after failing to find a commercial publisher. It reappeared in Stirring Science Stories in February 1941. Clark Ashton Smith (1893–1961) was a California poet, painter, sculptor and writer of weird fiction, one of the central figures of the Weird Tales circle alongside H.P. Lovecraft and Robert E. Howard, with whom he maintained a long correspondence.Become a supporter of this podcast: https://www.spreaker.com/podcast/the-classic-ghost-stories-podcast--7002956/support.*To buy my paperback books:* https://books.by/tony-walker-booksThe Classic Ghost Stories Newsletter — short essays on the genre, odd discoveries, and recommendations. Free, fortnightly. Subscribe: https://www.classicghost.com/#/portal To buy my ebooks and audiobooks: payhip.com/TheClassicGhostStoriesPodcastOr, if you'd just like to make a one-off gesture of thanks for my work https://buymeacoffee.com/10mn8sk *Intro and Outro Music by The Heartwood Institute*

FEAR IS A PRISON • WE ARE THE COSMOSwith Eve HowardEve Howard Rumble Link to Prime Declassified Podcast -https://rumble.com/user/Prime_Declassified_Podcast?e9s=src_v1_cmdEve Howard on Facebook - https://www.facebook.com/Evethunderhoward.4Tonight on the Typical Skeptic Podcast I'm joined by my good friend Eve Howard for a deep and emotional conversation about life, loss, consciousness, healing, and awakening.As the anniversary of her father's passing approaches, Eve reflects on how grief transformed her understanding of fear, purpose, and what it truly means to live. Together we'll explore the idea that fear itself may be the real prison — and that balance, awareness, and remembrance are the keys to reclaiming our power.

Hiranya Peiris is playing a starring role in a movie that promises to tell perhaps the greatest story of all time. However, it's a movie with a difference – there's no director and no script. The Legacy Survey of Space and Time is one of the most ambitious projects in the world of astronomy, with a mission to create a decade-long time-lapse movie of the visible universe, to answer fundamental questions about its origin, evolution and, ultimately, its fate.Hiranya is Professor of Astrophysics 1909, the prestigious Chair at the Institute of Astronomy at Cambridge University. Over her career she's been one of the pioneers of a revolution in astronomy, bridging fundamental physics with the observational data coming back from space, to establish the first evidence-based standard model for the origin, evolution and fate of the universe. The endeavour has transformed the field from the ‘wild west' of physics to the modern era of precision cosmology.Ironically, it was another movie, of sorts, Carl Sagan's documentary series ‘Cosmos', that first sparked Hiranya's interest in the universe as a young girl. Always keen to inspire women to follow in her footsteps and choose careers in science, if this interview were a live show she'd have reserved the front row for schoolgirls.