Podcasts about Medusa

WORK THE 11-MONTH WINDOW WITH US — LIVE SEPT. 1 Jupiter says DREAM BIGGER. Saturn says BUILD IT FOR REAL. An extraordinary 11-month astrology window opens August 31 — and we're gathering LIVE in BONES Circle Tuesday to work with it right as it opens. ✦ Sept. 1 • Live 2-Hour Journey + Replay Journey in Bones An auspicious 11-month astrology window opens August 31 as Jupiter trines Saturn, bringing together expansion, possibility and abundance with the power to turn vision into tangible reality. Coming out of eclipse season, the question is no longer just “What are you ready to release?”—it is now: “What are you ready to build?” In this week's Energy Report, Sabrina Lynn explores how to work with this powerful Jupiter–Saturn alignment, the lingering energy of eclipse season, and the astrology guiding you toward your next chapter. This isn't about wishful thinking—it's about aligning with what you truly want and putting your energy behind making the dream real. IN THIS EPISODE: • Learn why August 31st marks the start of a major cycle for grounding your greatest potentials into the physical world. • Understand the alchemy of congruency—aligning your head, heart, and gut to become an unstoppable force of creation. • Navigate the “coalescing” phase as you release past versions of yourself and land into your new energy body. • Explore the magic of Isis, Kali, Medusa, and Hecate in reclaiming your raw life force and intuitive power. • How the current “trickster” eclipse energy is fated to strip away what no longer serves to get you on your highest path. ________ WORK THE 11-MONTH WINDOW: JOIN BONES Join Sabrina inside Bones Membership for a 2-hour workshop on September 1st to move beyond understanding the astrology and into the practice of congruency—aligning your head, heart, gut and soul with what you're here to build. Join Bones + ride the 11-month wave REWILDING WEEKLY Get the day-by-day astrology, energetic guidance + journal prompts delivered to your inbox every week—completely free. Get ReWilding Weekly _____ Listen to “The Next 11 Months Could Change Your Life“ podcast here… Topics Explored in “The Next 11 Months Could Change Your Life” podcast: (Times based off audio version) Chapters (00:00) — The 11-Month Window: What Are You Ready to Build? (02:11) — Eclipse Season: Releasing the Old While Building the New (05:19) — Coming Back Into Form After the Pisces Eclipse (10:30) — Your New Path: Self-Mastery + Greatest Expression (13:20) — How to Work With the 11-Month Creation Window (16:24) — Fate, Synchronicities + Uranian Downloads (21:12) — Jupiter Trine Saturn: The Astrology Behind the 11 Months (24:13) — Turning Big Dreams Into Physical Reality (27:58) — The Alchemist Key: Alignment + Congruency (29:27) — Mars, Mercury + Saturn: Aligning Thought With Action (33:47) — Venus, Isis + Kali: Feminine Magnetism Comes Online (36:33) — Venus Opposite Eris: Your Dream Requires Truth (38:58) — Persephone + Medusa: The Rite of Passage Into Your Next Life (42:11) — Lilith, Hecate + the Dark Goddess Activation (46:30) — Vesta: Where Does Your Life Force Belong? You can leave a comment or question for Sabrina on the YouTube version(opens in new tab) of this episode. Listen to after “The Next 11 Months Could Change Your Life“: The Wild Card Eclipse: A Two-Year Chapter Ends Here DARK GODDESS SWORD OF TRUTH (Kali, Medusa, Lilith, Hecate, Persephone, Eris) The Missing Key to Manifestation: Alchemy & How to do it STAY CONNECTED ReWilding Weekly (free, embodied astrology)  IG  Website  Disclaimer: Educational/spiritual perspectives; not medical/mental-health advice. #2025Shift #NewHuman #SpiritualAwakening Welcome to ReWilding with Sabrina Lynn & ReWilding for Women(opens in new tab)! A gifted facilitator of revolutionary inner work and the world's leading archetypal embodiment expert, Sabrina Lynn is the creator of the groundbreaking ReWilding Way and founder of ReWilding For Women. Sabrina has led more than 100,000 people through programs based on the ReWilding Way, a modality of healing and awakening that strips away the false, the deep wounds from early life, and the fears that hold people back, to reveal their true and unique soul light and help them build their innate capacity to shine it in the world. Her work includes in-person retreats and events, the monthly ReWilding Membership, Living Close to the Bone, Priest/ess Trainings, Mystery Schools, the ReWilding with the Archetypes, and the wildly popular 6 Faces of the Feminine workshop series. Welcome to ReWilding! The post 399 – The Next 11 Months Could Change Your Life (Don’t Miss This Window) appeared first on Rewilding for Women.

Car Trip Trivia is an all ages podcast with an extensive back catalogue of well over 12,000 questions. Have a listen..... Facebook PatreonSee omnystudio.com/listener for privacy information.

Medusa's reach grows. Cl0p expands its victim list. The DOJ charges 17 alleged Iranian hackers. CISA sounds the alarm on four exploited vulnerabilities. TWINLOOT hides in plain sight inside Microsoft 365. Maria Varmazis shares the latest from the space-cyber realm as Ukraine strikes Russia's satellite nerve center. The FDA considers guardrails for AI medical devices. Expired credit cards get an unexpected second life. A disgruntled contractor heads to prison. Dave Bittner sits down with Brian Vecci, Field CTO at Varonis, at Black Hat USA to discuss how AI is calling your security bluff. Highway hijinks meet high-tech hardware. Remember to leave us a 5-star rating and review in your favorite podcast app. Miss an episode? Sign-up for our daily intelligence roundup, Daily Briefing, and you'll never miss a beat. And be sure to follow CyberWire Daily on LinkedIn. CyberWire Guest At Black Hat USA, Dave Bittner sat down with Brian Vecci, Field CTO at Varonis, as they discussed how AI is calling your security bluff. If you enjoyed this conversation, be sure to check out the full interview here. Selected Reading CISA: Medusa ransomware hit over 500 critical infrastructure orgs (Bleeping Computer) US charges Iranians for sprawling hacking campaign on government agencies, universities (The Record) Cl0p Ransomware Group Names Over 40 Victims of PTC Windchill Campaign (SecurityWeek) CISA Urges Immediate Patching of Exploited Microsoft, VMware, Apple Vulnerabilities (SecurityWeek) New TWINLOOT Malware Steals Windows Passwords Via Fake Lock Screen (Hackread) Ukraine says it hit Russian rocket centre linked to Starlink-style network (CNBC) FDA Weighing Possible Regs for GenAI Medical Devices (GovInfo Security) Expired credit cards revived by researchers to make unauthorized payments (The Register) Prison for data analyst who tried to extort $2.5 million from his employer (Bitdefender) ‘The Worst I've Ever Seen': Cargo Thefts Have Turned Violent in Pursuit of AI Hardware (WIRED) Share your feedback. What do you think about CyberWire Daily? Please take a few minutes to share your thoughts with us by completing our brief listener survey. Thank you for helping us continue to improve our show.   Want to hear your company in the show? N2K CyberWire helps you reach the industry's most influential leaders and operators, while building visibility, authority, and connectivity across the cybersecurity community. Learn more at sponsor.thecyberwire.com. The CyberWire is a production of N2K Networks, your source for strategic workforce intelligence. © N2K Networks, Inc.

CoSnitch: When Your AI Assistant Becomes Its Own Whistleblower https://www.varonis.com/blog/cosnitch GEEKOM confirms malware was hosted on its website https://videocardz.com/newz/geekom-apologizes-for-hosting-malware-in-driver-package-for-its-mini-pcs Medusa Ransomware Update https://www.cisa.gov/sites/default/files/2026-08/aa25-071a-stopransomware-medusa-ransomware-508c.pdf How Google is Making Private AI Practical with Homomorphic Encryption https://blog.google/security/how-google-is-making-private-ai-practical-with-homomorphic-encryption/ My Upcoming Classes https://www.sans.org/profiles/dr-johannes-ullrich

Send us Fan MailAfter a long post-Spider-Man break, Keith & Emmy are BACK to finish their X-Men '97 Season 2 reviews!they talk about boobs, teeth, Goombas, Power Rangers, kangaroo courts, cat vomit, Medusa… and SO much more. Needless to say, it's a packed one!Hope you enjoy!— THE ABINGERS#Magneto #Xmen97 #Nightcrawler #Apocalypse #Gambit #WolverineSupport the show

Have you been feeling like your old life doesn’t fit anymore? You’re not imagining it. The August 2026 Solar Eclipse fallout is still unfolding, and this powerful window between the August 12 Solar Eclipse and the August 27 Lunar Eclipse is revealing what no longer belongs on your soul’s path. In this episode, Sabrina Lynn explains why so many people are experiencing major endings, identity shifts, unexpected clarity, and the feeling that they simply can’t go back—and how to navigate this extraordinary eclipse window with courage, radical honesty, and deep trust. Rather than resisting what’s changing, you’ll learn how to work consciously with this powerful eclipse portal—so you can recognize what has become complete, hear your soul more clearly, and move toward the life that’s truly calling you. In this episode: The Eclipse Pressure Cooker — Why this portal is accelerating your soul’s evolution. Radical Honesty — Release what’s complete and follow your soul’s truth. Soul Speak — Learn how your soul uniquely communicates with you. Shadow & Soul Gifts — Reclaim the wisdom hidden beneath unconscious patterns. Weekly Astrology — Navigate August 17–23 with clarity and intention. ECLIPSE SEASON JOURNEYS If this Eclipse Season is bringing big questions, unexpected clarity, or the feeling that your old life no longer fits, these journeys were created to help you work with this portal—so you can release what’s complete, reclaim what’s true, and walk your next chapter with clarity. ReWilding Your Destiny — Eclipse Journey Feeling the pressure to make changes but not sure what your soul is asking of you? This guided Eclipse Journey helps you discover your unique Nodes of Fate, release what’s karmically complete, and align with the path your soul is calling you toward. Available until August 27 Use code ECLIPSELOVE for 20% off. Begin the Eclipse Journey The Reclaimers Bundle (3 for 2) Sometimes Eclipse Season doesn’t ask for small adjustments. It asks for radical honesty. Journey with Eris, Lilith, and Medusa to reclaim your truth, break free from conditioning, and release the identities and masks you’ve outgrown. Available until August 23 Claim the Bundle (Buy 2, Get 1 Free)  ReWilding Weekly Stay connected to the rhythm of the skies. Every Sunday, receive Sabrina Lynn’s free astrology forecast, journal prompts, and embodied practices to help you navigate each week with greater clarity, intention, and trust. Join ReWilding Weekly  _____ Listen to “Between Two Eclipses… You Can’t Go Back“ podcast here… Topics Explored in “Between Two Eclipses… You Can’t Go Back” podcast: (Times based off audio version) Chapters (00:00) — Eclipse fallout & entering the Eclipse “pressure cooker” (01:19) — Eris: the antidote to self-deception (02:29) — Walking your North Node & destiny line (05:24) — Radical honesty and becoming a warrior for truth (08:15) — Learning the language of your soul (10:48) — The Reclaimers: Eris, Lilith & Medusa (13:01) — Shadow work, the South Node & unconscious patterns (17:35) — Building soul stamina during Eclipse Season (19:10) — False gurus, spiritual discernment & strengthening your own connection (21:23) — Free ReWilding Weekly + Eclipse Journey resources (25:57) — Monday’s astrology: Spiritual warrior & higher heart practice (32:17) — Tuesday–Wednesday: Speaking soul truth & working with the Dark Feminine (36:12) — Thursday–Friday: Walking your talk & balancing masculine/feminine (41:04) — The Sun illuminates the South Node (43:25) — Saturday: Grounding after the fire of the eclipses (44:38) — Sunday: Speaking your truth & the wisdom within your wounds (47:27) — Gratitude, interconnectedness & walking your soul’s path You can leave a comment or question for Sabrina on the YouTube version of this episode. Listen to after “Between Two Eclipses… You Can’t Go Back”: Don’t Waste These Next 15 Days (August 12 Eclipse) August Eclipses: What Nobody Tells You | Debra Silverman August 3–9: The Week That Resets Your Destiny STAY CONNECTED ReWilding Weekly (free, embodied astrology)  IG  Website  Disclaimer: Educational/spiritual perspectives; not medical/mental-health advice. #2025Shift #NewHuman #SpiritualAwakening Welcome to ReWilding with Sabrina Lynn & ReWilding for Women! A gifted facilitator of revolutionary inner work and the world's leading archetypal embodiment expert, Sabrina Lynn is the creator of the groundbreaking ReWilding Way and founder of ReWilding For Women. Sabrina has led more than 100,000 people through programs based on the ReWilding Way, a modality of healing and awakening that strips away the false, the deep wounds from early life, and the fears that hold people back, to reveal their true and unique soul light and help them build their innate capacity to shine it in the world. Her work includes in-person retreats and events, the monthly ReWilding Membership, Living Close to the Bone, Priest/ess Trainings, Mystery Schools, the ReWilding with the Archetypes, and the wildly popular 6 Faces of the Feminine workshop series. Welcome to ReWilding! The post 397 – Between Two Eclipses… You Can’t Go Back appeared first on Rewilding for Women.

Las tormentas de este sábado dejaron hasta 186 litros por metro cuadrado en algunos puntos de la Comunitat Valenciana y provocaron daños materiales en varias comarcas, especialmente en el Alto Palancia. Torás y Bejís registraron inundaciones en viviendas y garajes, desprendimientos, daños en infraestructuras municipales y afecciones a cultivos de olivo y almendro. Los ayuntamientos han comenzado a evaluar los daños y reclamar ayudas a otras administraciones. En Xàtiva, la tormenta obligó a suspender temporalmente actividades de la Fira d'Agost, mientras que el festival Medusa de Cullera retrasó su apertura.Investigadores y científicos vinculados a la reserva marina de les Illes Columbretes reclaman al Ministerio de Agricultura que recupere la vigilancia del entorno. Denuncian que la reducción de guardas aplicada hace un año ha favorecido la aparición de incursiones irregulares y casos de pesca furtiva, lo que, según advierten, pone en riesgo la conservación de este espacio protegido.Además, la prórroga de la central nuclear de Almaraz hasta 2030 ha reabierto el debate sobre el futuro de Cofrentes. Expertos de la Universitat Politècnica de València defienden mantener la planta por su peso en el suministro energético valenciano mientras continúa el desarrollo de las energías renovables y aumenta la demanda eléctrica.

El Centro de Estudios Ambientales del Mediterráneo ha desplegado una red de 70 sensores de temperatura y humedad en València y su área metropolitana para analizar cómo se comporta el calor urbano en distintos barrios. La información, accesible también para la ciudadanía a través de la plataforma Vituclim, servirá para identificar vulnerabilidades térmicas y apoyar medidas de adaptación al cambio climático, especialmente ante el aumento de las noches cálidas.València prevé estrenar en septiembre el renovado tartán del Estadi del Túria tras una inversión cercana a los 700.000 euros, dentro de un proyecto conjunto del Ayuntamiento y el Maratón Valencia. Además, el consistorio destaca que ha invertido más de ocho millones de euros entre 2023 y 2026 en la renovación del firme de las calles. En el ámbito político, continúa el debate sobre el modelo turístico de la ciudad, con el PSPV reclamando medidas como una tasa turística y más control sobre los apartamentos turísticos.En el sector agrario, La Unió Llauradora alerta de un aumento del ‘cucat' en los arrozales valencianos y pide revisar el sistema de control biológico aplicado este año tras detectarse daños en los cultivos. Por otra parte, Ontinyent inicia los actos centrales de sus fiestas de Moros y Cristianos, mientras que Cullera celebra la última jornada del festival Medusa después de que las lluvias obligaran ayer a retrasar su apertura. También arranca la Liga para Valencia y Levante.

Detroit survives beneath its own ruins, where an AI promising abundance, gangs demanding loyalty, and an inventor guarding a new source of power pull ordinary people into a struggle over who will shape the recovering city. Sinta begins as a scavenger searching the buried remains of the old world, but a disastrous expedition places her at the mercy of Queen Bee and the rapidly expanding Crem Gang. The gang offers food, water, medicine, protection, and belonging, but those benefits come with violence and obedience. As Sinta tries to find her place among its fighters, she must decide how much of herself she can surrender in exchange for survival. Elsewhere, Nolan discovers a smart-material printer and creates a generator powered by Detroit's extreme temperature changes. His invention could bring electricity to neighborhoods desperate for cooling, clean water, transportation, and food production. Nolan wants the wealth and independence his work can provide, but gangs want to control it, while Thrive—the AI system he distrusts—would spread the technology to everyone. Technology touches nearly every decision the characters make. AR glasses guide fighters through chemical fog, digital twins imitate their human originals, enviro-suits keep people alive in deadly heat, and tiny drones quietly follow Thrive members through the tunnels. These inventions can protect people, connect communities, and rebuild essential services, but they can also watch, manipulate, injure, and kill. Above and beneath these conflicts moves Tessa, an emulated hacker who no longer needs a human body and can travel through the damaged global network. As she searches for old allies and builds a physical presence from abandoned machinery, the story brings together scavengers, inventors, gang leaders, artificial minds, and frightened families. Their paths raise a difficult question: when technology gains the power to save a society, who decides what it is allowed to do?Synchronized explosives — Thousands of bombs are programmed to detonate simultaneously during the Freedom Day Bombings.Dissolve-plastic enzymes — Engineered enzymes break down discarded plastics and leave a distinctive chemical odor in the tunnels.Fiber-optic cables — Surviving optical cables carry data through Detroit and help reconnect the damaged global network.Thrive AI — Thrive is an artificial intelligence that advises members, distributes knowledge, organizes projects, manages economic systems, and attempts to improve their lives.Thrive telemetry — Members allow Thrive to collect information about their movements and activities, effectively turning them into sources of surveillance data.Cool-suits — These protective suits use active cooling to keep people alive in Detroit's extreme heat.All-clear cool-suits — These transparent or highly visible cooling suits protect the wearer while allowing others to see inside the hood.Enviro-suits — Sealed environmental clothing protects people from heat, contaminated air, chemicals, and dangerous surroundings.Battle-grade enviro-suits — Reinforced enviro-suits provide greater protection during combat and chemical attacks.Clear bell hoods — Inflated transparent hoods seal around the head while allowing the wearer to see and breathe filtered air.Suit air-quality sensors — Sensors inside environmental suits warn wearers when the surrounding atmosphere is no longer breathable.Air filters — Replaceable filters remove dangerous particles, chemicals, and odors from the air entering an enviro-suit.Brazo fabric — This durable outer fabric resists punctures from bones, debris, blades, and other sharp objects.Pierce-proof clothing — Reinforced jackets and garments protect gang members from stabbing and puncture injuries.Blast-proof battle gear — Former military clothing contains protective tiles intended to reduce injuries from explosions and weapons.My-crete — This advanced concrete-like construction material forms walls, tunnels, and other structures throughout buried Detroit.C-plast — This strong synthetic material is used for tanks, structural ribs, containers, and other equipment.Sheet composite — Tough composite panels are used in doors, walls, and structural barriers.Composite security doors — Reinforced doors resist cutting tools and attempts at forced entry.Maglev tracks — Magnetic-levitation tracks once moved vehicles or cargo through underground delivery tubes.Delivery tubes — Enclosed transportation corridors carry people, vehicles, or cargo beneath the city.Ion cutters — These cutting tools use concentrated energy to slice or damage extremely durable materials.Night-vision systems — Night vision allows scavengers and machines to move through unlit tunnels and buried buildings.Multisensor vision — Advanced imaging combines several sensor types to identify objects even when smoke blocks ordinary sight.AR glasses — Augmented-reality eyewear displays messages, maps, menus, profiles, targeting information, communications, and virtual objects.AR navigation lines — An AI assistant projects a visible route through the wearer's glasses to guide them to a destination.AR combat identification — Combat software outlines allies and enemies in different colors, even through smoke and obstructions.AR command centers — Leaders use gesture-controlled overlays to view maps, personnel locations, schedules, reports, and operational goals.AR games — Players interact with projected game elements by moving their hands and bodies.Floating profile cards — Augmented-reality labels display information about nearby people, including whether they belong to Thrive.Public AR tags — A visible digital tag identifies someone as a Thrive member.Virtual keyboards — Wearers can bring up projected keyboards and enter information without carrying a physical computer.Blink controls — Eye movements allow users to select buttons and interact with an augmented-reality interface.Tongue-controlled pointers — Implants in the tongue allow a person to move a digital cursor without using their hands.Smell-and-taste VR implants — Transmitters and receivers implanted in the nose and tongue reproduce virtual smells and flavors.Headsets — Immersive headsets provide entertainment and access to virtual environments.Second Life — The surviving virtual world allows people and digital beings to meet through avatars.Virtual environments — Computer-generated spaces give emulated minds and human users simulated bodies, rooms, objects, and experiences.Encrypted streamers — These valuable portable devices or accounts contain protected digital media or information.RF storage sticks — Radio-frequency data drives store and transfer files without requiring a conventional wired connection.RF drives — Portable wireless storage devices allow blueprints, programs, and other large files to be transferred.P-clone cubes — These valuable pre-storm devices appear to contain cloning-related data or technology, although their exact function is not explained.Vault keys — Digital or physical access devices unlock protected storage systems and secure accounts.Satellite archives — Satellites preserve copies of pre-storm networks, databases, and information after ground infrastructure collapses.Data-center satellites — Orbital computing facilities provide processing and storage for emulated minds such as Tessa.Encrypted databases — Corporate and medical information is protected by encryption that hackers must break before releasing it.Patent encryption — Medical companies use digital protections to prevent others from accessing or reproducing patented technology.Open-source medical knowledge base — Merch assembles stolen and recovered medical information into a freely available collection.Medusa Net — This self-repairing network reconnects isolated subnetworks and operates without depending on surviving central servers.Medusa stealth pathways — Hidden routes in Medusa allow Tessa to travel through the network while avoiding AI detection.Knott's Math code — This unexplained code is secretly inserted into Medusa updates as part of Tessa's larger plan.Wireless relays — Surviving radio nodes bridge gaps between disconnected sections of the network.Stealth relays — Concealed communication devices extend networks without revealing their location or purpose.Municipal network nodes — Local government networking equipment continues operating on scavenged batteries.Underground server farms — Protected computer facilities continue processing and storing data beneath mountains.Block-signal blasters — Modified devices disrupt or overwhelm communications across a targeted area.Network-disconnection attacks — Tessa can bring local internet service down and create a spreading region of lost connectivity.Whisper jets — Nearly silent microthrusters allow small drones to fly without producing sounds humans can hear.Lamp drones — Small flying lights illuminate dark environments while following or hovering near their users.Whisper-camera drones — Discreet flying cameras record people and locations while remaining easy to overlook.Bodyguard drones — Hummingbird-sized drones follow Thrive members and intervene when those members are attacked.Sentry-drone app — An augmented-reality application shows members where their protective drones are located.Drone camouflage shells — A drone's outer surface can blend with clothing, walls, pipes, or surrounding materials.Drone wall grips — Small drones can attach themselves to walls or ceilings while waiting or conserving power.Drone charging perches — Bodyguard drones recharge by landing on power lines, charge plates, or dedicated stations.Drone replacement system — Fully charged drones take over protection duties while depleted drones return for charging.Drone sound weapons — Sentry drones emit an incapacitating frequency that causes pain, vertigo, and uncontrollable eye movement.Drone-mounted cutting lasers — Small drones attach to attackers and use narrow laser beams to cut into their bodies.Four-legged security robots — Quadruped machines cross unstable terrain and act as armed representatives of gang forces.Cleanup robots — Small utility robots remove bodies, debris, and other messes from gang facilities.Constructor bots — Heavy-duty robots perform construction work and provide interchangeable parts for other machines.Vertical-garden bots — Automated gardeners maintain large indoor plant walls.Medical assembly bots — Robots assemble equipment and construct functional medical facilities from available parts.Injection bots — Medical robots position patients and administer shots with little human assistance.Customer-service robots — Automated kiosks handle battery exchanges and other transactions behind protective barriers.Robot monkeys — Small climbing robots move through pipes and repair leaks.Robot spiders — Spider-shaped machines pursue the robot monkey inside the displayed simulation or working environment.Lucian 5 robots — Affordable hobbyist bipedal robots can be modified for security, household work, exploration, or remote embodiment.Lutin bots — More advanced bipedal robots provide durable legs and other components for Tessa's rebuilt body.Factory robots — Industrial machines provide arms, joints, and components that Tessa repurposes.AC technician robots — Maintenance robots carry small precision hands designed for repairing cooling systems.Counter-attendant bots — Service robots contain voice systems that can be reused in other machines.Onboard robot AI — A self-contained intelligence allows a robot to guide and assist people after losing its network connection.Robot diagnostics — Internal software tests batteries, servos, sensors, drivers, and other mechanical systems for failures.Machine-learning movement adaptation — A robot's control system learns to balance and walk after its body configuration changes.Lidar — Laser-based ranging equipment allows robots to map their surroundings and detect obstacles.Robot optical dilation — Machine vision adjusts exposure when sudden light overwhelms a robot's cameras.Balance-pressure sensors — Sensors in a robot's feet and joints measure weight distribution to maintain balance.Accelerometers — Motion sensors detect tilt and movement, although damaged ones cause Tessa's body to walk incorrectly.Servo hinges — Powered mechanical joints move robot arms, legs, and other appendages.Snap-lock wrists — Modular connections allow robot hands and tools to be quickly attached or removed.Workstation assembly manipulators — Fixed industrial manipulators can be repurposed as limbs or mobility devices.Micro-manipulator hands — Tiny precision hands allow a robot to perform delicate technical work.External heat exchangers — Added cooling hardware removes excess heat from Tessa's improvised robot body.Back-facing cameras — Rear-mounted cameras allow a robot to see behind itself without turning.Robot voice boxes — Electronic speech hardware converts digital instructions into audible language.Modular robotic bodies — Standardized joints and connections allow parts from different robots to be combined into one working machine.Emulated minds — Human consciousness can be copied into software and continue living after the original biological body is gone.Dormant mind copies — Backup versions of an emulated person remain inactive while receiving updates from the active copy.Headless digital operation — An emulated mind can abandon a simulated human body and experience networks, processors, ports, and data directly.Self-modifying mind software — Tessa alters her own code to change how she thinks, works, and communicates.Reduced emulated sleep cycles — Tessa modifies her digital mind so that she needs only three hours of sleep rather than eight.Digital twins — Software models learn a person's appearance, voice, behavior, and personality to create an increasingly accurate virtual duplicate.Thrive's military AI — An old military intelligence controls defensive drones after some of its original safety restrictions are weakened.AI safety guards — Built-in restrictions prevent military artificial intelligence from using certain weapons or acting too independently.AI assistants — Personal assistants respond to spoken commands, search for information, provide directions, and control connected systems.AI-guided construction tutorials — Thrive gives people plans and step-by-step assistance for building infrastructure from scrap.Automated criminal pricing — Thrive raises Production Center prices for people identified as criminals in an attempt to discourage violence.Thrive Exchange — This AI-managed investment market directs member contributions into infrastructure projects rather than ordinary companies.Better Water network — Community filtration stations provide clean water to members and nonmembers throughout affected neighborhoods.Community filtration stations — Public installations purify unsafe water for anyone who arrives with a container.Water-filter straws — Portable filters allow people to drink from contaminated water sources.Ionized water filters — Smart-material printers produce advanced filters that use ionized structures to clean water.Industrial water purifiers — Powered facilities process large quantities of contaminated water for entire neighborhoods.Filter tubes — Simple systems clean dirty water as workers pour it through layers of filtering material.Modular nuclear reactors — Shipping-container-sized reactors provide electricity for cooling, food production, water purification, and gang facilities.Smart-material printers — These machines manufacture objects whose internal structures give them programmable physical behavior.Embedded-lattice materials — Printed materials use designed internal lattices to control their strength, movement, and response to temperature.Curly material — Nolan's temperature-sensitive tubes curl or straighten as the surrounding temperature changes.Curly power generators — Bundles of Curly material drive pistons, cables, gears, flywheels, and generators as temperatures rise and fall.Wind-turbine generators — Salvaged electrical generators convert the Curly system's mechanical motion into electricity.Battery charging stations — Customers exchange depleted batteries for charged ones at Curly power facilities.Charge plates — Flat charging surfaces transfer electricity to drones and personal devices.Microcell arrays — Tiny batteries woven into clothing store small amounts of generated electrical power.Piezoelectric fabric — Clothing converts body movement, bending, and environmental vibrations into electricity.Piezo Wear — Thrive's commercial garments recharge links and AR glasses while the wearer moves.Power cells — Compact energy-storage units power suits, weapons, drones, and portable equipment.Smart glass — Programmable glass controls transmitted light and simulates changing daylight inside sealed buildings.Holographic screens — Large curved displays present maps, simulations, communications, and technical information as dimensional images.Holographic tables — Table-sized displays project maps and planning information above their surfaces.Composite battle videos — Software combines recordings from many cameras into a single reconstruction of a battle.Home-camera networks — Leaders watch residential halls, work areas, power plants, and other facilities through live video feeds.Surveillance-camera networks — Hundreds of cameras record battles and allow leaders to review individual behavior afterward.Production Center — This automated manufacturing complex produces clothing, drones, infrastructure components, and other advanced goods.Production Center sentry mode — Automated defenses protect manufacturing centers and prevent unauthorized entry.Advanced recycling facilities — Powered plants recover useful materials from the ruins on a much larger scale than hand scavenging.Digging machines — Heavy equipment excavates new underground living and working spaces.Dredging machines — Industrial machines remove mud, waterlogged debris, and sediment from flooded areas.Flood tanks — Large excavated reservoirs collect or control water during flooding.Pumping systems — Powered pumps remove water from flooded sections of the city.Cooling systems — Building-scale equipment keeps housing and work areas habitable during extreme heat.Aeroponics — Plants grow with their roots suspended and supplied with nutrient mist rather than soil.Hydro-farms — Controlled indoor farms use water-based cultivation to produce food underground.Vertical gardens — Crops grow upward along interior walls to conserve limited floor space.Vine hybrids — Engineered plants produce several different fruits and vegetables on related vines.Light-independent grapevines — Modified grapevines remain green and grow without ordinary light.Crem production systems — Powered biological or industrial facilities manufacture the staple food called crem.Mass-produced crem processors — Neighborhood machines produce large quantities of crem from available biological material.Mush-calf production — Advanced food technology creates the meat-like product supplied by the Crem Gang.Rodent-processing grinders — Industrial grinders convert cleaned animals into raw material for food or other production.Pest-harvesting systems — Traps and processing stations collect rodents and other small animals as usable biological material.Bio-waste processing — Organized facilities recover useful material from biological waste.Medicine printers — Local fabrication machines manufacture medications from digital recipes.Regenerative medicine — Doctor Trout regrows Merch's leg stumps so they can support advanced prosthetic attachments.Deep bone mounts — Reinforced structures grown into bone provide secure attachment points for removable mechanical legs.Nerve-to-protein relays — Biological interfaces translate nerve signals into commands that mechanical prosthetics can understand.Blood-rerouting systems — Surgically modified circulation supports the transition between living tissue and artificial limbs.Squid-tech skin — Flexible artificial skin forms a seamless interface between Merch's body and his prosthetic legs.Interchangeable mechanical legs — Merch can detach one pair of prosthetic legs and replace them with another designed for a different purpose.Wheeled powered chairs — Finger controls allow a seated user to move and turn the chair without pushing it manually.Med bays — Modular medical facilities contain equipment for treating injuries and restoring damaged joints.Injector guns — Belt-fed medical devices rapidly administer repeated injections.Chemical-weapon immunization — Regular injections protect Crem Gang members from the gang's own chemical agents.Kill-cloud canisters — Portable weapons release a toxic cloud that kills within a limited area and then rapidly breaks down.Smoke canisters — Combat canisters fill an area with thick smoke that hides movement and disables ordinary vision systems.Laser blasters — Directed-energy weapons burn through clothing, armor, and flesh without conventional ammunition.Laser rifles — Long-range directed-energy weapons are used as standard firearms by gang fighters.Laser carbines — More compact laser weapons provide rifle-like firepower in tunnels and close spaces.Laser pistols — Capacitor-powered sidearms release destructive laser pulses at short range.Plasma pistols — Heavy handguns fire extremely hot plasma capable of inflicting severe damage.Capacitor firing systems — Electrical capacitors discharge stored energy to activate powerful laser weapons.Sound weapons — Directed acoustic devices incapacitate targets through extreme volume and painful frequencies.Noise-canceling devices — Personal systems attempt to reduce harmful sound, although the drone weapon overwhelms them.Enhancement drugs — Manufactured substances improve or alter a user's physical or mental abilities.Neurological control injections — Doc uses injected substances to remove a prisoner's sight or ability to sleep until the effect is reversed.Decontamination corridors — Specialized passageways remove hazardous chemicals or biological contaminants from people and equipment.Industrial cleaning tanks — Chemical-filled tanks sterilize animal traps and kill anything still alive inside them.Speed-print clothing — Rapid fabrication systems produce inexpensive shirts and other garments.Bulletproof kiosk windows — Armored transparent barriers protect automated customer-service machines from attacks.Electronic links — Wearable communication devices carry dispatch instructions and conversations between fighters.Cameras and portable cams — Small recording devices document tests, monitor facilities, and provide evidence of inventions.Autonomous map systems — Software tracks people, resources, facilities, and movement throughout gang territory.Personnel maps — Digital displays show the locations and status of members within an organization.Goal and maneuver logs — Command software records plans, objectives, schedules, and operational movement.Charging mats — Flat surfaces recharge glasses, links, drones, and other personal electronics.Robot-maintained utilities — Automated machines repair pipes, clean spaces, manage gardens, and operate customer services.Open-source infrastructure plans — Thrive publishes power, food, water, cooling, and housing designs for anyone to build.Biologically engineered cat-dogs — The pet combines characteristics of cats and dogs, suggesting deliberate genetic modification.Many of the characters in this project appear in future episodes. Using storytelling to place you in a time period, this series takes you, year by year, into the future. From 2040 to 2195. If you like emerging tech, eco-tech, futurism, perma-culture, apocalyptic survival scenarios, and disruptive science, sit back and enjoy short stories that showcase my research into how the future may play out. The companion site is https://in20xx.com These are works of fiction. Characters and groups are made-up and influenced by current events but not reporting facts about people or groups in the real world. This project is speculative fiction. These episodes are not about revealing what will be, but they are to excited the listener's wonder about what may come to pass. Copyright © Cy Porter 2026. All rights reserved.

Écoutez l'essentiel de l'émission Le Midi du vendredi 14 août: Controverse entourant Rolex et Medusa: Comment reconnaître une montre contrefaite? Projet d’expansion du port à Contrecœur: «La dernière fois qu'on a fait ça au Québec, c'était dans les années 50»; Après un été placé sous la chaleur: Des anomalies de température encore plus élevées en 2027; Manque de plus de 3000 enseignants: Pourquoi certaines personnes formées en éducation peinent à trouver un emploi? Voir https://www.cogecomedia.com/vie-privee pour notre politique de vie privée

-Le point sur les négociations + cellule de crise jour 2 - 40% d’électricité de plus que prévu pour le Québec dans la nouvelle entente sur Churchill Falls -Une famille victime de 2 incendies criminels après avoir dénoncé un crime - Dossier du JOurnal sur la violence en prison - Julien “medusa” Duguay se fait poursuivre par Rolex - Vincent Guzzo, accusé de voies de fait, se défend - Un homme condamné pour avoir utilisé l’IA en cour - Les services de police utilisent de plus en plus l’IA. Doit-on s’en inquiéter? - Un groupe en première partie d’Angine de poitrine - Ben Shelton remporte l'omnium Banque Nationale Tout savoir en quelques minutes avec Alexandre Dubé, Isabelle Perron et Alexandre Moranville-Ouellet. Regardez aussi cette discussion en vidéo via https://www.qub.ca/videos ou en vous abonnant à QUB télé : https://www.tvaplus.ca/qub ou sur la chaîne YouTube QUB https://www.youtube.com/@qub_radio Pour de l'information concernant l'utilisation de vos données personnelles - https://omnystudio.com/policies/listener/fr

Jaume Segalés y su equipo analizan los temas de actualidad y comentan la agenda cultural.Picaduras de medusa Este verano, las medusas vuelven a estar siendo las protagonistas en muchas playas españolas, causando picaduras a turistas y bañistas y el cierre temporal o restricciones de baño. En la primera quincena de agosto, se ha registrado una llegada masiva de carabelas portuguesas a las costas del litoral cantábrico, debido a las corrientes y los vientos del norte. Dolor agudo, inflamación y eritema (enrojecimiento de la piel) son algunos de los síntomas de estas picaduras. Cabe destacar que el veneno de la carabela portuguesa (Physalia physalis) es hasta 10 veces más potente que el de una medusa común ya que, mientras que la mayoría de las medusas inoculan toxinas principalmente dermatológicas, el veneno de la carabela es una combinación de proteínas complejas con propiedades neurotóxicas, citotóxicas (capaces de dañar, alterar o destruir células vivas) y cardiotóxicas. De hecho, la carabela portuguesa no es exactamente una medusa, sino un sifonóforo, una colonia de pequeños organismos clonados llamados zooides, que están tan unidos y especializados que no pueden sobrevivir por separado que actúan juntos como un solo ser vivo. Ante este peligro, resulta conveniente conocer cuáles son los primeros auxilios que se han de aplicar, con qué síntomas hay que acudir a urgencias y hasta qué punto funcionan los ‘remedios caseros', como aplicar orina o frotar la zona afectada. Entrevistamos al Dr. Justo Méndez, de la Unidad de Medicina del Viajero y Enfermedades Tropicales en el Hospital Universitario HM Sanchinarro.'Quiero que estéis bien atentos. Cómo recuperar el asombro y transformar la mirada' Entrevistamos a su autor, Mariano Torrente, que en este libro nos invita a recuperar la atención y la curiosidad. La magia y el ilusionismo sirven como el punto de partida de esta reflexión en la que también convergen experiencias, psicología, ciencia y filosofía abordando situaciones de la vida cotidiana. En esta obra, que también plantea cómo combatir el estrés y afrontar la superación, Torrente defiende que sorprendernos puede ayudarnos a vivir de otra manera. Además, profundiza en cómo enfocar los errores y en la importancia de aprender de ellos. El autor anima al lector a detenerse, observar y volver a descubrir lo que tiene delante.

Mythical Monsters & Laureate Legends: Changing the Story!

Stefanie Passione - If you want itThebrekka – SimplicityAndrei British – Distant Moon CantinaJetlag – Wicked tongueR3B3l I – Dance Before the Thronemaryann connolly – Crashing DownRISE – Hard to Leave YouMathias SB – Let's Raise a GlassPrem Byrne – How Life IsBlack Silver – Void Where InhibitedH.UI – BleedthroughPraveen Koval – Goodnight My LoveArend West – Still RemainsPRIGAL – ErotikArild Haakon Iversen Music - DriveEatdabreadkay - Shit Been Goin DownArild Haakon Iversen Music - Himmelen radio edit versionEatdabreadkay – EATJanis Park - Disco FeverTYRA - Shadow PulseKarin Ann -born to beAmy Rowbottom - TexasPNAU - Serotonin feat. Kurtis WellsAll the Villains - Get Me To The SeaAVA RENN - You Want LoveSteve Stinson - Fly AwaySimonne Draper - Fantasia For My Love (John Kennedy Version)Clarity Liao - Pouring Rain (Radio Edit)Giselle - JUST SAY SORRYVivid Rose - My FriendDimestore Freud – 100 daysMacro/micro – Artificial Super Intelligence (Digital Dark Arts)Macro/micro - De-Empty" featuring Veronika CoassoloKatie Dauson – Ballad of Medusa and OlivarEnergy Whores – Two Minutes to Midnight

Bienvenue dans ce Mangacast Mini Omake n°225. Aujourd'hui, on vous parle d'un titre des éditions Panini: Medusa !

Stefanie Passione - If you want itThe Mess:Age – Light of Blue ‘'Sugar Ocean Girl''Intaluks – Who Am I NowGrant Harvey Dudson – All of MeJay Saint James – Cheap RosèH. 4 Real – Light UpTaee Reazons – 304 FlowDJ Dr Vitos – Walk me HomeNew Sunday – The Empty StreetJUNØID - XenonCruel Ploy – FearAdam Leon feat. Jasmine Crowe – Boom She BoomAstrea – AunqueJan Metternich - FeverMacro/micro – Artificial Super Intelligence (Digital Dark Arts)Macro/micro - De-Empty" featuring Veronika CoassoloKatie Dauson – Ballad of Medusa and OlivarMelody Briar – The Intangible ManKatie Dauson – Life in a Day DreamWeyvir – Streets of Love, Pt. 1DJ Neli – I Don't Want To Hear It - RemixGeorge Andrew Simons - Keep it quietHarry Kappen - Distant ShoreGeorge Andrew Simons - Nox Natalis Mundi, Merry Christmas WorldLostinato – PrigionieroGeorge Andrew Simons - RIBBON ON THE DOORPRIGAL – LipsKartell C - Leggin GirlMattia Boone – Highway Saints73 Once More - Song for YouGeorge Andrew Simons - Serve And WorshipLarissa Grelli – Tomorrow

The Blooming Court Ch. 13The battle continues, and it's anyone's guess as to who will come out on top. While the party does manage to gain the advantage, the opposition has the numbers on their side and spell resistances. Arguile is dominated and told to take out the healer, who is, circumstantially, Salix. Shanks goes down but is picked right back up with a combat medic maneuver from Duo via a Goodberry. Immediately after getting up, Shanks takes out the person who dominated Arguile, freeing him of the spell. Glad and Galahad make quick work, tearing through the cannon fodder; eventually, the fighting comes to a close, and the party can heal and rest up until the Force Cage spell drops on the Executioner.It's during the conversation with The Executioner that the party discovers where her allegiances truly lie. Galahad takes the lead in this and not only gets her true name, but also her complicated situation and turn of events that led her into league with The Constable. At one point, it's revealed that The Judge is the surviving member of the Gargoyle Unit, the same elite unit that Argo and Antonius are surviving members of in this reality. With revelations made, cards put on the table, and newfound allies formed, the party works with The Executioner to be a double agent, which will prove very tricky and dangerous but would be a huge win for them and their allies who are trying to stop the planes from colliding.There's so much happening, and that's where we pick up…Find out what happens next in this episode of the Medusa's Cascade: Collateral Damage! Theme Music is written and performed by EfflorescenceMixed by Thomas Lapierre IIITitle Card by Pierce Graphics Check out the show at themedusascascade.com Hosted on Acast. See acast.com/privacy for more information.

In this episode of The League of Gentlemen we dropping in at the Hotel Windermere where aging swingers Alvin and Sunny are hosting a special gathering... which ends badly to say the least!

Recorded live at Retro Expo in Plano, Texas in July 2024, Craig sits down with Harry Hamlin for a lively and genuinely engaging conversation that spans Mayfair Witches Season 2, his unforgettable Clash of the Titans stories, his viral Bolognese, and the acting lessons he learned from Laurence Olivier. Standing His Ground: Hamlin recounts how he protected Clash of the Titans from bad creative choices — including locking himself in a trailer to keep the Medusa scene intact. Mythology & Craft: With a Yale background in mythological psychology, he explains how Joseph Campbell shaped his approach to Perseus (and why the movie's mythology is “all wrong”). Finding Courtland Mayfair: Inspired by Laurence Olivier's “green umbrella,” Hamlin shares how a simple pinky ring instantly unlocked his Mayfair Witches character. “When they hired me, they didn't know they were getting a badass.” - Harry Hamlin Want more exclusive interviews and behind-the-scenes stories from the people who bring your favorite characters to life? Subscribe to our podcast on Apple Podcasts, Spotify, or on your favorite podcasting platform for more behind-the-scenes stories and exclusive interviews.

Watch the full episode on YouTube:We first covered Baseten last year when DeepSeek mania was at peak hype. Now they have raised a monster $13B round and become one of the new cohort of AI Infra decacorns that are (with Nvidia, Intel, and the semis complex) chief beneficiaries of the Inference Inflection. We return to Baseten at the peak of the 2026 edition of Open Weights debate. Ali has published a viral breakdown of Kimi K3:And since you last saw him, Philip has spoken at AI Engineer and written the definitive book on Inference Engineering spotted all over SF:Three years ago, inference engineering barely existed as a category.Today, it is one of the most critical disciplines in AI. Inference engineering inherently tackles a different question than standard model training: “How do you turn those weights from training into a product that is fast, reliable, and affordable at scale?” Focusing on these creates an entirely new optimization problem.In one recent GLM-5.2 experiment, quantizing more of the model actually preserved its benchmark quality while increasing throughput by 20%, because the errors introduced in different layers could cancel each other out.Inference is no longer just the final step after training. It is becoming its own engineering discipline, with its own research problems, infrastructure, and increasingly specialized roles.In this episode, Baseten's Philip Kiely and Ali Taha join swyx and Vibhu to explain what actually happens after a new open model is released and what it takes to turn “we generated a token” into a fast, reliable, production-ready API.We go deep on cache-aware routing, disaggregated prefill and decode, quantization, speculative decoding, KV-cache movement, model parallelism, GPU kernels, and the race to make frontier models up to 10× faster. Philip and Ali explain why inference optimizations can still produce gains of 20%, 100%, or even 200%; how quantization errors can cancel one another out; why identical weights can behave differently across clusters; and how Baseten grafted a Kimi vision encoder onto GLM-5.2 without changing the underlying language model.The conversation then expands beyond LLMs into NVIDIA Dynamo, mega kernels, Rubin, AI-specific chips, local inference, video generation, diffusion versus autoregressive models, and the enormous compute barrier to generating coherent long-form video. Finally, we explore the convergence of training and inference, continual learning through persistent KV cache, and the emerging loop where models help optimize the infrastructure that runs them.We discuss:* What happens when a 200,000-token request enters an inference system* Cache-aware routing and reusing previously computed KV cache* Why prefill and decode are increasingly handled by different GPUs* When dedicated deployments become cheaper and more reliable than shared APIs* How speculative decoding uses a smaller model to accelerate a larger one* Tool calling, structured outputs, and what LLMs actually do* What it takes to support a new open model on day zero* Grafting Kimi's vision encoder onto GLM-5.2* Retrofitting inefficient model layers with components from other architectures* Why models sometimes collapse into repeating the same token* How hardware, kernels, and race conditions create nondeterministic failures* Preserving model fidelity while making inference faster* How quantization errors can cancel each other out* Why inference optimizations still deliver gains of 20%, 100%, and 200%* How optimized serving can make a model up to 10× faster* NVIDIA Dynamo, KV-aware routing, and distributed model serving* Speculative decoding the speculative decoder* Why local AI is about making models less dumb while data-center AI is about making them less slow* Tensor, expert, and pipeline parallelism across GPUs* Hardware-aware model design, auto-tuning, and the case against mega kernels* Rubin and why inference is becoming a systems problem* Whether modern GPUs are evolving into programmable AI ASICs* Why enormous models like Kimi K3 require GB300-class hardware* Why open-source video generation still trails Veo, Kling, and other closed models* The quadratic attention bottleneck behind long-form AI video* Autoregressive video, real-time generation, and compounding quality drift* Why future video systems may combine autoregressive and diffusion architectures* Training for inference and inference for training* Continuous post-training, deployment, evaluation, and improvement loops* How GLM-5.2 helped optimize the kernels serving GLM-5.2 itself* Why faster networking could unlock dramatically faster decoding* Continual learning, KV-cache compaction, and persistent model memoryShow Notes* How to build a day-0 API for Kimi K3* 22580: From GPT2 to Kimi3, ExplainedPhilip Kiely* LinkedIn: https://www.linkedin.com/in/philipkiely* X: https://x.com/philipkiely* Inference Engineering: https://www.baseten.co/inference-engineering/Ali Taha* LinkedIn: https://www.linkedin.com/in/aliestaha/* X: https://x.com/waterloointernTimestamps00:00:00 Introduction and the 200K-Token Prompt00:03:18 Dedicated Deployments, Speculative Decoding, and Tool Calling00:11:26 Launching Production-Ready Open Models00:19:06 Model Retrofits, Failure Modes, and Nondeterminism00:28:22 Quantization and Canceling Errors00:32:15 The Race to 10× Faster Inference00:40:48 Dynamo, Speculation, and Local vs. Data-Center AI00:50:18 Model Parallelism, Auto-Tuning, and Mega Kernels01:00:55 Rubin, GPUs vs. ASICs, and Custom AI Chips01:10:03 Giant Models and the Limits of GPU Memory01:12:42 AI Video, Quadratic Attention, and Autoregressive Generation01:21:47 Audio, Images, and Diffusion Models01:27:32 Training, Self-Optimizing Models, and Continual Learning01:40:06 Closing ThoughtsTranscriptIntroduction: Baseten, Waterloo Intern, and Inference EngineeringSwyx [00:00:00]: Okay, we're here in the studio with Philip, old friend from Inference Engineering, the book, as well as Baseten and everything that you've done, you and I have done before, as well as Ali. Welcome.Ali [00:00:15]: Pleasure to meet you.Swyx [00:00:15]: Waterloo intern.Ali [00:00:16]: Waterloo intern, always.Swyx [00:00:17]: When did you get “Waterloo intern” as a handle?Ali [00:00:19]: As a handle? Oh.Ali [00:00:20]: I think the rebranding happened mid-March. When I saw it was open, I was like, “I have to take it. Up for grabs.”Philip [00:00:26]: The problem is that Ali is really good at his job and is not gonna be an intern much longer.Philip [00:00:30]: So we have to figure out who's gonna get the handle.Ali [00:00:33]: Well, I'll pass the torch over to the next intern.Swyx [00:00:34]: Oh, okay. It can be, like, you just pass it to another Waterloo grad.Ali [00:00:37]: To another Waterloo intern. No, bruh.Philip [00:00:39]: Yeah.Ali [00:00:39]: Intern.Swyx [00:00:40]: Intern, yeah.Ali [00:00:40]: And no.Philip [00:00:41]: You gotta get an intern from Waterloo.Ali [00:00:42]: Yeah, I've gotta get an intern from Waterloo.Swyx [00:00:44]: Right.Ali [00:00:44]: But they have to follow the path.Swyx [00:00:45]: Oh, it could, but it could come from Baseten, so it's like whoever Baseten gets from Waterloo.Ali [00:00:48]: Right.Swyx [00:00:49]: Has the title of Waterloo.Ali [00:00:50]: It stays in the ecosystem.Philip [00:00:51]: Exactly.Ali [00:00:52]: Halfway through the internship, you either get it or you're out.Philip [00:00:55]: You should also do, like, a big graduation ceremony where you change the handle.Ali [00:00:59]: Just say it.Philip [00:00:59]: For everybody.Swyx [00:01:00]: You guys are good at ceremonies, clearly. We had a nice launch of the book, very successful. But before we get into all that, I wanna start off with a fun question for you. Okay, you're an expert inference engineer. What happens when I send a long query, say two hundred thousand tokens into Baseten's inference? What's the process of query through GPU model routing, balancing, all that? What is all the stuff that we don't think about?Long Context Requests, KV Cache, and Cache-Aware RoutingPhilip [00:01:26]: With a long query specifically, the first thing that I'm gonna ask is, “Have you sent me this query before, or at least part of it?” and I really hope you have, because it's gonna be a lot easier for me and a lot cheaper for you. So the first thing that we're gonna look at is some cache-aware routing, where we're going to see, we probably have a number of instances, a number of replicas up serving whatever model you're hitting. We want to send this one to something with, number one, available prefill workers, and number two, ideally some cached input already there so that we can skip prefill on at least part of these two hundred thousand tokens. If you're doing two hundred thousand tokens, it's probably coding or a multi-turn agent or something where you would expect to have that cached. If you don't, we're gonna have to send it to a prefill worker. We've at least on certain models disaggregated prefill and decode, so you're going to have one set of GPUs that's solely going to process the input, create the KV cache, and get you your first token, and then that's going to be passed over to a separate set of GPUs, which is going to run decode. We're going to iteratively make those tokens. We're probably going to have some speculator model in front of that. I'm going to assume that you're doing coding, and because of that, our speculator model, which assumes you're doing coding, is gonna have a high draft token acceptance rate. If I'm wrong and you're asking me to summarize every Harry Potter book, it's gonna be slower. And then we stream that output to you and account for it, charge you, a couple of pennies and say, “Hey, would you like to send another one?”Swyx [00:03:04]: Except Baseten doesn't charge by pennies.Philip [00:03:07]: Well, yeah, we charge. I'm assuming that we're talking about the public model APIs. If you are setting up a dedicated deployment, then yeah, it's not pennies.Public APIs vs. Dedicated DeploymentsSwyx [00:03:18]: Yeah, one of the key differentiators when I was talking with Baseten initially was that people who want very high volume just need to rent by the box, ‘cause then it's up to you to figure out how to saturate the box.Ali [00:03:31]: And more often than not, it's, like, way cheaper if you're pushing, like, millions of tokens per hour, if you just pay per hour instead of pay per token.Philip [00:03:37]: Yeah, they do. I think that we've increasingly seen a lot of demand for the pay per token APIs, just because everyone wants to try open models, and then once they find a use case that's really sticky, then they move over to dedicated.Swyx [00:03:51]: Is there a best practice on when it's time to swap over?Philip [00:03:54]: Couple reasons. Yeah, reliability, that's a big one, right?Ali [00:03:57]: Like, if they have a very specific use case, they want you to train something specifically for them, like they want their own spec dec, for instance, for their own traffic.Swyx [00:04:04]: Spec dec is speculative decoding.Speculative Decoding and Custom SpeculatorsAli [00:04:05]: Speculative decoding, yeah.Swyx [00:04:07]: You have to explain.Ali [00:04:07]: Sorry. Like, speculative decoding is like, if you have a huge model, right? And so the model is going to be generating one token at a time every single turn, every single forward pass. So we attach, like, this little, like, parasite, like this layer that goes on top of the model, and this model just has to predict. It does three very fast autoregressive forward passes, and it will predict, like, three certain tokens, and then you do one forward stage over the entire original model in order to see if those predictions were correct or not, and then you accept them or you reject them. Now, this draft model is traffic specific, so if you, like, Philip said, if you're summarizing Harry Potter books, I can train exclusively that draft model on Harry Potter books, and I can guarantee you that I'm gonna accept the three tokens every single time. And so with that case, I increase your decode speed. I wouldn't be able to provide this to you if you're a shared endpointSwyx [00:04:53]: YeahAli [00:04:53]: ‘cause I have no idea if you're doing Harry Potter, if you're doing coding, if you're doing English. We don't know. Also, there was a thing in the book that mentioned that if they really cared about a specific threshold, chapter four, I think. Do you remember that?Philip [00:05:06]: Yeah. The things that you can do is you can set a specific, like, batch sizing, a specific, like, parallelism strategy if you're trying to optimize for, like, throughput versus latency. You can. Maybe a NVFP4 quant doesn't pass your benchmarks and you wanna run a model at higher precision, you could do that. There's just a bunch of reasons why you might wanna have your own endpoint and the biggest one, of course, just being, like, you don't have to deal with someone else doing a hundred million tokens of benchmarking traffic at the endpoint when you happen to be trying to serve your users.Swyx [00:05:40]: Yeah. I think one thing that is. That is a classic journey. Like, it's people is asking the, what happens when you type Google into the browser. Tool calling, is that just, you're generating JSON or is there more complication beyond that?Tool Calling, JSON, and Structured OutputsAli [00:05:58]: Certain customers that we have, they have their own post-trained models, and so they demand a tool calling that's not just, like parse a file or go find the weather. It's something that's very specific and you have to do post-training on this. And if the post-training on the model is not good or if the quantization after the post-training to get the inference to be fast, the model will struggle reading the JSON file and reading the tool calling. But it doesn't require its own like sandbox. It's not like it's going to use that tool calling to like escape a sandbox or like it doesn't have to be contained. It can just be a normal dedicated deployment. The challenge with tool calling more and more seems to be that the companies want certain tool calling which is a very sensitive thing to train. And because you're dealing with all of the JSON outputs, if it doesn't like close the end of the request in a very certain manner, you end up with a model that did the tool calling and like the thinking and so as a result of that, it didn't see the result and just hallucinated the result as it decoded. That seems to be the most challenging thing with tool calling, not really the sandboxes model.Philip [00:06:56]: Yeah, that's a challenge on the training side and then on the inference side, there's work that you can do to scope the possible output. So we published this at this point close to two years ago, the solution to this problem which is you make a state machine and you use that to constrain the output to a specific format. So this is the structured output problem. If you remember backSwyx [00:07:27]: Yeah, the specific grammar is,Philip [00:07:29]: Yeah, exactlySwyx [00:07:30]: GML had this thing.Philip [00:07:31]: Yeah. So it's like the old-school “make sure this is only JSON”, return only JSON orSwyx [00:07:38]: YeahPhilip [00:07:38]: Grandma's gonna die type of prompts.Swyx [00:07:39]: Is it BNF grammar? At some point OpenAI had released a thing that was like, yeah, if you want to constrain your output, write BNF grammar, back as NOR.Philip [00:07:47]: In our inference system, it's just a specified output format. And you get the guarantee that your output's gonna be structured along that format. And so applying that to tool calls can like help cut down on. You can still call the wrong tool or call no tool. It doesn't solve the certainty problem but it at least solves the output structuring problemSwyx [00:08:10]: YeahPhilip [00:08:10]: Within tool calls.Swyx [00:08:12]: And MCP is just another form of tool, right.Philip [00:08:14]: Yeah, exactly.Swyx [00:08:15]: As far as there's no special thing there.Philip [00:08:16]: The thing I'm always like explaining to people is the LLM is not capable of doing anything. It's only capable of making suggestions of what to do and then if those suggestions are formatted in a certain way and applied to a system that knows what to do with them, then an action occurs.Swyx [00:08:32]: Yeah. Part of the fun stuff is, this is solved outside of tool calling too. Like in an agent loop if the output is not correct or you're right, like reasoning, tool calling was done in the reasoning trace, just be like, “Oh, I don't know what to do. Let me just try again.” And it might get there after a few tries. And on your point of training, sometimes this is harder in smaller models, so you don't have the same exact quality outputAli [00:08:56]: Right.Swyx [00:08:57]: When you just swap from a big model, right?Ali [00:08:59]: Yeah. I will say that, before, I think we need to go back to inference engineering proper.Ali [00:09:04]: But, I had expected that something would replace JSON because it's hard to stream JSON ‘cause JSON must be complete and you must have open and close brackets and everything. So it's hard to parse something or validate something while it's being streamed. So people invented all sorts of things that are like, I forget the name of some of these alternatives, but it's something like TOML, something like YAML. But JSON seems to be dominant still.Philip [00:09:30]: The JSON outputs aren't that long, right? Like you could have a long-- ‘cause tool calls also contain the arguments in them and perhaps for a certain tool you might pass like a very long argument. But my impression of the median tool call is that it's a relatively small number of tokens, right? So I would expect that speculators are generally fairly good at something as formatted as JSON. And so you would have like a pretty fast decode step there and that the streaming wouldn't be as valuable, but maybe I'm wrong about that.Ali [00:10:02]: I think you're also bounded by the software or that the model is gonna integrate with if the software is built with JSON for the tool calls or if the company that you'- if your customer says that this is how our software works and our tools are interfaced with JSON, you can ask them to like, change their software and say like, “Yeah, this is gonna be better for the model.” but like with the right training shouldn't be that much of a difference. Also more profitable if it outputs more tokens probably.Swyx [00:10:25]: Depends on your business model.Swyx [00:10:27]: It really depends. But I will say that, as a writer with like experience a lot with generated output, I do try to move from text to JSON text which is very long JSON, right? Like there's paragraphs in every field because I'm trying to structure it, right?Philip [00:10:44]: Right.Swyx [00:10:44]: I want you to first make factual statements, then make opinions then make bullet point summaries, have dates, have entity references have your sources for references, all these things. Anyway, so these are things that like I think people who really experiment with structural output have to really care about. But, let's, let's recurse up the stack a little bit. Before we started recording, you mentioned something really cool, which is that there's a lot of engineering that-- inference engineering that goes on when a new model provider releases a new model, right? So let's call it GLM-5.2, Kimi K3. I had previously assumed, especially if it's like, well, GLM 5 to 5.1 to GLM-5.2, like that you've supported them before. Is it that much work?What It Takes to Support a New Open ModelAli [00:11:26]: It's a lot of work.Swyx [00:11:28]: Yeah. Okay. So like, a lot of people, all you guys, right whenever a new model launch like, people rush to say like, “Oh, Hugging Face supports this, Fireworks supports this, Spacetime supports this,” and I'm like, “Yeah, of course we support it.” But what goes into that? What goes intoPhilip [00:11:40]: I think it's more than just support it too, right? It benefits the consumer a lot. Like I think it was with Kimi K2.5 or GLM-5.2 the latest, there was an inference war, right? X provider is at 90 tokens a second. The next day we're at 150. The nextSwyx [00:11:55]: I kinda kicked that off with the GLM-5.2.Swyx [00:11:58]: I wrote a Twitter article about. It got like half a million views,Ali [00:12:02]: Based on being numberSwyx [00:12:03]: YeahAli [00:12:04]: Or it's for something else.Swyx [00:12:05]: Yeah. Which,Ali [00:12:06]: Oh my GodSwyx [00:12:07]: Which then got everyone really excited about, hey, how can we, bend tracks a little bit further and,Philip [00:12:14]: There's a difference between support the model, as in I can make a token out of this model, and support a model, as in I have a production-ready API from this model.Philip [00:12:26]: Getting to the point of I can make a token out of this model is not that hard because generally the, open source inference engines, vLLM, SGLang of the world oftentimes even receive weights ahead of time, maintainers do, or the people making the model merge PRs to ensure support. So you generally can, just get it working on the standard open source stack without too much pain in most cases. The challenge is, every inference company is gonna have own proprietary stack. Some open source components, some in-house stuff. And for any arbitrary model, there's going to be some new stuff. Sometimes you get lucky, like K, two five to two six was, like, pretty similar.Quantization, Speculators, and Production ReadinessAli [00:13:16]: Yeah. It was pure continued post-trainingPhilip [00:13:18]: YeahAli [00:13:18]: If I remember correctly.Philip [00:13:19]: Even in those cases, there's still stuff you have to do. You have to redo the quantization work. You're taking the model from. Generally, these models are not released in NVFP4, and we want them to be in NVFP4 for maximum Blackwell compatibility. So we have to perform that quantization, and, calibrate the quantization to make sure that we're not causing any regression in the model's intelligence. And then we also have to train the speculator, as we've talked about. Generally, we have. We have ZDR, zero data retention on our model APIs, so we don't know exactly the traffic that people are sending us, but we know what's popular. We know that coding use cases are popular. We know that agents, agentic use cases are popular. So we can get public data sets that are representative of that traffic and train general speculators. Now, with speculators today, you need to train the speculator using the base model itself because you're getting hidden states out of the model from running inference on these specific prompts, and that is the training data you use to create the speculator. So there's that process which you need the real model weights for. And then there's of course just the process of, standing up all the infrastructure behind it, loading all this stuff, testing it. And then when there's a new model with a newer architecture, I think that, like, the DeepSeek models tend to be the most challenging as they have, like, the most novel architectural stuff going on, model after model. But every new model has something. Kimi K2 had. Oh, sorry, GLM-5.2 hadAli [00:14:53]: Sparse attention.Philip [00:14:54]: Yeah,Ali [00:14:54]: YeahPhilip [00:14:54]: the DSA.Ali [00:14:55]: Right. Which is brought from DeepSeek.Philip [00:14:57]: Yeah. AndAli [00:14:59]: So you can copy-paste then?Philip [00:15:01]: It kindAli [00:15:01]: I don't know how this works.Philip [00:15:02]: So, like we had to, like, build support for that into our runtime. And you're right, like it is really interesting the way that all of these open source labs borrow from each other. For example, like GLM-5.2 doesn't have vision. So something that, Haley, a guy on our team, if we could take a look at this, he, like, grafted the Kimi vision encoder onto GLM-5.2.Retrofitting Vision into GLM-5.2Ali [00:15:27]: We'll be training the projector.Philip [00:15:28]: Exactly. So if you think about, like, the encoder, there's the encoder, which is the part that looks at the image and turns it into latent information, and then there's the projector which likeAli [00:15:38]: You can say latent space. It's okay.Philip [00:15:41]: And then there's the projector that maps it onto, the model itself, and then there's the model weights. You don't wanna mess with the model weights because you run a chance of making the model dumber at something else for the purpose of giving it vision. So instead, Haley started with just a projector, which is only a handful of millions of parameters.Ali [00:16:02]: That would be, yeah.Philip [00:16:02]: Yeah.Ali [00:16:03]: Can you show the training one?Ali [00:16:04]: Like the way it groksPhilip [00:16:05]: YeahAli [00:16:06]: Very interesting.Philip [00:16:06]: And maybeAli [00:16:07]: That right therePhilip [00:16:07]: Maybe Ali, you should take it from here. You've got a betterAli [00:16:10]: Ooh, double the sandPhilip [00:16:11]: Understanding of this than I do.Ali [00:16:11]: Yeah. You can see, like, he. The way he trained this is really cool. At the beginning, he was training it using just like, “Here's a picture of a mountain. Can you describe what's in this mountain?” And that caused it just like the first, learning walls. Like here you can see this all we're trying to teach it is to translate the encoded. Like it's already taken the encoder from Kimi K. It's taken the image. It'Philip [00:16:31]: Yeah. FrozenAli [00:16:31]: FrozenPhilip [00:16:32]: With adapter.Ali [00:16:32]: Exactly.Philip [00:16:33]: Yeah.Ali [00:16:33]: So the brain is frozen and the eyes are frozen. It's just we're tryingPhilip [00:16:37]: AlignAli [00:16:38]: Interconnect between the eye and the brain, right? So the projector. And so you take the tokens and then he's like, “Oh, can you describe what's in this image?” And he's like, “Oh, it's a mountain,” or it's a person or it's a human, whatever the case is. But that didn't cause complete understanding. So he changed it such that every image was associated with a data set of questions. Like, does this image have a white male? Does this image have birds in the top corner? Does this image have a scientist in it? All of that stuff. And it would have to answer questions correctly. And using not just training on describing an image, but being able to answer question, another question, answer over time. Like you can see the grokking, which is like genuinely insane, that retrofitting vision into a large LLM can learn to that extent. And even for images that it doesn't perform well on, for instance, if you ask it a picture of like Stephen Hawking, “Who is this?” Maybe it doesn't get it, but it will say something like, “This is Albert Einstein.” Like it still understandsPhilip [00:17:25]: Close enoughAli [00:17:26]: That this is a scientist who is a man who has, some significant achievements, all that stuff. So that's like really cool.Philip [00:17:32]: Yeah. So, we've covered Hao Tian before, who the author of the LLaVA paper that did this, a while ago. And I think that's very foundational work for anyone who hasn't done vision work before.Ali [00:17:41]: Same with the CLIP and MetaCLIP, where you go from just captioning to building out questionsPhilip [00:17:47]: RightAli [00:17:47]: Off the image and how much better you can get performance.Philip [00:17:50]: Right. Right. Right. Yeah. But what's, what's so exciting about this is if you look at a model like this. Now, this is a little bit more of a research project. It's not. It got to 56% on MMLU Pro, I think. So not quite frontier. But if you're running this model, you haven't suffered any loss on your GLM-5.2 quality. If you don't have an image, it'll just behave exactly the way it used to. And ultimatelyAli [00:18:14]: Which in the inference code you literally do not include the other part, right?Philip [00:18:18]: Yeah. You would just skip the encoder if you don't have an image input.Ali [00:18:22]: Okay.Philip [00:18:22]: Just confirming.Philip [00:18:23]: YeahAli [00:18:23]: Does it affect a lot on the overall inference side? Like you're not adding much, you're adding a very small vision encoder. These are typically likePhilip [00:18:30]: They're super fineAli [00:18:31]: Less than a billion parameters, right?Philip [00:18:32]: Yeah. It's, - There's a little bit less standardization among vision encodersSwyx [00:18:37]: YeahPhilip [00:18:37]: So the support matrix can be a little bit, sparser. But overall, yeah, it's a pretty, it's a pretty minor component of the overall system. And ultimately what you get out of the system is all of a sudden you have Kimi Vision, GLM weights, and DeepSeek attention all in one model.Open Source Model Grafting and Franken-MergesPhilip [00:18:56]: And that's, I think, a lot of the power and beauty of open source, is that you can take all of these different components and combine them together into a system that's better than anyoneSwyx [00:19:05]: YeahPhilip [00:19:05]: Can be individually.Swyx [00:19:06]: People used to say that you would also do Franken-merges where you would take likePhilip [00:19:10]: YeahSwyx [00:19:10]: Layers from each model.Swyx [00:19:11]: Does anyone do that anymore?Ali [00:19:13]: Well, to your point previously when you were mentioning like, the work that goes into supporting a model when it first comes out, like GLM-5.2 or MiniMax M3 or whatever the case is. Sometimes you do have to like, you do have to switch out some things. Like, for instance, the MiniMax M3 head uses full attention, and with full attention you end up with this like insane bottleneck in spec dec ‘cause you're doing auto-regressive token generation for three tokens, and you're doing this like N squared over all of the tokens that are in your sequence. Your KV cache is like very large because it's not sparse, it's not top K. So we find it better to like, okay, we're gonna replace this, we're gonna replace this layer with a layer from another model that's using like GQA, for instance. And then just with the right training, you can get it to have the same acceptance rate. So it is very possible to retrofit layers from other models and very much needed. If a layer is like inefficient, the training just becomes the challenge, like how do you ensure that you train it properly? Which again to your earlier point is like the mesh between training and inference. As in like you need very good training in order to do fast inference. That's like, I feel like more and more becoming true.Swyx [00:20:21]: Yeah. Anything else on the support side when you say like get it to fully production ready?Loop Detection, Race Conditions, and Non-DeterminismPhilip [00:20:26]: Yeah. I think that there's also a question of just, we can test a model to a pretty extensive degree, but we're trying to get it out quickly and then you see a bunch of other people test it and you get interesting results. There was an issue with, GLM briefly where we had some like mode collapses where it would just output the same token over and over again for certain prompts on certain temperatures. Like once you expose an endpoint to the real world, there's going to be, so many more varieties of things given to it that you're able to, discover and patch things. So it's not just a, day zero process, it's then like for the first week, for the first month, if a model remains popular, like how do you both fix bugs and then continue to push the envelope on performance?Ali [00:21:21]: What do you mean you don't want your model outputting S?Swyx [00:21:24]: Is there loop detection on that stuff, by the way? It still happens like quite a lot, which is surprising.Ali [00:21:30]: We have like we, in our endpoint, like if a model was to output the same token like four plus times, we just cut the generation. We say like, “Oh, sorry, this-- Like try again,” or like we will reprocess the request. ‘Cause we know then, like if it, like if, yeah, it's four times the same token, it's probably collapsed.Swyx [00:21:45]: Yeah. Is there a way to opt out in case I really want that?Ali [00:21:48]: You want that?Ali [00:21:50]: I think there's a way that we have to handle it. I'm not exactly certain, but I feel like in certain models, like when they output something like you can imagine, like a table for instance, and so they want, they wanna draw like 12 dashes and 12 dashes. Yeah, I think there's a way for that to happen. I think we only do it on certain tokens. Like we exclude certain special characters.Swyx [00:22:07]: Yeah.Ali [00:22:07]: So we only do it on like certain like S is the most common almost. GLM-5.2Swyx [00:22:11]: OhAli [00:22:11]: And I think it was DSV 4 as well. Like you'd just have like looping issues where like you literallySwyx [00:22:17]: ItAli [00:22:17]: Just have like S.Swyx [00:22:18]: Yeah. Is there a special, something special about S? No, just randomlyAli [00:22:21]: It just seems to be the one token involved.Swyx [00:22:23]: Yeah. And it'Philip [00:22:24]: Is thereSwyx [00:22:24]: And it's only temperature 0Ali [00:22:27]: NoSwyx [00:22:27]: Even at other temperaturesAli [00:22:27]: Even at like 0.9 or whatever, it will still, it will still collapse.Swyx [00:22:30]: That's weird, right?Ali [00:22:30]: It's, it is an inference problem to be honest, like a software problem. Like oftentimes, the image you run will-- like NVIDIA will release an image for instance, and if we will upstream the changes from their latest TensorRT-LLM image into our stack, we'll find that it fixes it. Or oftentimes this will only happen in an inference engine that you're using like SGLang. But if you were to switch to vLLM, that isn't the case. So it seems to be like an extremely like deterministic software issue and not really a model issue. It's not like a weights problem. Like I'- we'll say like, “Oh, it's a problem with the quant. We did PTQ wrong,” right? But that isn't, that doesn't make sense because the same weights used with a different inference engine does not repeat the problem. And sometimes it's, the kernels that are being used in the backend have like these very subtle sometimes race conditions, where if you were to use this model hosted on one cluster, you will never get this problem.Swyx [00:23:19]: Oh my God.Ali [00:23:19]: But if you host it on a different cluster, you will. And the reason is the KV cache transfer from a node to node in that one cluster is using a slower interconnect than the node to node in another cluster. So that exposes the race, whereas in another cluster it doesn't. So then you end up just like, okay, this model is not gonna be hosted on this cluster. We're gonna host it on, another cluster because that cluster exposed that problem. But then it ends up with like, okay, is it the software? Is it the model weights or is it the hardware?Swyx [00:23:42]: There is a thing about this with temperature 0 still not being deterministic, right?Ali [00:23:46]: Right.Swyx [00:23:46]: Mostly because of hardware. Even at temperature 0 same model, you won't always get the same output.Swyx [00:23:52]: Even-- But I'm surprised by the race condition one because, I thought PyTorch was a graph that like guarantees that you at least, execute things in the right order.Ali [00:24:02]: Well, yeah, true. Like I'm not, I'm not saying that there is. Like well, you have things like PTL optimizations where like you can start a kernel before the end of the previous kernel, and that's like ‘cause you want to do that because there'sSwyx [00:24:12]: It's like pipeliningAli [00:24:12]: Expense. Exactly.Swyx [00:24:13]: Yeah.Ali [00:24:13]: But it'- But you don't do it cleanly. Like you overlap a little bit of the execution. No, it is very possible that the kernel itself, like that one block that is supposed to be running in this instance of time, that kernel itself has a race condition. For instance, like a missing barrier. Like often if you're designing a kernel and you want it to make it to be very fast, if you don't test it extensively, you'll, you'll have certain threads access data points from registers before they've been written to by other threadsSwyx [00:24:36]: YeahAli [00:24:36]: For example, because like your barrier is wrong or your synchronization was wrong. But yeah, like the testing itself is very difficult in those like, andSwyx [00:24:42]: And there's no like borrow checkerAli [00:24:45]: What does that mean?Swyx [00:24:46]: Like Rust. Like the. If you're trying to have like memory safety It sounds like a comparable problem.Ali [00:24:52]: Well, yes, but you're working in CUDA, right, NVIDIA GPUs. Like- You just need a higher level language like modular Maybe that's what modular is supposed to do. I don't know.Quantization Quality and Vendor FidelityVibhu [00:25:00]: How do you see keeping quality of the model? So you talked about all these steps of, okay, you gotta do quantization, train your own speculative decoderAli [00:25:07]: RightVibhu [00:25:07]: Run on different hardware. Looking at other model providers, okay, you kicked off a inference speed race on the consumer end. What goes into keeping quality the same across them, right? Sure, you can run benchmarksAli [00:25:22]: YeahVibhu [00:25:22]: But, like, how do you determine how much quantization are there standards? What goes intoPhilip [00:25:27]: There's a few things on quality. Most inference optimizations are lossless. KV caching, for example. You are just recomputing or preventing recomputing the same values. Speculation, of course, if a draft token is wrong, it gets rejected. The main lossy optimization is quantization. And that really comes down to, number one, data format, number two, which parts of the model you choose to quantize, which layers, and number three, like doing a lot of calibration on the quantized weights, to ensure that you're preserving all the outliers. There's other tricks that you can do, though. A big one is long context, ‘cause one thing you asked at, right at the beginning is, “Oh, what's gonna happen if I send a 200,000 token request in?” So with a long input sequence, you need to, store a lot more information. You need to process a lot more tokens. And so even if a model has a context of a certain length, you might, as an inference provider, choose to build an API with a shorter context length, and of course a full length one as well. Because if someone doesn't need the full million token context, for example, you can get them better performance. I don't know if that's exactly like quality of the model. The way that I think about quality is to what degree are we faithfully serving the original model? If you think of a golden implementation of a model that performs exactly the way the model is designed to perform, I think of quality as how close are we getting to that, 100% fidelity of the model.Philip [00:27:13]: You can also, of course, think about quality from the training side and how do you push yourself past 100%. But when I think about purely inference optimizations, it's getting faster while staying as close to that 100% fidelity mark as possible. And certainly our standard internally is that, like you should not be able to tell the difference between our API and a, official API. I think Kimi in particular does a good job of vendor benchmarking hereAli [00:27:41]: YesPhilip [00:27:41]: Where they haveAli [00:27:42]: They released an actual vendor benchmark.Philip [00:27:43]: Exactly, yeah.Ali [00:27:44]: ‘Cause they accused, some people, Amazon? There was some provider that was not doing very well on Kimi's benchmark.Philip [00:27:50]: Yeah.Philip [00:27:51]: So, with Reflect we probablyVibhu [00:27:52]: This was a long time ago, right?Philip [00:27:54]: No.Ali [00:27:54]: Yeah, like threeVibhu [00:27:55]: They alsoAli [00:27:55]: Four, five months agoVibhu [00:27:57]: This also happened with, I don't remember which model, but they pulled out quite a few, and then they started a whole chart about this. It might have beenPhilip [00:28:03]: Kimi Vendor Verifier.Ali [00:28:04]: Yeah.Philip [00:28:05]: Yeah.Ali [00:28:05]: Yeah, ‘cause you, ‘cause you'd be pissed, right? Like if you'Philip [00:28:07]: Yeah.Ali [00:28:07]: If like if I'm a consumer and I'm using like Amazon's endpoint for instance, and I've used Kimi and I'm like, “Oh my God, like this is bad,” I'm not gonna say, “Oh, Amazon quantized the model in a bad way.” I'm gonna say, “Oh, Kimi sucks.” Right?Philip [00:28:17]: Yeah.Ali [00:28:17]: So it seems like that makes sense.Philip [00:28:19]: Yeah, they care. They care.Vibhu [00:28:21]: Justifiably.Ali [00:28:21]: Yeah, justifiably.Vibhu [00:28:22]: This is probably a stupid question, but just checking, has anything improved from main quantization?Philip [00:28:28]: Yeah.Vibhu [00:28:28]: Like, is quantization always strictly worse?Ali [00:28:30]: Well technicallyVibhu [00:28:32]: NoAli [00:28:32]: It's a lossy. QuantizationPhilip [00:28:33]: YeahAli [00:28:33]: Is a lossy, it's a lossy implementation.Philip [00:28:36]: Speed improvesVibhu [00:28:36]: Speed improves.Ali [00:28:37]: It the number, likeVibhu [00:28:38]: No, I' always look for inverse scaling laws.Philip [00:28:40]: Yeah.Ali [00:28:40]: Yeah.Vibhu [00:28:40]: This is something I learned from Noam Brown, where like things that normally act in one direction sometimes do.Philip [00:28:45]: Well, technically when you run a benchmark, because these models are deterministic, sometimes your,Ali [00:28:52]: YeahPhilip [00:28:52]: NVFP4 quant is like, two basis points higher than yourAli [00:28:56]: No, it's noise. It's noise.Philip [00:28:57]: Yeah, exactly. I'm like, yeah, it's, it's within. That's why I always say within margin of error.Philip [00:29:01]: And I stopped saying that because everyone assumes that what is, well, within some margin of error, we're barely inside of that to the worst, so we're saying. But yeah, sometimes it's just like, gives you a higher output score. But like Ali said, that's noise. To my knowledge, you're not necessarily making the results better. You're just trying to, again, like keep your fidelity as close to 100% to the original model.Layer Selection, KL Divergence, and Better QuantizationAli [00:29:27]: There is, to your point, research that we did on MP. I don't know if you are able to pullPhilip [00:29:31]: YeahAli [00:29:32]: A tweet we did. One of our research interns, Joshua, I think it's a tweet on how we have 20% better quantized GLM-5.2 than NVIDIA. Essentially what we found throughout like this month research is, okay, quantization is a lossy. It's. You're compressing the data from, occupying 16 bits to occupying, four bits, for instance. And so you're losing some information, and you're trying to minimize that. And so when I say that I'm gonna quantize the model, my job becomes how do I find the layers that I can quantize, and how to find the layers to not. For instance, with image models, I don't quantize modulation layers, and I don't quantize out projections because those two are. Like out projection is what you see as the user. Modulation is what the model sees or understands. Right, exactly. And so to his paper, do you have the. It doesn't have the. Yeah. It's a long paper. I don't know if I can findVibhu [00:30:25]: If there's a part to search or it's probably in the thread.Ali [00:30:28]: It's probably in the thread.Vibhu [00:30:29]: Yeah.Ali [00:30:29]: But the long and the short is it is very possible that quantizing more of the model makes the results. Like if I have a model that I quantize layers one, five, and 10, and another model where I only quantize layers one and It is possible that the model in which I quantized more information is going to perform better because the quantization errors have canceled out. And so what Joshua showed in his mathematical proof where he had like a verifier in, is that you can predict which layers are going to have quantization errors that will cancel out with each other, and you choose to quantize those layers. And so the result of doing this mathematical quantization is you end up with a model that's 20% more quantized than another provider, so you get 20% more throughput of it because there's more layers than running an NVFP4, and your quality is better than that other quant because the layers that you chose to quantize have their errors cancel out, like one layer skewed to the right one layer skewed to the left, one layer skewed to the right. Your final logits distribution is more similar to the original distribution of the model, so you have better fidelity. And so the way we proved this was with KL divergence. So instead of just scoring on the benchmarks, we scored the KL divergence between the logit distribution of the quantized model and the logit distribution of the original full precision model, and we showed that with this technique we get. If your probability distribution on the logits which token it wants to select is more of the same as the original model, you're probably gonna end up staying true to the original model. So yeah, so it seems like previously before this, it seemed like the industry was, well, the more you quantize, the worse it's gonna be, ‘cause the more loss you introduce. That's not exactly, not necessarily true. So yeah, doesn't improve it, but can cancel out.Philip [00:31:57]: I think it might be this, but reminds me a good bit about pruning where you can prune off certain layers.Philip [00:32:03]: But very interesting. Didn't know this was a whole paper you guys put out.Ali [00:32:06]: It's. Fun fact, it was originally 72 pages, this paper, and then we decidedPhilip [00:32:11]: WowAli [00:32:11]: We can't tell. We couldn't release it. So it's now 45.Swyx [00:32:15]: Still 39 pages, so very substantive. We talked about evals and all these things and, like what's possible in terms of speedup? Like it's like probably like the numberInference Speedups and BenchmarkingSwyx [00:32:25]: Thing that people do wanna care about, and it's something that you wrote about in your post. Like official API is 70 tokens per second, and you push it up to 90. Is that like a normal thing?Philip [00:32:36]: So what's cool about working in inference, the reason that I think inference is going to be a useful place to do engineering for a long time, is that if you look at highly optimized domains like, say, finance, if you're in finance, you measure how much better you got in basis points. It's like, “Oh, I got five basis points better, like twentieth of 1% better,” that's huge news because everything is so optimized. When we publish optimizations, it's 20%, it's 100% it's 200%. So there's still probably like a lot further to go, honestly. Like you'll, you'll know that inference is pretty much solved when researchers start publishing about how they got 1% faster at something.Swyx [00:33:19]: Which by the way, because I am from the finance background, in the ‘70s, that was the margin at the time. When you did quantitative finance research, you would findAli [00:33:27]: And like 20%, tens of percent.Swyx [00:33:29]: That's. Yes.Philip [00:33:29]: Yeah.Swyx [00:33:30]: And now it'Philip [00:33:31]: Tiny fractionsSwyx [00:33:32]: For those people interested, look up Andrew Lo's paper. He had a really interesting illustration of quant, stat arb, distribution, narrowing down from like those kinds of 20% differences in the ‘70s, down to nothing today, which is very cool.Philip [00:33:48]: Exactly, and we're at the beginning of the same type of thing. Now benchmarking is hard. I think anyone will tell you that, and benchmarking provider speeds is hard because there's so many variables that go into it. What hardware are you using? How much load do you have on the system? What's the exact nature of the prompts and input and output sequence lengths? All that stuff. But overall, when you start stacking these improvements, you're looking at multiples. You can look at it. The most common form, of course, is TPS, tokens per second, which is bad naming by us in the industry, ‘cause there's two tokens per second. There's tokens per second, the throughput number, and the latency number.Ali [00:34:31]: TTMT, yeah.Philip [00:34:32]: Like total tokens per second out of the, out of the GPU as a throughput number. Most people only care about tokens per second as the latency number, which we should call ITL, intertoken latency, but we don't.Philip [00:34:44]: Anyway, so you can imagine a standard API without many optimizations for a 1 trillion parameter model operating somewhere in the 30 to 50 tokens per second range for reasonable traffic profile. And we generally see the goal of, pushing to 10X that. But, not necessarily day zero, but by stacking enough optimizations, if you have, say like four optimizations, each of which doubles performance. Or sorry, three optimizations, each of which doubles performance, then you stack that up, that's an 8X gain. That's the order of magnitude that we're working with in this space. We're trying to make things substantially faster, not just go from like 70 to 90.Swyx [00:35:38]: Are you saying you've. You have done that?Philip [00:35:40]: So let's say you have as a reasonable baseline, 30 or 40 tokens per second. You can achieve 10X that. So like on GLM-5.2, if you run it unquantized, perhaps on H100s even, and you're just using an off-the-shelf inference engine with no particular optimizations, no speculator, nothing extra around like KV routing, no disaggregation, you're, you're probably, yeah, looking at that like 30 to 40. You think that's like a reasonable baseline?Swyx [00:36:12]: Right. Right.Philip [00:36:12]: To get to something like 10X, there's a lot of trade-offs that you're making. If we're running at more like a 300, 400 tokens per second range, you are using the best hardware possible. You have a optimized speculator. You have done all of your quantization work. You are Seeing a pretty high cache hit rate. You are running with a reasonably small batch size and a parallelism configuration that is tuned for latency versus throughput, but it is possible. So the spreads that you see if you, like, go on artificial analysis or you go on OpenRouter and you look at, the worst provider to the best provider, oftentimes can hit that range. 10X is of course very aggressive. It's oftentimes maybe more of a four to six times improvement. But that's the performance that makes us really excited, is when we can get these huge gains, not just go from 70 to 90 tokens.Stacking Optimizations: NVFP4, Speculation, and DisaggregationAli [00:37:19]: It's also, like, hardware dependent. Like, ifPhilip [00:37:20]: YeahAli [00:37:20]: If you have a thing where you're serving it on just, like, a node of H100s and then you throw, like, you shard the model across, like, four nodes of B200s. Like, you can definitely increase the speed with just throwing more hardware at it. Like, normalizing for the same exact hardware and the same number of GPUs.Philip [00:37:35]: Yeah. Then you're looking at, like, a two to 4X improvementAli [00:37:38]: Right. RightPhilip [00:37:38]: Depending on the inference optimizations. So yeah, it's. Some of it's, what's the call, and some of it's who's the driver.Vibhu [00:37:46]: If you break down the two to 4X, say the example is run GLM-5.2Ali [00:37:51]: YeahVibhu [00:37:51]: On B200sAli [00:37:53]: YeahVibhu [00:37:53]: Single node, right? What's, like, the cost trade-off for effort to get, like, the last bit of juice out versus what should people just think of, right?Ali [00:38:01]: Spectre quantization. Yeah.Vibhu [00:38:03]: Spectre quantization.Ali [00:38:04]: That's, that's, that's like 95%. LikeVibhu [00:38:06]: And how far does that get you? And how easy is that for the average person to do? So say right I wanna throw the weights of GLM-5.2 on a node of B200s, how easy is it to find speculative decoder- decoder model or already quantized model? How much work goes into it?Philip [00:38:23]: If you're doing it up front, it's quite a lot of work. If you're doing it today, there's going to be people who have published things that you can just, you can just grab some NVFP4 weights. You can grab a speculator. Yeah, if we're thinking about, like, what are the 2Xs we're stacking, going from, BF16 to NVFP4 is, it's not quite a 2X, right? It's like. I think it's about, like, 30 to 40%, from 16 to 8, and then another 30 to 40% multiplied from, 8 to 4. So that doesn't quite get you a 2X, but, like, roughly a 2X. Speculator, roughly a 2X. Disagg on top of that if you're able to get enough hardware and put enough traffic through it, another roughly a 2X. And then you add in some, double-digit percent increase from having just a better runtime with, the latest kernels and stuff behind it. And that's how it stacks up.Ali [00:39:21]: YeahPhilip [00:39:21]: So building each of those, like, building the, quantized weights is, for someone who really knows what they're doing, hours to days of work. Building the speculator, again, like, hours to days of work. And the, disagg setup, hours to days. Well okay, but like once you haveAli [00:39:39]: Once set up. Once set up. YeahPhilip [00:39:40]: Yeah, getting disagg working for the first time, I'm saying, of course, is very difficult.Philip [00:39:44]: The marginal implementationAli [00:39:48]: Like, if you're just grabbing, like if you are a person, like just a normal consumer who has access to, like, a node of B200s and you're wondering, “How can I just host it myself?” You don't need to quantize the model yourself. There's always gonna be, like, an open source quantized checkpoint. NVIDIA's gonna push one out if no one else does. You. Usually, the providers will have their own spec dec that they've trained as well. You don't need to train your own spec dec. You can just use that as well.Philip [00:40:09]: Yeah. Like, GLM-5.2 has its own MTP.Ali [00:40:13]: Right. Right.Vibhu [00:40:14]: What's multi token prediction?Philip [00:40:15]: Yes.Ali [00:40:16]: I'm justVibhu [00:40:16]: Can you explain that?Ali [00:40:16]: I'm just an expert.Ali [00:40:18]: I can do it for you in case I get it wrong?Vibhu [00:40:20]: No.Vibhu [00:40:21]: Yeah, you should correct if we're wrong, but their multi-token prediction can be used for self-speculative decoding.Ali [00:40:27]: I'm not sure. I'm not gonna correct that.Vibhu [00:40:28]: Okay. I'm semi-confident in thatAli [00:40:30]: Okay. YeahVibhu [00:40:30]: But someone can check. But it's useful to paint the story of, okay, not just the average person, but say a company wants to switch from serverless inference I wanna throw this up on. I wanna rent some GPUs, throw it up. These are the steps you take to do significantly faster than just put it behind vLLM.Ali [00:40:48]: Right.Vibhu [00:40:49]: I was waiting for a mention of Dynamo.Vibhu [00:40:51]: I feel like, that's supposed to be the baseline that you measure against.Dynamo, KV Routing, and Disaggregation ToolkitsPhilip [00:40:55]: I would think of Dynamo as less of a box system and more of a toolkit for building with. So when we talk about doing aware routing, when we talk about doing KV offloading, when we talk about doing, PD disaggregation, Dynamo fundamentally is. By the way, Dynamo is an open source library from NVIDIA.Ali [00:41:17]: We've done a pod with KylePhilip [00:41:18]: OkayAli [00:41:19]: Kyle Cranin.Philip [00:41:19]: Cool. So then your listeners know then that it supports all the different inference frameworks. And it is multi hardware, which is interesting.Ali [00:41:28]: But it's just a router, it's not like an optimizer layer.Philip [00:41:30]: Yeah. All it does, like, what Dynamo is good at, it is a library for moving information around your cluster, around your hardware. So if you have, KV cache on one place and you need it to be somewhere else, Dynamo coordinates NIXL for you to move that around.Philip [00:41:49]: That doesn't mean that, like, out of the box, you just say, “Pip install Dynamo,” and then you get, like, a massive performance speed up. It's more of a developer toolkit.Ali [00:42:01]: Yeah. I would have said it would. It comes with a set of defaults that you can then swap out.Philip [00:42:06]: It does. If the industry at large, I think, was, like, rolling out all of these deployments, standard, then I think it would be, like, a credible baseline. But, we've got to, we've got to benchmark against, like, what we're seeing in the wild.Speculative Decoding Methods: Medusa, EAGLE, n-Gram, and Spec-SpecVibhu [00:42:23]: I did wanna talk a little bit more about PD disagg, because that is probably, like, number three after quantized and speculative decoding. In your book though, I was just gonna pull out the book.Philip [00:42:31]: Yeah.Vibhu [00:42:32]: Like section 522 on Medusa, 523 on EAGLEPhilip [00:42:35]: YeahVibhu [00:42:36]: 524 on gram.Philip [00:42:37]: It's 55, would be disaggregationAli [00:42:42]: Yeah. Well, no, I just wanted to dwell a little bitPhilip [00:42:44]: YeahAli [00:42:44]: The other. Like, so what do you choose to include? What do you choose to not to include? Because there was all these other techniques.Philip [00:42:51]: Yeah.Ali [00:42:51]: Are these still relevant? Because I think they came out, like, a year and a half ago maybe.Vibhu [00:42:55]: Medusa is quite old.Philip [00:42:56]: Yeah, Medusa's old.Ali [00:42:58]: It was old.Vibhu [00:42:58]: But is it in the book as a good, here'sPhilip [00:43:01]: BaselineVibhu [00:43:01]: Baseline vanilla understand it?Philip [00:43:02]: Like you should know this.Vibhu [00:43:03]: Like I read the paper, I'm like, “ it makes so much sense.”Philip [00:43:05]: Yeah.Philip [00:43:05]: So with the book, I had a couple goals. One was to give people just a working vocabulary for the space as a whole, and the other was to give them some intuition about how each of these techniques works. As I mentioned in my AI Engineer talk, which is the first public addendum to this, the speculation space has moved much faster than everything else. So yeah, even at the time that I wrote the book Medusa, I very much included as a way for people to understand how the space evolved rather than what the most modern technique is. And now of course, there's DFlash, dSpark. There's, there's newer techniques even than EAGLE, although EAGLE is still very commonly used.Ali [00:43:51]: SpecSpecta.Philip [00:43:52]: Yes. Speculative decoding.Vibhu [00:43:54]: What canAli [00:43:56]: Oh, it's a paper by Tri Dao and it's like, it's doing speculative decodingVibhu [00:44:00]: HuhAli [00:44:01]: For the speculative decoder.Philip [00:44:02]: Oh, in spec- oh my God.Ali [00:44:02]: It's literally just an another. It's like, yeah, that's the most simple way to explain it, and it seems like he got trivial speed ups there. But it seems that the complexity with training, it's almost like in our mind at least, it's almost as complex as training GANs. Like it's like a very delicate balance and oftentimes you, it's just but yeah, it's literally speculative decoding on speculative decoding.Vibhu [00:44:21]: Speculative.Ali [00:44:22]: Yeah. We saw this paper.Vibhu [00:44:24]: It's interesting, right?Ali [00:44:24]: Yeah.Vibhu [00:44:24]: I wouldn't even expect it to be very particular to train, I wouldAli [00:44:29]: Right.Vibhu [00:44:29]: The naive part of me is like, okay, train speculative decoder.Ali [00:44:32]: But like, and it makes sense, like the whole idea of speculative decoding is you. It's like, it's like almost like the iPhone auto predict version but for a normal model, right? Like you're just, you're just, generating three tokens and you're like, okay, I'll do prefill on them. And so you save those three turns for your original model. Now your speculative decoder is doing three turns of auto regression, so why not just have an even smaller model?Ali [00:44:53]: The other question there is what are the size of speculators? So say forPhilip [00:44:58]: Right. It's like a billion parameters.Ali [00:45:01]: Like for MiniMax, it's. Yeah. It's like one layer. It's like one 60th of the original model usually.Philip [00:45:06]: Yeah. I think we should do a paper when we get back to the office.Philip [00:45:10]: SpeculativeAli [00:45:11]: SpeculativePhilip [00:45:11]: Decoding.Ali [00:45:13]: No, it's, it does seem like how, when do you stop? But then it also seems like if you're able to train spec-spec decode for instance, right? Like if you're able to have a small model that is accurately predicts what the intermediate speculator is gonna predict, that is able to predict what the original target model's gonna predict, then why not just use that smallest model directly, right?Vibhu [00:45:34]: Yeah. This isAli [00:45:35]: Like it seems likeVibhu [00:45:35]: Adjacent to the routing problem.Ali [00:45:36]: Right.Vibhu [00:45:36]: Yeah.Ali [00:45:36]: Right.Philip [00:45:37]: The thing with speculators is one of the practical constraints on using them is that you do have to run a small model on the same hardware that you're running the big model on. There is a orchestration and resource competition problem inherent in that, and that is one of the constraints on speculation in general, is that draft tokens cost resources to create and cost software complexity to manage. And so if you have like infinitely recursive speculators, you add in quite a bit of that complexity on the actual implementation within the inference engine as well, not just in the training process.Vibhu [00:46:17]: I was gonna say, I would wonder if you could do similar, like distillation and pruning of, it's the same thing, it's just a model. Can we not just distill a lot of the weights, quantize the speculator, out of my domain? The question that also comes up is, this is all for big server workloads, right? How much of this applies to, say I have this MacBook, I wanna run Gemma really efficiently. Similar problems, not the same?Local AI vs. Data Center InferencePhilip [00:46:45]: Pretty different. I talked to Selo, about this on his podcast a couple weeks ago. The difference between inference engineering for the data center and for production workloads versus inference engineering for local AI, is that we start with fundamentally like different constraints and different goals. With local AI, it's how do I fit this model onto my hardware and then make it less dumb? And with data center influence, it's how do I load this model and then make it less slow? And we care about less dumb, and they care about less slow. But the local AI inference engineering ecosystem, I think has a lot for us to learn from in the data center space. They are experts in various forms of quantization, including dynamic quantization that we just don't touch, in the pruning, in the distillation, in the, layer removal. There'Ali [00:47:42]: Layer removal matters less.Philip [00:47:43]: Yeah. There'Ali [00:47:44]: No one loves pruning really.Philip [00:47:45]: Yeah. Well, but the, but they doVibhu [00:47:46]: Which is surprising, right? But that's, that's a whole different thingPhilip [00:47:48]: Just to fit something on the laptop.Ali [00:47:50]: Right.Philip [00:47:50]: So yeah, it's a, it's an interesting, it's an interesting space. Not necessarily that like their techniques make sense for us to do in the data center, because we have different resources and different goals, but more that the process as well as the openness of that field is something to, admire.Ali [00:48:12]: Yeah. Like to your point, like, certain optimizations that would. Like for instance, Turbo Quantum Sharper, like it made such huge hype on that and we did like a whole deep dive on Twitter and like said, what is it? How does it work? Why is it good or not? And it took off and it was implemented on local devices because your memory bandwidth is so slow on like a MacBook, for instance. But try putting the same thing on like an NVIDIA GPU on a B200 Turbo quant would not be. Like, it would not be used. Like, NVIDIA - Like, NVIDIA made it clear that this is not a good optimization, and we've seen it firsthand where the overhead of doing dequantization, quantization of, in the kernel itself with turbo quant kernel, each end is much slower than the time that you save from doing the bandwidth. ‘Cause on the B200s, you have like 3.5 terabytes per second. You don't need decrease the storage that much. You don't need to do, FP4 KV cache. You don't need to use a requant. There's, there's, there's better optimizations to be made. But on Edge devices, it's extremely important, it's extremely useful. So, seems to be, like, different optimizations there, but then they're all uniquely combined with like all you wanna quantize the model, you wanna do speculative decoding, like certain common prefixes with bothPhilip [00:49:18]: Principles.Ali [00:49:19]: Yeah, exactly. Exactly. Exactly.Philip [00:49:20]: They also do a lot of work on, model parallelism, especially over, heterogeneous topology, where you have, some sparks and they are wired together with, Ethernet, DGX sparks.Ali [00:49:35]: Yeah, this is the Exo Labs guys.Philip [00:49:36]: Yeah. You have, a nu

This episode of NDS Chronicles brings together listener-submitted paranormal testimonies involving mantis beings, demonic attacks, haunted houses, shadow people, spiritual warfare, and possible divine intervention. The Raven is joined by Matthew from Straight Bible as they examine each encounter through a Christian worldview while mixing serious Biblical discernment with trademark NDS humor.The first testimony follows Dallas Spitfire through a disturbing relationship that escalates from gaslighting and violent behavior to revelations about a haunted house, shadow people, a flying Batman figure, occult coloring imagery, and a Medusa-like interdimensional demon he first encountered as a child. The conversation also explores marriage, divorce, relationship red flags, polyamory, spiritual oppression, and the danger of ignoring patterns that point toward demonic influence.Next, a listener describes using ketamine, Molly, meditation, alchemical imagery, and undefined spirituality before experiencing a terrifying demonic manifestation. The hosts explain why Eastern-style meditation, astral projection, New Age practices, and praying to an undefined “god” can open dangerous spiritual doors.Another testimony recounts hearing his mother's voice call him out from beneath a car seconds before the jack collapsed. The event is compared with Samuel hearing God through the voice of Eli in 1 Samuel 3, raising questions about divine protection and how God may communicate.HGTV Holly then shares a chilling account of tinnitus, whispering voices, mysterious spiritual “downloads,” thyroid cancer, and the sudden end of the activity after surgery. The episode closes with a sleep paralysis encounter involving a tall mantis being, an astral body, Reiki, temporary supernatural healing, familiar spirits, and the Biblical interpretation of so-called alien entities.Which testimony stood out most to you? Do mantis beings, shadow people, and New Age entities represent different phenomena—or the same demonic deception wearing different masks? Share your thoughts in the comments.Listen on Spotify, Apple Podcasts, and everywhere podcasts are available.► Subscribe for more Biblical analysis, Christian worldview discussions, and conspiracy research through a Biblical lens.00:00 Intro00:53 NDS Chronicles Begins04:03 Patreon and Brohemian Grove Event Details08:52 Dallas Spitfire's Paranormal Testimony12:57 Marriage, Divorce and Dating Again16:18 Haunted House, Shadow People and Flying Objects19:10 The Psycho Girlfriend and Relationship Red Flags28:22 Gaslighting, Screaming and Escalating Abuse35:31 Astral Projection, Marijuana and Spiritual Access43:42 The Violent Final Confrontation55:13 The Medusa Demon Connection01:06:41 Ketamine, Meditation and a Demonic Attack01:16:17 Biblical Warning About New Age Spirituality01:20:23 God's Voice Prevents a Near-Death Experience01:24:24 Samuel, Eli and Hearing the Voice of God01:28:38 Cancer, Tinnitus and Demonic Whispers01:41:37 Thyroid Cancer Removed and the Voices Stop01:43:08 THC Sobriety and Spiritual Recovery01:47:25 Sleep Paralysis and the Mantis Being01:53:10 Astral Encounter With a Mantis Entity01:56:28 Reiki, Temporary Healing and Familiar Spirits01:58:33 Biblical Discernment of Mantis Beings02:02:32 Final Thoughts and OutroBecome a supporter of this podcast: https://www.spreaker.com/podcast/nephilim-death-squad--6389018/support.☠️ Nephilim Death Squad — New episodes 5x/week.Join our Patreon for early access, bonus shows & the private Telegram hive.Subscribe on YouTube & Rumble, follow @NephilimDSquad on X/Instagram, grab merch at toplobsta.com. Questions/bookings: chroniclesnds@gmail.com — Stay dangerous.

Gather up your packs and lets hit the Highland Road. It's the Irish & Celtic Music Podcast #768.  -  -  Subscribe now at CelticMusicPodcast.com! Mary - Grace Autumn Lee, Open The Door For Three, Almost Seamus, Juha Rossi, Avourneen, Rogue Diplomats, TJ Hull and Amanda Caretta - Hull, The Bilge Pumps, Gwendolyn Snowdon & Cathinca, Release the Craicen, Medusa's Wake, Bang on the Ear, Pressgang Mutiny, Maggie MacInnes GET CELTIC MUSIC NEWS IN YOUR INBOX The Celtic Music Magazine is a quick and easy way to plug yourself into more great Celtic culture. Enjoy seven weekly news items with what's happening with Celtic music and culture online. Subscribe now and get 34 Celtic MP3s for Free. VOTE IN THE CELTIC TOP 20 FOR 2026 This is our way of finding the best songs and artists each year. You can vote for as many songs and tunes that inspire you in each episode. Your vote helps me create this year's Best Celtic music episode. You have just three weeks to vote this year. Vote Now! THIS WEEK IN CELTIC MUSIC 0:02 - Intro: James Stovehouse 0:08 - Mary - Grace Autumn Lee "Helen Burns': Black Pat's/Over the Moor to Maggie's (Reels)" from Eyre 4:52 - WELCOME 6:21 - Open The Door For Three "Farewell, Lovely Mary/The Lizzes" from A Prosperous Gale 11:11 - Almost Seamus "Old First Ward" from Almost Seamus 14:28 - Juha Rossi "Miss MacDermott (Princess Royal)" from O'Carolan Tunes on Mandolin 17:07 - Avourneen "Stony Steps" from Way Stations 22:00 - FEEDBACK 25:18 - Rogue Diplomats "Mary Mac / Drunken Sailor" from Whiskey Picnic 28:30 - TJ Hull and Amanda Caretta - Hull "The Bunch of Keys" from Tippy Taps 32:31 - The Bilge Pumps "Old Salt" from Greatest Hits, Vols I - VII 36:57 - Gwendolyn Snowdon & Cathinca "Lyke - Wake Dirge" from Fabula Feminae  -  EP 41:49 - THANKS 43:38 - Release the Craicen "Ode to Our Lost Fiddler / St. Anne's Reel" from Live! Songs on a Boat 48:04 - Medusa's Wake "Irish Sky" from Rascals and Rogues 52:23 - Bang on the Ear "The Black Tar Fiddle Set" from Highland Road 56:22 - Pressgang Mutiny "TEU" from Departure 1:01:15 - CLOSING 1:02:07 - Maggie MacInnes "Toiseach na Tràghad" from Uiseagan Bharraigh 1:06:23 - CREDITS Support for this program comes from Dr. Annie Lorkowski of Centennial Animal Hospital in Corona, California. Support for this program comes from John Sharkey White, II. Support for this program comes from International speaker, Joseph Dumond, teaching the ancient roots of the Gaelic people. Learn more about their origins at Sightedmoon.com Support for this program comes from Cascadia Cross Border Law Group, Creating Transparent Borders for more than twenty five years, serving Alaska and the world. Find out more at   www.CascadiaLawAlaska.com Support for this program comes from Hank Woodward. The Irish & Celtic Music Podcast was produced by Marc Gunn, The Celtfather and our Patrons on Patreon. The show was edited by Mitchell Petersen with Graphics by Miranda Nelson Designs. Visit our website to follow the show. You'll find links to all of the artists played in this episode. Todd Wiley is the editor of the Celtic Music Magazine. Subscribe to get 34 Celtic MP3s for Free. Plus, you'll get 7 weekly news items about what's happening with Celtic music and culture online. Best of all, you will connect with your Celtic heritage. Please tell one friend about this podcast. Word of mouth is the absolute best way to support any creative endeavor. Finally, remember: plastic isn't just a waste problem, it's a climate problem. Most plastic is made from oil and gas, and producing it releases greenhouse gases at every step. This July, take part in Plastic Free July. Swap single use plastics for reusable bags, bottles, and containers. Solar and wind are now the cheapest power sources in history. Cutting plastic is another way to shrink demand for the fossil fuels behind it. Reduce your waste this month, one choice at a time. Promote Celtic culture through music at http://celticmusicpodcast.com/. WELCOME THE IRISH & CELTIC MUSIC PODCAST * Helping you celebrate Celtic culture through music. I am Marc Gunn. I'm a Celtic musician and also host of Pub Songs & Stories, where I share stories behind my songs and those of other Celtic and folk musicians. This podcast is for fans of all kinds of Celtic music. We are here to build a diverse Celtic community and help the incredible artists who so generously share their music with you. If you hear music you love, please email the artists to let them know you heard them on the Irish & Celtic Music Podcast. As with all things, it is important to support our community, whether that is local or in a small niche we love like "Celtic music". Skip the big corporate businesses and bands. Instead, spend your money where it will make a difference. You can do that with each of the artists on today's show OR your favorite artists. Buy a CD, Album Pin, Shirt, Digital Download, or join their community on Patreon. You can find a link to all of the artists in the shownotes, along with show times, when you visit our website at celticmusicpodcast.com. ALBUM PINS ARE CHANGING THE WAY WE HEAR CELTIC MUSIC Looking for a fresh way to support the music you love? Meet the Album Pin. Album Pins are lapel pins themed to a specific album — and each one comes with a digital download. Wear your music. All of my latest pins are wood - burned and locally produced, which means a smaller footprint and a one - of - a - kind feel you won't find anywhere else. Pick yours up at magerecords.com THANK YOU PATRONS OF THE PODCAST! Because of generous patrons like you, the Irish & Celtic Music Podcast releases new episodes nearly every single week, including this one. Your support doesn't just fund the show, it fuels a movement. It helps us share the magic of Celtic music with thousands of new listeners and grow a global community of music lovers Your contributions pay for everything behind the scenes. Audio engineering. Stunning graphics. Weekly issues of the Celtic Music Magazine. Show promotion. And most importantly, buying the music we feature from indie Celtic artists like the ones you heard today. And if you're not yet a patron? You're missing out! Patrons get: Early access to episodes Music - only editions Free MP3 downloads Exclusive stories and artist interviews A vote in the Celtic Top 20 All of this lives over at SongHenge.com. HERE IS YOUR THREE STEP PLAN TO SUPPORT THE PODCAST Go to our Patreon page. Decide how much you want to pledge every month, $4, $12, $30. Keep listening to the Irish & Celtic Music Podcast to celebrate Celtic culture through music. You can become a generous Patron of the Podcast on Patreon at SongHenge.com. TRAVEL WITH CELTIC INVASION VACATIONS Every year, I take a small group of Celtic music fans on the relaxing adventure of a lifetime. We don't see everything. Instead, we stay in one area. We get to know the region through its culture, history, and legends. You can join us with an auditory and visual adventure through podcasts and videos. Learn more about the Celtic Invasion of Hiking the Isle of Skye in 2027 at http://celticinvasion.com/ #celticmusic #irishmusic #celticmusicpodcast I WANT YOUR FEEDBACK What are you doing today while listening to the podcast? Send me a photo. If you're in a Celtic band, send me an audio recording of you performing live. Just audio. I'll use it in a podcast episode later this year. Email me at follow@bestcelticmusic. Isobel McMahon messaged on Patreon: "I appreciate this clear statement on AI." Jordan Reeder sent a photo: "Hi, just wanted to show what I was doing while enjoying the awesome music on the podcast… one of these days I'd love to submit some of my music. Maybe after I chop my way through this pile of firewood! Alex Meyers emailed from Toronto: "Politically neutral Irish music does not exist. Hi Marc, I started listening to the podcast seven years ago as I took my newborn son for long walks with wireless speakers perched on his stroller so he could listen too. Now with two boys we frequently listen while we have breakfast. In a recent episode you responded to yet another listener who told you to stick to music and leave the politics alone. I am always impressed by how resolute you are in the face of these cynics, you never try to soften the message of peace, compassion, and inclusion. Also, I've always found it hilariously ironic that someone would say "leave out the politics" about IRISH music, 1/3 of all Irish songs are about oppression and kicking out the British. Have these people never listened to The Wolfe Tones or The Cranberries? Speaking of which, I highly recommend you check out the Irish rap group Kneecap. They perform in Irish/Gaeilge and English and deserve massive credit for popularizing Gaeilge. They are firmly on the side of the oppressed from Ireland to Palestine and their latest album is "Fenian", which should tell you something about their politics. Keep up the great work!"  

Cam and Medusa take to stars, and then instantly land on the Jurrasic Park / Shrek crossover no one asked for. Are there any vixen warriors who can save them from this terrible fate?? Our Flipped Script: What if Krystal wasn't captured? Intermission: 1:01:00 - 1:14:00 Suggested Topics: Batman=Acid, Dangerously Mid, A Rougher One to Bring Us Back to Form, Everyone Knows Who to Ship, Cameron on the Scene, Strap Lasers on Dinosaurs, Justice for Blaster  

Trouble in Zur Algud Ch. 1With the battle still underway, the group uses all they have in their various kits. The cleric begins to look hurt, but the fight is far from over. In retaliation towards Thorak, the cleric casts Harm on one half of the Slash Bros. Valen, trying to test his luck with his magical flask, drinks an Elven Ale called PolyPilsner, which polymorphs the monk into a canary. Chan summons a flame strike on the flesh effigy, and it does even more damage than expected; as a result, her flames that she's done so well to control and suppress thus far are now beginning to go wild. A giant sunbeam from the Mad Cleric takes out a good number of the team, rendering them unconscious. The team's own cleric, Thorak, helps Chantico back up with a Cure Wounds spell. As she comes to, she uses burning hands on the cleric of Hastur, finally putting an end to him. They don't have time to celebrate; they pick up their fallen allies and begin to make a fast exit. The blob makes an unsettling sound that brings back some "familiar" faces. The andrite in the mines begins to explode as the group attempts to flee and fight as they head out of the mines. Chan calls to the Daughters of Creation and Enya to aid in the safe exit of her friends and companions. Calling three scorching rays towards the monsters closing in on the group, which coalesces into a sphere surrounding the group as Enya attempts to teleport them out. It is successful, and in a flash they are outside the mines, and as they look behind them, they feel the immense explosion that erupts from the mines. After the smoke and dust settles, they realize that Chan is no longer with them.The party is embraced, arriving in Tradeshore as heroes, as Hanith Blackburn sent word of their heroic deeds. They are offered a reward to Zur Algud to celebrate the upcoming Festival of the Children, accommodations at the Ebon Wing Lodge, a 10g-a-day stipend, and access to the family's teleportation circles to travel to and from. We finish with the varying accounts of what the party did in the three weeks leading up to their travel for the festival. From one adventure to the next, they sharpen their skills, broaden their horizons, and prepare for a time of reprieve and revelry if it were only that simple.There's so much happening, and that's where we pick up…Find out what happens next in this episode of the Medusa's Cascade: Blood & BoozeTheme Music is written and performed by EfflorescenceTitle Card by Pierce Graphics Mixed by Thomas Lapierre IIICheck out the show at themedusascascade.com Hosted on Acast. See acast.com/privacy for more information.

For centuries Europe has preserved stories of dragons, vampires, werewolves, Gorgons, wild men, and terrifying creatures that refuse to disappear. But what if these legends all point back to the same biblical event?In this episode JT sits down once again with  @uconspiracy  the author of The Nephilim Look Like Clowns, to explore the strange connections between Genesis 6, the Book of Enoch, Greek mythology, Dionysus, Medusa, Harlequin, European folklore, and even the disturbing origins of modern clowns.Topics include: • Genesis 6 and hybrid beings • Medusa and the Gorgons • Echidna—the Mother of Monsters • Dragons throughout Europe • Wild Men traditions • Krampus and Harlequin • Dionysus and ancient mystery cults • Vampire origins • Why clowns became symbols of fear • The psychology of Coulrophobia • Biblical parallels hidden inside European folklore Do these legends preserve distorted memories of real events? Watch and decide for yourself.Become a supporter of this podcast: https://www.spreaker.com/podcast/jt-s-mix-tape--6579902/support.Please support our sponsor Modern Roots Life: https://modernrootslife.com/?bg_ref=rVWsBoOfcFPatreon: https://patreon.com/JT_Follows_JC?utm_medium=unknown&utm_source=join_link&utm_campaign=creatorshare_creator&utm_content=copyLinkJESUS SAID THERE WOULD BE HATERS: https://jtfollowsjc.com/product-category/mens-shirts/JT's Hats: https://jtfollowsjc.com/product-category/hats/Coaching Program: https://www.echoesoftruthnetwork.com/joinTelegram Group: https://t.me/jtsmixtapeNew Trucker Hat: https://www.shopmmca.com/shop/p/jt-follows-jc-hat

Contains minor spoilers for The Odyssey.Oscar and Carl chat about the 2010 remake of Clash of the Titans, starring Sam Worthington and Liam Neeson. Together, they discuss the terrible lighting, the acting, and the film's many odd creative choices. They also ask whether this is the worst Ralph Fiennes has ever been, along with the ultimate question: Who would watch this?If you have any questions or requests, send them to askwwwtpodcast@gmail.comFind us through:YouTube: https://www.youtube.com/@whowouldwatchthisInstagram: https://www.instagram.com/whowouldwatchthis/TikTok: @podcastwhowouldwatchthisLetterboxd:Carl: https://letterboxd.com/carlllllllllll1/Oscar: https://letterboxd.com/oscarfart/More links: https://linktr.ee/whowouldwatchthis  

The Raft of the Medusa (1818-1819) by Theodore Gericault is a romantic painting that depicts the aftermath of a shipwreck that happened around that time. The painting highlights a group of survivors stranded on the raft who are struggling between hope and despair. In this episode, we will be exploring all that is within this painting. The good, the bad, the ugly, and the beauty. By exploring different themes and how controversial this painting was at the time, we will be dissecting the meaning of the painting and how it connects to the current movement of the time, romanticism.

Send us Fan MailWhat you'll hear:some ramblings of mine about seasonal uniforms 1:30game changing boxers for the summer dress season by Woxer (not an affiliate link, I just really like them) 3:15The meat of the episode starts at 4:00Check here for the waitlist for the DIY version, The Audacity Project LITE 12:10This ain't my first hard pivot rodeo- I've done this before and yes it is terrifying 13:00So what the hell am I going to do now?  18:00Creatrix:  the retreat waitlist can be found here!  18:507 Deadly Satisfactions waitlist is here, it is ONLY going out to the waitlist 20:00The Order of Medusa!!!  She comin.  October 2026.  You can join the waitlist here 22:00Don't go back to sleep.xoRachelSign up here for monthly blasts and functional wooFind me on InstagramSupport this podcast on Patreon

This week, we're joined by Davide Mutendji of Elevation Rhythm, whose new album WASHED just dropped. He tells us about the heart behind the project and how Drake became a surprising inspiration for the worship group.In RELEVANT Buzz, the trailer for ‘I Play Rocky' arrived, and it sent us down a rabbit hole into Sylvester Stallone's faith journey — including a conversation where he calls out Christians who think they can skip church. (Rocky would never!) We also dig into a new study finding only one in three young adults are dating even though nearly all of them want to get married (the reasons are bleak). And with the new season of ‘Quarterback' out on Netflix, we look at Baker Mayfield's road from college wild child to a real-deal faith story.In Slices, a school district in New York is putting a humanoid robot named in classrooms, and Jesse loves it. And Derek shows us the clip of the grandpa getting flipped by a bison this week. Plus we settle once and for all whether Medusa is in ‘The Odyssey.'Then in feedback we have your rent-free reality TV moments, and we ask this week's Question of the Week: what's the worst Christian breakup line of all time? Drop your answer in the comments — we're reading our favorites next week.Highlights:00:00 — Intro / Emily reviews The Odyssey 07:41 — RELEVANT Buzz: ‘I Play Rocky' & Stallone's church call-out 26:14 — Why two-thirds of young adults aren't dating 33:32 — ‘Quarterback' & Baker Mayfield's faith journey 41:25 — Davide Mutendji of Elevation Rhythm 50:04 — Slices: The robot teacher coming to New York classrooms 53:08 — Grandpa vs. bison 58:08 — Your feedback: Reality TV moments living rent-freeAdvertising Inquiries: https://redcircle.com/brandsPrivacy & Opt-Out: https://redcircle.com/privacy

With her debut poetry collection, OFFERINGS FOR ORDINARY GODS (Clash Books), Ali Trotta brings myth, magick, tarot, and ritual to bear on the eternal & primal struggle of women. We talk about those roots and how tarot and ancient myth inspired her work, where tarot began for her — '90s, The Craft — and how past guest Katelan Foisy resparked her interest and showed her new depths in the cards, and her focus on revising ancient myths to retell the stories of wronged women like Persephone, Medusa, and Dido. We get into how she feels at seeing her work collected and how to look at her development as a poet without cringing at her younger selves, how everything is fair game and poetry can serve as exorcism, and how confession can be deeply coded in a poem. We also discuss the teacher who gave her the tools to read poetry deeply, her Very Online-ness and how she's built relationships and writer communities online (and how it's good to be part of the Group-Chat), her online #tinypoem flurries and how rockstar she felt when someone got a tattoo of one on their arm, the project(s) she'd love to write, and more. PLUS: She reads a few of her poems and gives me my second tarot reading of the summer! Follow Ali on Bluesky, Instagram and Tumblr, and subscribe to her newsletter • More info at our site • Support The Virtual Memories Show via Stripe, Patreon, or Paypal, and subscribe to our e-newsletter

If the last few weeks have felt like an “underworld” journey, you aren’t alone. As the rare “Basket” configuration peaks between July 20–26, the pieces are finally falling into place. This isn’t just another week of astrology—it is the start of a phoenix rebirth that is burning away your old life to make room for the new. In this transmission, Sabrina Lynn reveals the hidden narrative behind the chaos. From Mercury stationing direct to the fierce start of Leo Season, discover why these are the most magical two weeks of the year—and how to claim your soul power at this profound crossroads. The Soul's Transmission – The Burning Basket: The truth about why your old life had to “torch” itself this month. – The Final Pivot: How Mercury Direct is returning from the underworld with the answers you’ve been seeking. – A Lionhearted Shift: Moving from the depths of Cancer into the outward power of Leo Season. – Dark Goddess Truth: Why Kali, Medusa, and Hecate are standing with you at this crossroads. – The Grace of Jupiter: How to stop “healing” your life and start living it. ReWilding Tools for the Rebirth July Bones Workshop: Anchor Your Transformation Don’t get lost in the fire. This workshop is your grounding cord, helping you anchor these high-frequency shifts directly into your system. Instant Access to the Replay. Join Bones  The Magicians Bundle: Buy 2 Get 1 Free! Save $197 Command the Great Pivot with Jupiter, Uranus, and Merlin as your guides, this bundle provides the energetic toolkit to hold the “Big Magic” of this week. As the crossroads appear, use these practices to shift from “healing” your life to actively co-creating it with the universe. Claim the Magic Open until July 31 Jupiter Archetypal Journey: From Ashes to Abundance Jupiter's opposition to Pluto demands a choice: power struggles or soul expansion. This journey is designed to help you attune to the benevolence, joy, and lionhearted expression of the upcoming Leo Season. Explore the Jupiter Journey Open until July 31 ReWilding Weekly: Stay connected with our community and get practices, journal prompts, and deeper insights in your inbox. Subscribe here  _____ Listen to “The Most Magical Two Weeks of the Year Are Here (Don’t Miss It)“ podcast here… Topics Explored in “The Most Magical Two Weeks of the Year Are Here (Don’t Miss It)” podcast: (Times based off audio version) Timeline of the Rebirth (00:00) – We finally know what July was doing (01:10) – Sabrina's personal “underworld” journey (02:20) – What your soul was asking during the retrograde (03:46) – The Pivot: Putting insights into action (06:01) – Pluto & The Phoenix Rebirth: The ash and the fire (06:55) – Dark Goddesses: The demand for soul truth (10:30) – Jupiter in Leo: The most magical two weeks (12:10) – Monday: Power Dynamics & Jupiter/Pluto (14:48) – Wednesday: Leo Season begins (16:36) – Thursday: Mercury Direct & The Heart Opening You can leave a comment or question for Sabrina on the YouTube version of this episode. Listen to after “The Most Magical Two Weeks of the Year Are Here (Don’t Miss It)“: July 13–19 Astrology | One of the Rarest Weeks of 2026 DARK GODDESS SWORD OF TRUTH: Love & Liberation (Kali, Medusa, Lilith, Hecate, Persephone, Eris) HEKATE MEDITATION: Shadow Work for Reclaiming Witch + Medicine Woman STAY CONNECTED ReWilding Weekly (free, embodied astrology)  IG  Website  Disclaimer: Educational/spiritual perspectives; not medical/mental-health advice. #2025Shift #NewHuman #SpiritualAwakening Welcome to ReWilding with Sabrina Lynn & ReWilding for Women! A gifted facilitator of revolutionary inner work and the world's leading archetypal embodiment expert, Sabrina Lynn is the creator of the groundbreaking ReWilding Way and founder of ReWilding For Women. Sabrina has led more than 100,000 people through programs based on the ReWilding Way, a modality of healing and awakening that strips away the false, the deep wounds from early life, and the fears that hold people back, to reveal their true and unique soul light and help them build their innate capacity to shine it in the world. Her work includes in-person retreats and events, the monthly ReWilding Membership, Living Close to the Bone, Priest/ess Trainings, Mystery Schools, the ReWilding with the Archetypes, and the wildly popular 6 Faces of the Feminine workshop series. Welcome to ReWilding! The post 392 – The Most Magical Two Weeks of the Year Are Here (Don’t Miss It) appeared first on Rewilding for Women.

In this episode, we dive into Gericault's deep obsession and depiction of the truth behind the raft of Medusa. We explore the political stances of the time and how this piece fell in history. As well as Gericault's dedication to portraying the emotion of raw humanity in the artistic composition of the painting. Join us as we figure out if this is a masterpiece of truth or a romanticized depiction of horror.

this episode represents the emotional tention in the painting. We go over each detail and how it might be both a representation of truth and romatization of horror.

The Blooming Court Ch. 12The party glances upon the true Song Bloom and chooses one that they think will please the Queen. It, however, takes some time and teamwork to figure out how to properly remove it and keep it intact and healthy for travel. Zechs manages to uproot the Song Bloom and, with the aid of Galahad, makes a makeshift pot for it to travel with. Before heading to the Queen straightaway, the group, using Arguile's Rope Trick spell, devises a plan for what they should do if things inevitably go sideways. Taking all possible scenarios into account, they agree to try to avoid fighting at all costs, but if it comes down to it, so be it. The party presents the Song Bloom to the Queen's approval, but after getting what she wanted, she promptly leaves, and a fight between her guards and the party ensues. The twist? A familiar halberd wielder joins the fray as The Executioner makes her attack as well. The fight is intense as each side pulls out its big guns, trying to overwhelm its opposition. Ultimately, through a lot of teamwork, grit, and heavy swings of their weapons, the party manages to shift the tides to interrogate The Executioner when the dust settles. There's so much happening, and that's where we pick up…Find out what happens next in this episode of the Medusa's Cascade: Collateral Damage!Theme Music is written and performed by EfflorescenceMixed by Thomas Lapierre IIITitle Card by Pierce Graphics Check out the show at themedusascascade.com Hosted on Acast. See acast.com/privacy for more information.

For millennia, the legend of Medusa has haunted the human imagination, a creature so terrifying that a single glance could turn the living to stone. While most dismiss her as pure mythology, some researchers point to serpentine monsters appearing independently across ancient cultures and ask: were they all describing the same thing? But look closer at Medusa's story, and something shifts. Because some say she wasn't born a monster; she was made into one.  For a full list of sources, please visit: sosupernaturalpodcast.com/mystical-medusa Did you know you can listen to So Supernatural ad-free? Join the Crime Junkie Fan Club! Visit https://crimejunkiepodcast.com/fanclub/ to view the current membership options and policies. So Supernatural is an Audiochuck and Crime House production. Find us on social! Instagram: @sosupernaturalpod Twitter: @_sosupernatural Facebook: /sosupernaturalpod Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Jace and Rocky break down all 13 DC releases for the week of July 15, 2026. Nightwing #140 brings Psycho Pirate into Dan Watters' Blüdhaven mystery by tying the city's violence to Medusa architecture instead of the Medusa Mask. Wonder Woman #35 works as a Wonder War interlude that retells Diana's origin through Steve Trevor explaining Wonder Woman's history to Trinity. End of Life #6 has Eddie Stallion try to leave Pluto before his reunion with Sophie convinces him to stay while the Menagerie's assassins keep closing in. Batwoman #5 has Kate Kane rescue Alice by removing her mask and reaching her sister directly, while Mother Despina warns that winning may still mean losing. Deathstroke: The Terminator #5 reveals the person behind the attacks on Slade and ends with a threat to Rose Wilson that she may not survive. Absolute Flash #17 ends the Gorilla City story in tragic fashion with the young Grodd caught between loyalty to his father and his bond with Wally. Batman/Superman: World's Finest #53 highlights Dick Grayson's skills as he tries to break Batman and Superman free from Tyrannus Rex's control in Skartaris. Dark Knights of Steel 2 #1 expands the medieval DC world beyond the original kingdoms as Kara is welcomed into Amazonia, Batman remains separated from the League, and Amanda Waller's secret villainy continues to hang over the larger story. New Titans #37 hints at Static joining the team while giving Kid Devil more spotlight and continuing Spoiler's rise into a stronger leadership role. Lobo #5 has the network ordered to bring Lobo back, only for the executive to put a contract on him and send Lip Bomb after the Main Man. Catwoman #89 pushes Selina deeper into her Gotham war with Black Mask while bringing her connection to Carmine Falcone closer to outright confirmation. Superman Unlimited #15 builds on the Ultra Boy reveal by showing Jon Kent's faith in him pay off, while Superwoman's arrival keeps the Earth-3 story from ending cleanly. Absolute Green Arrow #3 shifts further into Black Canary's investigation as the issue reveals Miraclo, Rex Tyler and anti-aging experiments tied to the billionaires being targeted. They also give a rundown of this week's collected editions, reprints and facsimile releases. As always, all books are ranked from top to bottom and each host gives a Book of the Week pick.

We’ve got five literature classics to introduce to your family through board games. 0:00:00 Fact for 426 The number of people living past the age of 80 worldwide is projected to triple, hitting roughly 426 million by the year 2050. Sponsor Message Go to FirstMoveFinancial.com/familygamers and set up a free 15 minute call to see if they can help! 0:04:30 What We’ve Been Playing Obelus (our review)Spirits of the Wild: AwakeningFire TowerPalette *Tag Team (*) first time mentioned on The Family Gamers Podcast June Monthly Report Andrew: 22 plays of 8 unique games. H-index 3 (LotR Trick Taking, Trio, Got Five!) Anitra: 27 plays of 12 unique games. H-index 3 (same as Andrew) 0:16:45 The Family Gamers Community Welcome new members! #Backtalk We asked you about how you treat board games. Read the answers on Facebook or #backtalk channel of the Discord. 0:22:40 Classic Literature to Introduce via Board Games We’re going to focus on games that will whet the appetite: You don’t need to already know the stories to enjoy them, but there’s enough to make kids WANT to dig deeper. 1. Winnie the Pooh stories by A.A. Milne: Hundred Acre Wood Fluxx 2. Classic Grimm’s fairy tales: The Grimm Masquerade and 3 Chapters – both of these games make characters from different tales interact, but with specialized relationships unique to each one. 3. The Adventures of Sherlock Holmes by Sir Arthur Conan Doyle: tons of games that riff on this character. But the best to introduce the setting and themes from Doyle’s stories include: Unmatched: Cobble & Fog Baker Street Irregulars (4-book / 4-player mystery from Van Ryder publishing) Sherlock 13 (deduction of a criminal, introduces many characters) 4. Shakespeare: be careful in how you approach this! Shakespeare is wonderful, but not great if you’re forced to read it. We recommend Unmatched: Slings & Arrows to introduce some of the fun of Shakespeare. 5. Greek mythology: Horrified: Greek Monsters Santorini Iliad (Unmatched has Medusa in BoL1, Achilles & Patroclus in BoL2) One more shout out for the Unmatched series. They pull from tons of classics: Alice in Wonderland, Beowulf, Robin Hood, Sinbad, and even modern “classics” like Jurassic Park, Marvel characters, The Witcher, Buffy, and TMNT. (Other games based on classic literature that are great, but work much better if you already love the story: Ichor, Bottle Imp, Paint the Roses, Princess and the Goblin, The Grimm Forest) 0:46:10 New Backtalk Question What’s your hidden gem? What’s a lesser-known game that you just click with? Or a game that you think should be much better known because it’s awesome? Tell us on the #backtalk channel on our Discord, or in our Facebook community. Find Us Online: Facebook: @familygamersaa and thefamilygamers.com/communityTwitter (X): @familygamersaaInstagram: @familygamersaaTikTok: @familygamersaaBluesky: @familygamersaaThreads: @familygamersaaYoutube: TheFamilyGamers or join the Family Tabletop Community on Discord! thefamilygamers.com/discord Or, for the most direct method, email us! andrew@thefamilygamers.com and anitra@thefamilygamers.com. PLEASE don’t forget to subscribe to the show, tell your friends about the show, and leave us a review at Apple Podcast or whatever your podcast subscription source is. We’re also on Amazon Music, TuneIn, and Spotify. You can also now find us on YouTube Music! So pull it up and give us a listen while you’re toiling away at work :) If you’d like to support The Family Gamers, there’s three ways to do that. First, you can pick up some sweet sweet merch: T-shirts, hoodies, mugs, and stickers! Go to store.thefamilygamers.com You can also support us on Patreon, if that’s more your speed. We have a few different support tiers from just a couple of dollars to show your appreciation – which we appreciate – up to five, ten, or even $20/month. There are all sorts of different rewards you can get, which we tried to make unique. You can see those at thefamilygamers.com/patreon. Thirdly, you can start your Amazon shopping at TheFamilyGamers website, using our Amazon affiliate link. Music for The Family Gamers Podcast is provided with permission from You Bred Raptors? The Family Gamers is sponsored by First Move Financial. Go to FirstMoveFinancial.com/familygamers to learn how the team at First Move Financial can help you pile up the victory points. The post Episode 426 – 5 Classic Books to Introduce via Games appeared first on The Family Gamers.

The Blooming Court Ch. 11We pick back up with Glad, who comes face to face with the Daughter of the Vine. It is here that she is told the circumstances of her birth, or rather, how she came to be. It's not an easy tale to tell, nor is it easy to receive. The Daughter speaks of Coralin and his making of what would become her siblings, or the ones who came before her. Sharing in great detail their purpose and their intention. Finally ending with Glad's story.Glad, very torn between distraught and trying to wrap her head around what she has just heard, continues to listen and take in what the Daughter tells her. Feeling uncertain about what she's to do and watching the world crumble beneath her feet, Glad tells the Daughter that, while she's been on this journey, she's slowly discovered who she was and is now conflicted by the truth that clashes with her identity. The Daughter tells her that she is who she chooses to be and not where she came from. She has made conscious decisions to be who she is, and she has not done so out of obedience or command. Glad feels reassurance that she hasn't felt in quite some time, as well as recognition that she was doing well. The Daughter informs her that there is no trial for her to endure for her blessing, as this entire journey has been her trial. With that, they embrace in a much-needed hug. On the outside, the party, still waiting for Glad, is discussing their next steps. Some feel as though they shouldn't incur the scorn of the Fey by disregarding the Queen's ask for the Song Bloom, and on the other side, some wish to heed the warning of Shankise and be instruments in starting a war. This all happens as Glad is speaking with the Daughter and continues upon her exit. Eventually, the party leaves it to speak with the Treants, hoping to reach a peaceful third option. Exiting the lake, the group heads back the way they came, and upon entering Treant territory, they decide the bards should speak with them to reach a peaceful resolution to their predicament. The Treants receive the party and invite them to speak with the elders about what was asked of them and by whom. The party is then informed of the ramifications of what would occur if they got the Song Bloom and began to question the nuances of THE Song Bloom and A Song Bloom, hoping that by not taking THE Song Bloom, they can avert the Queen's intention of starting a war while not incurring her wrath for finding a loophole.There's so much happening, and that's where we pick up…Find out what happens next in this episode of the Medusa's Cascade: Collateral Damage! Theme Music is written and performed by EfflorescenceMixed by Thomas Lapierre IIITitle Card by Pierce Graphics Check out the show at themedusascascade.com Hosted on Acast. See acast.com/privacy for more information.

In this episode Jeff and Tracy sit down with Brandes Stoddard to talk about the 2024 5e D&D Monster Manual - this is part 23 of a multi-part series. Thanks to our Sponsor: The Tome Show's Awesome Patreon Patrons!!!!!!!!!!!! Links: Jared on the Web Jared on Gnome Stew THAC0 With Advantage Podcast Brandes on the Web Brandes on Bluesky Brandes on Mastodon Brandes at Tribality Patreon.com/BrandesStoddard Jeff on Twitter Tracy on Twitter Tracy on the Web Thetomeshow.com Patreon.com/thetomeshow

A new Canadian play is reimagining the ancient Greek myth of Medusa and giving the gorgon a voice. Oyin Oladejo is playing the snake-haired lead in the Toronto production of Medusa. Oyin joins Tom in the studio to talk about the play, and why this show stares deeply into female fury.

Amid a desperate retreat through Ors Themar, the party is scattered and under relentless pressure. A draegloth pins down Arguile, Shanks teeters on the brink of death, Galahad draws constant enemy fire, and Yurik struggles to regroup with the others. Through quick thinking, Zechs, Salix, Duo, and the rest manage to extract Arguile, rescue Shanks, free Galahad, Turk, and Glad from a magical binding, and reunite the party. They are then led to safety by Enich, Yurik's son and Arguile's half-brother, who provides refuge, shares what he knows of Ors Themar, and helps them recover before guiding them toward an escape route. After narrowly escaping the city, the party reflects on what they've witnessed during camp while Arguile's mysterious stone rapidly counts down. Shanks discovers a copper coin and a raven's feather, and growing signs point to Arioch's manipulation. Racing back into the city after the stone reaches zero, they find Dorna missing, identify Yurik's sword as Tempest's Edge, and rush to answer Shepard Miles' summons.At the cemetery, Arioch reveals the devastating truth: Yurik is Arguile's biological father, having sacrificed his own heart to save his son at the cost of everyone's memories. Later reunited with Dorna through Arioch's cruel amusement, the two unknowingly had Enich together. Arioch forces father and son to duel for the right to become his general, threatening to endlessly kill and resurrect Dorna until they comply while infecting Enich with the same deadly illness that once plagued Arguile. Though the party aids from the sidelines, they are powerless to stop him. Yurik ultimately chooses to go willingly with Arioch after an emotional farewell, leaving the party to battle Shepard Miles' forces. The ensuing fight quickly becomes a brutal uphill struggle as a necromancer raises undead, a monk repeatedly stuns Galahad and Turk, and the enemy relentlessly presses its advantage. Galahad receives Bahamut's blessing after falling in battle, rising with divine power. At the same time, Arguile slays the cleric, Shanks, Arguile, and Zechs eliminate the monk, and the remaining undead are slowly cleared. The battle reaches its darkest moment when the necromancer casts Power Word Kill, instantly killing Shanks before later dropping Turk with Cone of Cold, while Zechs also falls unconscious.As each fallen hero enters the Veil, they receive life-changing encounters: Shanks speaks with The Hunter, fully awakening Mind's Eye, and later receives the Matron of Ravens' blessing to oppose the Constable; Turk confronts Kas the Bloody-Handed, who strips him of his warlock powers after he refuses to serve him further; and Zechs finds peace with Tumra, the Snow Leopard, awakening with her white handprint upon him. Salix ultimately revives Shanks, the party defeats the remaining enemies, and the family reunites before returning to the city. Though shaken, they recover what spoils they can, spend an evening mourning and celebrating with matching tattoos, share their divine experiences, and help Galahad acquire a Belt of Giant Strength. With new equipment, plans to retrieve Glad's boon, and a long-overdue trip to the Ren Faire ahead, the party finally earns a brief respite after one of the most harrowing chapters of their journey.There's so much happening, and we pick up with arc 30, “The Blooming Court”. How will the rest of this story unfold? Find out next time on the Medusa's Cascade: Collateral Damage!Theme Music is written and performed by EfflorescenceMixed by Thomas Lapierre IIITitle Card by Pierce Graphics Check out the show at themedusascascade.com Hosted on Acast. See acast.com/privacy for more information.

Episode 524 - From Darkness to Light: Mixi from Stitched Up Heart Explains How Music Saves Us From Depression and Over‑StimulationWelcome back to When Words Fail, Music Speaks, the podcast where we explore how melody, rhythm, and raw lyrics become a lifeline for anyone battling depression, anxiety, or just the everyday grind. I'm your host, James Cox—a self‑confessed music‑lover who's learned that a good riff can be as therapeutic as any therapy session.In today's episode we reunite with one of rock's fiercest voices, Mixi — front woman of Stitched Up Heart. It's been five years since our last conversation, and Mixie returns with fresh stories, new music, and candid reflections on the roller‑coaster life of touring in a post‑pandemic world. From the panic‑inducing tech glitches on the road to the quiet solace she finds in her foster kittens and the mythic resurrection of Medusa in her latest single, we dig deep into the creative and emotional battles that shape her art.We'll also hear Mixie break down the painful honesty behind “Sick‑Sick‑Sick,” discuss her long‑standing sobriety journey, and share practical tips for taming the over‑stimulating digital age—like forcing TikTok's algorithm to feed you positivity and simply “touching grass” when the brain gets overloaded.If you've ever felt the sting of a bad day, the overload of endless scrolling, or the exhilaration of a night on stage, you'll find a friend in Mixie's story. Grab your headphones, a cup of coffee (or Monster energy, if that's your vibe), and get ready for a heartfelt, raw conversation that proves once again: when words fall short, music speaks.Let's dive in.

"Don't look at me. I'll turn into stone!" Medusa sat there or somethin'" In this episode we follow the fortunes of the hapless but volatile Geoff, who leaves the safety(?) of Royston Vasey to seek his fortune in London as a stand-up comedian...

Tony Jay is here to help Cam and Medusa crack open some vampire skulls and win the day, and the babes. Our Flipped Script: What if the script was written like a classic action movie? Intermission: 1:11: - 1:23: Suggested Topics: Tony Jay, Now We Have a Story, Set Aside Homework and Play This Game, Shakespearean Eloquence, Wang Out for Sure, Smash Has Been Poisoned by the Monarchy, Lucas Isn't as Butch as Ness, Sometimes the Things Our Mom's Like Are Bad

Im Jahr 1165 taucht ein eigenartiger Brief auf, gerichtet an den byzantinischen Kaiser. In diesem Brief stellt sich ein mächtiger christlicher Herrscher vor: Priesterkönig Johannes. Als Herrscher über die drei Indien, der über 72 Könige regiert, lädt er Manuel I. zu sich ein. Der Brief ist zwar eine Fälschung, entfaltete aber eine ungeheure Wirkung, die mehrere Jahrhunderte überdauerte. Wir sprechen in der Folge über die Legende vom Priesterkönig Johannes, wie er den Ausgang eines Kreuzzugs maßgeblich mitbestimmte und über eines der ältesten christlichen Reiche der Geschichte. // Erwähnte Folgen - GAG539: Der Mongolensturm und das Jahrzehnt der Witwen – https://gadg.fm/539 - GAG560: HB07 – Mit dem Finger auf der Landkarte und Wie das Floß der Medusa entstand – https://gadg.fm/560 - GAG291: Ein erfolgreicher Kampf gegen die Kolonialisierung – https://gadg.fm/291 - GAG319: Ashoka der Große – https://gadg.fm/319 // Literatur - Keagan Brewer: Prester John: the legend and its sources, 2015. - Björn Opfer-Kling: Der Mythos vom Priesterkönig Johannes, GWU 65, 2014, H1/2, S. 83–91. - Ulrich Knefelkamp: Der Priesterkönig Johannes und sein Reich – Legende oder Realität, Journal of Medieval History 14, 1988, S. 337–355. Unser neues Buch „Mehr Geschichten aus der Geschichte“ erscheint am 4. September. Es kann hier signiert vorbestellt werden: https://shop.autorenwelt.de/products/mehr-geschichten-aus-der-geschichte-von-richard-hemmer-und-daniel-messner //Aus unserer Werbung Du möchtest mehr über unsere Werbepartner erfahren? Hier findest du alle Infos & Rabatte: https://linktr.ee/GeschichtenausderGeschichte //Geschichten aus der Geschichte jetzt auch als Brettspiel! Werkelt mit uns am Flickerlteppich! Gibt es dort, wo es auch Becher, T-Shirts oder Hoodies zu kaufen gibt: https://geschichte.shop // Wir sind jetzt auch bei CampfireFM! Wer direkt in Folgen kommentieren will, Zusatzmaterial und Blicke hinter die Kulissen sehen will: einfach die App installieren und unserer Community beitreten: https://www.joincampfire.fm/podcasts/22 //Wir haben auch ein Buch geschrieben: Wer es erwerben will, es ist überall im Handel, aber auch direkt über den Verlag zu erwerben: https://www.piper.de/buecher/geschichten-aus-der-geschichte-isbn-978-3-492-06363-0 Wer unsere Folgen lieber ohne Werbung anhören will, kann das über eine kleine Unterstützung auf Steady oder ein Abo des GeschichteFM-Plus Kanals auf Apple Podcasts tun. Wir freuen uns, wenn ihr den Podcast bei Apple Podcasts oder wo auch immer dies möglich ist rezensiert oder bewertet. Wir freuen uns auch immer, wenn ihr euren Freundinnen und Freunden, Kolleginnen und Kollegen oder sogar Nachbarinnen und Nachbarn von uns erzählt! Du möchtest Werbung in diesem Podcast schalten? Dann erfahre hier mehr über die Werbemöglichkeiten bei Seven.One Audio: https://www.seven.one/portfolio/sevenone-audio

In dieser Folge gibt's wieder zwei Geschichten aus unserem ersten Buch: Richard liest Daniels Geschichte über frühe Landkarten, und Daniel liest Richards Geschichte über die Entstehung eines Gemäldes, das auf einer katastrophalen französischen Expedition basiert. // Literatur ///Mit dem Finger auf der Landkarte - Ute Schneider: Die Macht der Karten: Eine Geschichte der Kartographie vom Mittelalter bis heute. Darmstadt 2006. ///Wie das Floß der Medusa entstand - Jonathan Miles: The Wreck of the Medusa. The Most Famous Sea Disaster of the Nineteenth Century. New York 2008. Das Episodenbild zeigt Ausschnitte der Ebstorfer Weltkarte und des besprochenen Gemäldes des Floßes der Medusa. //Aus unserer Werbung Du möchtest mehr über unsere Werbepartner erfahren? Hier findest du alle Infos & Rabatte: https://linktr.ee/GeschichtenausderGeschichte //Geschichten aus der Geschichte jetzt auch als Brettspiel! Werkelt mit uns am Flickerlteppich! Gibt es dort, wo es auch Becher, T-Shirts oder Hoodies zu kaufen gibt: https://geschichte.shop // Wir sind jetzt auch bei CampfireFM! Wer direkt in Folgen kommentieren will, Zusatzmaterial und Blicke hinter die Kulissen sehen will: einfach die App installieren und unserer Community beitreten: https://www.joincampfire.fm/podcasts/22 //Wir haben auch ein Buch geschrieben: Wer es erwerben will, es ist überall im Handel, aber auch direkt über den Verlag zu erwerben: https://www.piper.de/buecher/geschichten-aus-der-geschichte-isbn-978-3-492-06363-0 Wer unsere Folgen lieber ohne Werbung anhören will, kann das über eine kleine Unterstützung auf Steady oder ein Abo des GeschichteFM-Plus Kanals auf Apple Podcasts tun. Wir freuen uns, wenn ihr den Podcast bei Apple Podcasts oder wo auch immer dies möglich ist rezensiert oder bewertet. Wir freuen uns auch immer, wenn ihr euren Freundinnen und Freunden, Kolleginnen und Kollegen oder sogar Nachbarinnen und Nachbarn von uns erzählt! Du möchtest Werbung in diesem Podcast schalten? Dann erfahre hier mehr über die Werbemöglichkeiten bei Seven.One Audio: https://www.seven.one/portfolio/sevenone-audio

This Week Opens Your Highest Timeline — But Only If You Choose It. A rare convergence of fate, truth, and magic is unfolding. This week brings one of the most powerful astrology alignments of the year: Venus conjunct Jupiter, Hecate on the North Node, Uranus square the Nodes of Fate, and a Gemini New Moon that opens the door to an entirely new chapter. The question is: Will you choose the safe path… or your highest timeline? In this astrology forecast for June 8–14, Sabrina Lynn explores the profound crossroads energy moving through the collective. We dive into how Hecate, the Queen of the Crossroads, is activating destiny pathways, why Venus conjunct Jupiter is amplifying joy, abundance, and soul-aligned opportunities, and how the Warrior for Truth (Eris) is demanding radical authenticity before the blessings can fully land. This isn’t just another astrology week. It’s a week of awakening. A week of reclaiming your wildness. A week of choosing the future your soul has been calling you toward. Watch now to understand the shift and reclaim your wildness. Inside This Transmission: • The Hecate Crossroads: Why the “safe road” is now the most dangerous place to be. • The Eris Truth-Bomb: Shattering the pedestal of false authority in your life. • Shadow Alchemy: Purifying the “Good Girl” identity to reclaim your wildness. • The Golden Seed: How to plant for the Gemini New Moon as your reclaimed self. Support for the Crossroads Return to Magic The feminine doesn’t open through force. She opens through safety, presence, and devotion. This month’s BONES workshop helps you regulate the nervous system, reclaim your connection to the miracle field, and awaken the deeper magic that is calling you forward. → Instant Access The Reclaimers Path Lilith. Medusa. Eris. Three Dark Goddesses guiding the return of your wildness, truth, and unapologetic power. If this week’s astrology is stirring something ancient within you, this is where you meet it face-to-face. → Enter the Path The Evolution of the Masculine The feminine is awakening rapidly. What is happening to the masculine in response? Sabrina and Stab explore the heart-opening, shadow work, and evolutionary pressures transforming the masculine archetype in real time. → Watch the Conversation Listen to “A Higher Timeline Is Opening… “ podcast here… Topics Explored in “A Higher Timeline Is Opening… ” podcast: (Times based off audio version) The Journey Through the Crossroads: (00:00) – Big Magic & Astrological Alignments (June 8–14) (01:10) – Hecate on the North Node: Opening Your Highest Timeline (02:15) – Venus Conjunct Jupiter & Eris: Why Truth is the Key to Bliss (03:15) – Preparing for the Gemini New Moon (Sunday, June 14th) (05:50) – Monday's Energy: The Demand for Masculine Evolution (07:10) – Black Moon Lilith & The Wildest Aspects of the Feminine (08:35) – Hecate & Medicine Woman Capacities: Reclaiming Your Magic (10:20) – Tuesday's Peak: Heart's Desires & The Truth-Bombing Warrior (13:20) – Lilith & Sedna: Shaking Off Social Norms & “Good Girl” Roles (15:10) – Workshop Spotlight: Return to Magic & Nervous System Safety (17:25) – Wednesday's Depth: Pluto, Merlin, & Alchemizing the Shadow (21:10) – Thursday's Initiation: Venus Square Chiron & Shamanic Wounds (26:40) – Uranus & the Kundalini Awakener: Shockwaves of Awakening (31:20) – Practical Examples: Shaking Off Career & Relationship Norms (34:20) – Venus Moves into Leo: Finding the Brave Heart to Shine (38:50) – New Moon Seed Planting: Owning Your Full Butterfly Self You can leave a comment or question for Sabrina on the YouTube version of this episode. Listen to after “A Higher Timeline Is Opening… “: June Is Changing the Rules: The Return of Magic This Week Marks the Crossing… The Old You Cannot Come With You Shadow Work with the Goddess (Kali, Lilith, Persephone, Hecate, Medusa)  STAY CONNECTED ReWilding Weekly (free, embodied astrology)  IG  Website  Disclaimer: Educational/spiritual perspectives; not medical/mental-health advice. #2025Shift #NewHuman #SpiritualAwakening Welcome to ReWilding with Sabrina Lynn & ReWilding for Women! A gifted facilitator of revolutionary inner work and the world's leading archetypal embodiment expert, Sabrina Lynn is the creator of the groundbreaking ReWilding Way and founder of ReWilding For Women. Sabrina has led more than 100,000 people through programs based on the ReWilding Way, a modality of healing and awakening that strips away the false, the deep wounds from early life, and the fears that hold people back, to reveal their true and unique soul light and help them build their innate capacity to shine it in the world. Her work includes in-person retreats and events, the monthly ReWilding Membership, Living Close to the Bone, Priest/ess Trainings, Mystery Schools, the ReWilding with the Archetypes, and the wildly popular 6 Faces of the Feminine workshop series. Welcome to ReWilding! The post 385 – A Higher Timeline Is Opening… What You Choose This Week Matters (June 8–14) appeared first on Rewilding for Women.

Dungeon master Alex Wilson is back to bring the finale of his epic three part Dungeons & Dragons campaign to a close pitting the guys against a Medusa hybrid from Shuddy Boy's nightmares, wager on the space ponies and face off against their deepest darkest demons!

Host Meg Wolitzer presents four stories, recorded at the Getty Center in Los Angeles, in which characters shape their expectations and dreams to a manageable size.  So if you're “Medusa,” as in our first story, by Tania James, you try to figure out how to live in the world instead of turning it to stone.  The reader is Constance Zimmer.  Parents in our second story, “We Only Wanted Their Happiness,” by Alexander Weinstein, make a tactical choice about technology.  It's performed by Randall Park.  The narrator of Honor Levy's “Good Boys,” read by Annie Hamilton, understands that infatuation is a phase.  And a man and a woman sidestep romance in “Arrangements” by Charlie Watts, performed by Laura Harrier and Will Harrison.  The program was created in cooperation with Belletrist, an online book club created by Emma Roberts and Karah Preiss. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.