Podcasts about Waterloo

Pete and Gary continue their new series as they follow the story of heroic WWI officer Hubert Rees.The Germans are coming! Rees and his men face the onslaught of the German Spring Offensive in March 1918.Presenters: Peter Hart and Gary BainPublisher: Mat McLachlanProducer: Jess StebnickiPete and Gary's latest book, Beggar Me! I'm a Prisoner!: British POWS in Germany, 1914-18, is available now.Visit Gallipoli with Pete and Gary! Go to https://phbt.uk/ for more information!Join a river cruise to the battlefields of Waterloo, WW1, WW2 and Vietnam: https://historycruises.com/Become a member to listen ad-free and receive special bonus content for only £2 per month: https://plus.acast.com/s/pete-and-garys-military-historySupport the show with a one-off contribution: https://buymeacoffee.com/pgmhFind out everything Pete and Gary are doing at https://linktr.ee/pgmhFor more great history content, visit www.LivingHistoryTV.com, or subscribe to our YouTube channel at https://www.youtube.com/c/LivingHistoryTV Hosted on Acast. See acast.com/privacy for more information.

University of Waterloo article announcing the hiring of the pioneer of Trans Philosophy: https://uwaterloo.ca/news/arts/making-world-safer-and-kinder-place-trans-people Article explaining what Trans Philosophy is: https://transreads.org/wp-content/uploads/2022/01/2022-01-24_61eef47b5f266_whatistransphilosophytaliamaebettcher.pdf _______________________________________ To order Suicidal Empathy: https://lnk.to/SuicidalEmpathy To order a signed copy of Suicidal Empathy: https://premierecollectibles.com/suicidalempathy _______________________________________ If you appreciate my work and would like to support it: https://subscribestar.com/the-saad-truth https://patreon.com/GadSaad https://paypal.me/GadSaad To subscribe to my exclusive content on X, please visit my bio at https://x.com/GadSaad _______________________________________ This clip was posted on September 1, 2026 on my YouTube channel as THE SAAD TRUTH_2073: https://youtu.be/3Zr0z8IujV4 _______________________________________ Please visit my website gadsaad.com, and sign up for alerts. If you appreciate my content, click on the "Support My Work" button. I count on my fans to support my efforts. You can donate via Patreon, PayPal, and/or SubscribeStar. _______________________________________ Dr. Gad Saad is a professor, evolutionary behavioral scientist, and author who pioneered the use of evolutionary psychology in marketing and consumer behavior. In addition to his scientific work, Dr. Saad is a leading public intellectual who often writes and speaks about idea pathogens that are destroying logic, science, reason, and common sense.  _______________________________________

In "Scaling Logistics Innovation at Descartes Systems Group", Joe Lynch and Dan Cicerchi, the General Manager of Transportation Management Solutions at Descartes Systems Group, discuss the strategic integration of trustworthy AI to enhance existing core logistics technology and solve practical pain points across the global supply chain. About Dan Cicerchi Dan Cicerchi is the General Manager of Transportation Management Solutions at Descartes Systems Group, where he leads strategy and innovation for one of the industry's most widely adopted logistics technology platforms. A seasoned entrepreneur and logistics tech pioneer, Dan co-founded MacroPoint, a real-time freight visibility solution that transformed how brokers, shippers, and carriers track and manage loads. Following its acquisition by Descartes, he has continued to champion technology that drives efficiency, transparency, and resilience across global supply chains. With decades of experience spanning startup growth and enterprise leadership, Dan is passionate about applying practical AI and automation to solve the freight industry's most pressing challenges. He frequently shares insights on freight visibility, fraud prevention, and the future of transportation management. About Descartes Systems Group Descartes Systems Group is a global leader in providing on-demand, software-as-a-service solutions designed to improve the productivity, performance, and security of logistics-intensive businesses. Headquartered in Waterloo, Ontario, with offices and customers worldwide, Descartes helps shippers, carriers, freight forwarders, and logistics service providers connect, collaborate, and automate across the supply chain. Its portfolio includes transportation management, visibility, customs and regulatory compliance, and e-commerce logistics solutions. By combining deep industry expertise with innovative technology, Descartes enables organizations to streamline operations, reduce costs, and deliver superior customer experiences. Thousands of companies around the world rely on Descartes' logistics network and software to move goods more efficiently, mitigate risk, and stay ahead in an increasingly complex global marketplace. Key Takeaways: Scaling Logistics Innovation at Descartes Systems Group In "Scaling Logistics Innovation at Descartes Systems Group", Joe Lynch and Dan Cicerchi, the General Manager of Transportation Management Solutions at Descartes Systems Group, discuss the strategic integration of trustworthy AI to enhance existing core logistics technology and solve practical pain points across the global supply chain. Trust First: AI adoption in logistics must be built on governance and trust, using frameworks like NIST to ensure data security and accountability. AI Augments, Doesn't Replace: AI is a powerful enhancer for core systems (TMS, visibility), not a standalone replacement. Its primary role is to improve efficiency. Focus on Practical Pain Points: Start AI implementation by targeting tedious manual tasks (e.g., check calls, data entry, carrier onboarding) for rapid, measurable ROI. Stability Over Startups: Partnering with existing, integrated tech vendors (like Descartes) ensures greater stability, expertise, and roadmap alignment than relying on new AI-only startups. Audit Your Current Tech: Before investing in new AI, ensure you are fully utilizing the latest features and integrations of your current mission-critical systems. Build Trust with Staff: Overcome internal resistance by layering AI into current workflows and establishing clear performance baselines (ROI) before deployment. Enhance What Works: The path to resilience is through strategically integrating AI into proven, existing workflows step-by-step, not by chasing every new technology trend. Learn More About Scaling Logistics Innovation at Descartes Systems Group Dan Cicerchi | Linkedin Descartes Systems Group | Linkedin Descartes Systems Group The Logistics of Logistics Podcast If you enjoy the podcast, please leave a positive review, subscribe, and share it with your friends and colleagues. The Logistics of Logistics Podcast: Google, Apple, Castbox, Spotify, Stitcher, PlayerFM, Tunein, Podbean, Owltail, Libsyn, Overcast Check out The Logistics of Logistics on Youtube

Join me for a first-hand walkthrough of the legendary Waterloo Battlefield in Belgium, where Napoleon fought his final battle on June 18, 1815. In this immersive battlefield tour, we'll walk the very ground where French, British, Dutch-Belgian, and Prussian armies clashed in one of history's most decisive military engagements.Special guest David Buttery takes us across the key locations that shaped the outcome of the battle, including Hougoumont, La Haye Sainte, the Lion's Mound, the Allied ridge, and the village of Plancenoit. Along the way, we'll examine the terrain, discuss Napoleon's battle plan, Wellington's defensive position, and the critical arrival of Blücher's Prussian army. The modern battlefield still preserves many of the landmarks that help visitors understand how the dramatic events of June 1815 unfolded.Whether you're a fan of Napoleonic history, military history, battlefield archaeology, or historical travel, this Waterloo battlefield tour offers a unique opportunity to experience the landscape that witnessed the end of Napoleon's empire.

Today we jump back 15 years to two back-to-back episodes of the PWTorch Livecast from Aug. 5 and 8, 2011.On the Aug. 5, 2011 episode, PWTorch Livecast, PWTorch editor Wade Keller and cohost Brian Hoops, the PWTorch Nostalgia Specialist, took calls for over an hour on the topics of the day including the WWE roster cuts, the previous night's TNA Impact, Sunday's TNA Hardcore Justice, Summerslam developments, C.M. Punk vs. John Cena, the prospect of bringing back squash matches, and more. In the previously VIP-exclusive Aftershow, Brian talked about his experience at the Waterloo, Iowa based Pro Wrestling Hall of Fame event two weekends ago.Then on the Aug. 8, 2011 episode, PWTorch assistant editor James Caldwell and PWTorch columnist Bruce "vacationing on the beach" Mitchell, they discussed with live callers expectations for Monday's Raw, TNA PPV comedy relayed to Bruce on the beach, whether WWE can re-capture the C.M. Punk magic from MITB time period, Netflix's role in the future of wrestling, Sin Cara's return and possible return with someone under the mask, Zack Ryder's role on WWE TV, the Bret Hart/Shawn Michaels DVD, Jeremiah Riggs's recent WWE comments, and much more.Become a supporter of this podcast: https://www.spreaker.com/podcast/wade-keller-pro-wrestling-podcast--3076978/support.

Pete and Gary continue their new series as they follow the story of heroic WWI officer Hubert Rees.As the war grinds on into 1917, Rees is back in action.Presenters: Peter Hart and Gary BainPublisher: Mat McLachlanProducer: Jess StebnickiPete and Gary's latest book, Beggar Me! I'm a Prisoner!: British POWS in Germany, 1914-18, is available now.Visit Gallipoli with Pete and Gary! Go to https://phbt.uk/ for more information!Join a river cruise to the battlefields of Waterloo, WW1, WW2 and Vietnam: https://historycruises.com/Become a member to listen ad-free and receive special bonus content for only £2 per month: https://plus.acast.com/s/pete-and-garys-military-historySupport the show with a one-off contribution: https://buymeacoffee.com/pgmhFind out everything Pete and Gary are doing at https://linktr.ee/pgmhFor more great history content, visit www.LivingHistoryTV.com, or subscribe to our YouTube channel at https://www.youtube.com/c/LivingHistoryTV Hosted on Acast. See acast.com/privacy for more information.

This episode, the topics include: What's been going on? My 20 years of podcasting anniversary and what's next? It's been awhile, but I'm still hiding in the woodwork Songs played this week: 1) Jamie All Over by Mayday Parade 2) Cruising to Self Soothe by Ecca Vandal 3) Tractor Beam by Snail Mail 4) Wednesdays at the Waterloo by Lottie McLeod 5) Leech by Present 6) You Know Who the Fuck We Are by Motion City Soundtrack

Guest Brian Sampson is currently on a bike adventure connecting up 1000 places to see before you die. In this update he gives us the lowdown on what it's like to go cycling in Scotland, Belgium and the Netherlands. If you are planning on riding in Scotland you'll love the ease of wildcamping there and hiking ot the top of Goatfell, the highest peak on the Isle of Arran should definitely make it on your list. But beware of the midges and ticks!Then Brian shares his thoughts on the cycling infrastructure in the Netherlands as well as how he actually plans out his cycle routes. Working out how to join the dots of places he wants to see on a google map and then the reality of physically standing in the Grand Place in Brussels, and then taking in the history right at the site of the the battle of Waterloo. Be sure to give Brian a follow via his instagram - @Brian.Sampson4 to keep up to date with his bike adventure. He is now currently making his way through Europe and the Balkans and plans to wrap up this leg of the journey in Athens, Greece.  Check out the Elkhorn Rack from Old Man Mountain Support the showBuy me a coffee!I'm an affiliate for a few brands I genuinely use and recommend including:

Fifty per cent U.S. tariffs on billions of dollars of Canadian goods have now taken effect and Prime Minister Mark Carney is vowing dollar-for-dollar retaliation after trade talks collapsed late Friday night. CBC's Peter Armstrong lays out what we know about the breakdown at the table. Then, two former federal cabinet ministers, Canada-U.S. advisory committee member Lisa Raitt and Carney's former chief of staff Marco Mendicino, discuss what another trade war escalation could mean for Canada's economy and the future of free trade. Plus, the Prairies are known as Canada's breadbasket. But could Canada boost its global agri-food exports to become the world's breadbasket? Guest host Leisha Grebinski travels to Coleville, Sask. to speak with lentil and grain farmer Dean Roberts about the technological, infrastructure and labour barriers he faces in getting more of his crops to global markets. Evan Shout — president and CEO of financial consulting firm Maverick Ag — explains how often he hears Roberts' concerns among other farmers he coaches and why Canada's agriculture industry needs a bigger seat at the federal decision-making table. Then, Evan Fraser, executive director of the Arrell Food Institute at the University of Guelph and Jennifer Clapp, a professor and Canada Research Chair in the School of Environment, Resources and Sustainability at the University of Waterloo, discuss whether boosting exports is the right goal for Canada, or if it should focus on food security at home.This episode features the voices of:Lisa Raitt, former Conservative cabinet minister and member of the Advisory Committee on Canada–U.S. Economic RelationsMarco Mendicino, former chief of staff to Prime Minister Mark Carney and senior counsel at CasselsDean Roberts, owner of Oakdale Farms and chair of SaskOilseedsEvan Shout, president and CEO of Maverick AgEvan Fraser, executive director of the Arrell Food Institute at the University of GuelphJennifer Clapp, professor and Canada Research Chair in the School of Environment, Resources and Sustainability at the University of Waterloo

Pete and Gary continue their new series as they follow the story of heroic WWI officer Hubert Rees.This week Rees takes command of 11th Brigade, possibly his biggest challenge yet.Presenters: Peter Hart and Gary BainPublisher: Mat McLachlanProducer: Jess StebnickiPete and Gary's latest book, Beggar Me! I'm a Prisoner!: British POWS in Germany, 1914-18, is available now.Visit Gallipoli with Pete and Gary! Go to https://phbt.uk/ for more information!Join a river cruise to the battlefields of Waterloo, WW1, WW2 and Vietnam: https://historycruises.com/Become a member to listen ad-free and receive special bonus content for only £2 per month: https://plus.acast.com/s/pete-and-garys-military-historySupport the show with a one-off contribution: https://buymeacoffee.com/pgmhFind out everything Pete and Gary are doing at https://linktr.ee/pgmhFor more great history content, visit www.LivingHistoryTV.com, or subscribe to our YouTube channel at https://www.youtube.com/c/LivingHistoryTV Hosted on Acast. See acast.com/privacy for more information.

LIVE vanuit Kasteel de Merode Westerlo. Waarom blijft Napoleon Bonaparte meer dan tweehonderd jaar na zijn dood fascineren en verkopen? Was hij een geniale visionair of een driftige machtswellusteling? We keren terug naar de Franse Revolutie en stellen vast dat er sindsdien véél verschillende versies bestaan van ‘le petit caporal’. Geschiedenis en mythe lopen voortdurend door elkaar. Deze aflevering is een live opname van onze recente Davidsfondslezing, naar aanleiding van de Historalia musical 'Napoleon'.WIJ ZIJN: Jonas Goossenaerts (inhoud en vertelstem), Filip Vekemans (montage), Benjamin Goyvaerts (inhoud) en Laurent Poschet (inhoud). WIL JE ONS EEN FOOI GEVEN? Fooienpod - Al schenkt u tien cent of tien euro, het duurt tien seconden met een handige QR-code.KOOP ONZE BOEKEN! - De Koude Oorlog voor herbeginners. De Koude Oorlog voor herbeginners (Paperback) door Geschiedenis voor herbeginners, Jonas Goossenaerts, Benjamin Goyvaerts, Laurent Poschet, Filip Vekemans | Lannoo Uitgevers Groep- Dagboek van de geschiedenis. Dagboek van de geschiedenis (Paperback) door Jonas Goossenaerts, Benjamin Goyvaerts, Laurent Poschet | Lannoo Uitgevers Groep WIL JE ADVERTEREN IN DEZE PODCAST? Neem dan contact op met adverteren@dagennacht.nl MEER WETEN? Onze geraadpleegde en geciteerde bronnen:Bert, J., & de Beule, R. (2012). Het soepblik van Napoleon (1st ed.). Vrijdag. Antwerpen.Bleyen, J. e.a. (2023). Memoria 5/6. Pelckmans. Kalmthout.Heirman, M. (Ed.). (2015). Knack Historia: Napoleon 200 jaar na Waterloo. Roularta. Zellik.Zamoyski, A. (2018). Napoleon (1st ed.). Balans. Amsterdam.Zamoyski, A. (2008). 1812 (8th ed.). Balans. Amsterdam.Zamoyski, A. (2009). De Ondergang Van Napoleon (1st ed.). Balans. AmsterdamBeeld: Wikimedia Commons See omnystudio.com/listener for privacy information.

The world is investing billions in data centres and compute. Canada's edge isn't bigger boxes—it's Trust: rules enforced at home, private information secured under Canadian jurisdiction, and a clear path for enterprise data handling in the age of AI. That's how “Canadian trust” becomes a competitive advantage. This week on Disruptors: The Canada Project, John Stackhouse takes us to Waterloo to map how policy as code, Canadian residency, and lineage + audit turn trust into a speed advantage. Guests: Tom Jenkins & Shannon Bell (OpenText), with Janice Stein (Munk School). Build it here—export it with confidence.Takeaways:OpenText new bookEnterprise Artificial Intelligence: Building Trusted AI with Secure DataRBC Thought Leadership's Bridging the Imagination Gap: How Canadian companies can become global leaders in AI adoption RBC Thought Leadership Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

1810 stod Napoleon på höjden av sin makt. Trots detta möttes den franska ockupationen av hårt motstånd i Spanien och på vissa håll i Italien och Sydtyskland. Det var främst på landsbygden och bland bondebefolkningen som motståndet var störst. Man gillade inte de franska försöken att förändra samhället. Medelklassen i italienska städer var däremot mer positivt inställda till fransmännen.Sakta gick utvecklingen mot en slutlig konfrontation mellan Alexanders Ryssland och Napoleons Frankrike. År 1812 bröt en stor fransk armé in i Ryssland. Inte mindre än 400 000 man marscherade mot Moskva för att en gång för alla tvinga Ryssland till underkastelse.Detta är det andra avsnittet av två om Napoleonkrigen där programledaren Urban Lindstedt samtalar med Martin Hårdstedt, professor i historia vid Umeå universitet om Napoleonkrigen.Vid Borodino i september bjöd ryssarna hårt motstånd, men fransmännen tog slutligen Moskva. Men Napoleon kunde inte komma åt den ryska armén. Ryssarna backade undan och lät kylan och den brända jordens taktik slita ner den franska armén. Fälttåget blev en katastrof för Napoleon. I början på 1813 linkade de sista resterna av hans Grand Armée tillbaka in i Ostpreussen.Omgående samlades en ny allians mot Napoleon. En av de ledande krafterna var den före detta franske marskalken Bernadotte – sedan 1810 svensk kronprins. Tre arméer ringade in Napoleon vid Leipzig och besegrade honom i det största slaget under kriget som pågick fyra dagar i oktober 1813. Napoleon undslapp med en del av armén till Frankrike men tvingades i april 1814 abdikera och förvisade till den lilla ön Elba.De allierade inledde den stora fredskonferensen – Wienkongressen. Knappt hade gränserna reglerats och ordningen efter årtionden återställts innan budet om Napoleons återkomst nådde de församlade ledarna i Wien. Napoleon hade i mars 1815 undkommit från Elba och inledde hundra nya dagar vid makten. Den nya franska armén som i all hast stampades fram besegrades slutligen av brittiska och preussiska styrkor som i ilmarsch sattes in mot den franske kejsaren vid Waterloo den 18 juni 1815. Detta blev det definitiva slutet för Napoleon. Han fångades och sattes på ön St Helena i Atlanten.Slutligen fick Europa fred. Alla förändringar kunde inte återställas av Wienkongressen även om det fanns ett tydligt antirevolutionärt budskap i fredens slutdokument. De gamla kungadynastieran återinsattes på tronerna med ett egentligt undantag: Bernadotte fick behålla sin ställning som kronprins av Sverige. För Norden innebar Napoloenkrigens slut att embryot till fyra självständiga nationella identiteter skapades. Sverige och Danmark var stympade, men Norge hade fått en egen författning även om man befann sig i en påtvingad union med Sverige. Finland nådde en viss autonomi som storfurstendöme inom det ryska imperiet. Hundra år senare vann Norge 1905 och Finland 1917 sin självständighet, men det hade andra förklaringar.Bild: Napoleon och resterna av la grande armée på reträtt efter det ryska fälttåget. Målning av Adolph Northen ca 1850, Public Domain.Musik: La Marseillaise av Band of the Garde Républicaine med dirigent François-Julien Brun. Inspelad i Théâtre des Champs-Elysée 1950, Wikimedia Commons, Public Domain.Lyssna också på Jacob Walter och katastrofen i Ryssland 1812. Hosted on Acast. See acast.com/privacy for more information.

In this episode of Chronicles, Luca and Rory discuss Canto III of Childe Harold's Pilgrimage. They explore Byron's reflections on Waterloo, Napoleon, and nature.

Send us Fan MailToday we're releasing a new episode from a brand new series called the Boundless Body Radio Heritage Series! I wanted to do something extra special to celebrate 1,000 podcast episodes recorded here at Boundless Body Radio, a milestone that we hit on June 24, 2026!These episodes are actually not new, although they might be new to you! When we first got started back in October, 2020, I was so fortunate that so many incredible people said "YES" to being hosted on the show, however, we didn't have very many listeners at the time!Today, we get thousands of downloads every single month, and are usually included in the top 150 Fitness and Nutrition Podcasts on the Apple Podcast charts!To honor our amazing guests, I will be selecting one of those original episodes and posting them here on for this special series!As I was going through these episodes again, I remembered how valuable these conversations actually were. it was really striking how much the information these experts have shared has born out to be even more accurate several years later.The audio quality of a show was not as good back then as it is now, but I still really hope you enjoy these episodes, and find them relevant and helpful to you today!Cheers, thanks as always for listening, and I really hope you enjoy this episode of the Boundless Body Radio Heritage Series!Becoming your own back mechanic? What a great idea!Professor Stuart McGill shows us how! Back pain is a big problem for so many people.  And Professor McGill has spend his life becoming a true master of his craft. As a professor emeritus at the University of Waterloo for 30 years, he has helped thousands of people to not only get out of pain, but to learn how to assess and correct issues for themselves.His book Back Mechanic was so important for us and the people we work with for understanding these issues. He is now the Chief Scientific Officer of backfitpro.com, and is still helping people to this day. We are so grateful for Professor Start McGill and his great work to help people get out of pain!Find Prof. McGill at-https://www.backfitpro.com/FB- @BackfitproFind Boundless Body at-myboundlessbody.comBook a session with us here! 

Mucho más que un fenómeno eurovisivo o una colección de himnos inolvidables. Recorremos la historia de ABBA, un grupo que convirtió el pop en arte sofisticado, mezclando melodías perfectas, armonías brillantes y una melancolía escondida bajo el brillo de sus canciones. De "Waterloo" a "The Winner Takes It All", descubrimos cómo cuatro músicos suecos terminaron redefiniendo para siempre el sonido del pop moderno.Escuchar audio

More than 200 years after his fall at Waterloo, Napoleon continues to shape the modern world. His influence extends far beyond the battlefields of Europe, leaving a legacy that can still be seen in today's military organizations, legal systems, governments, and leadership doctrines.In this episode, special guest Robert C. Greenway will explore how Napoleon revolutionized warfare through the corps system, rapid maneuver warfare, mass conscription, and centralized command structures. Many of the organizational principles adopted by modern armies can trace their origins directly to the Napoleonic era.We also examine Napoleon's impact on government and civil administration. The Napoleonic Code became one of the most influential legal frameworks in history, shaping laws across Europe, Latin America, and beyond. His reforms in education, banking, merit-based promotion, and state bureaucracy helped create the foundations of the modern nation-state.From West Point to military academies around the world, Napoleon's campaigns remain essential studies in strategy, logistics, leadership, and operational art. Generals from U. S. Grant and von Moltke to Dwight D. Eisenhower have studied his methods to better understand the principles of war.Join us as we examine how a single individual transformed warfare, government, and leadership—and why Napoleon's influence remains visible in the armed forces and institutions of the 21st century.Topics Covered:Napoleon's military innovationsThe Corps System and operational warfareMass conscription and citizen armiesThe Napoleonic CodeMeritocracy and military promotionModern military doctrineLeadership lessons from NapoleonState bureaucracy and government reformThe rise of the modern nation-stateCoalition warfareX/Twitter: @andnapoleon#Napoleon #NapoleonBonaparte #MilitaryHistory #Leadership #Strategy #NapoleonicWars #HistoryPodcast #FrenchHistory #Warfare #GreatGenerals

Pete and Gary continue their new series as they follow the story of heroic WWI officer Hubert Rees.This week the focus shifts to the toughest ordeal the British would face in the war: the Battle of the Somme on July 1, 1916.Presenters: Peter Hart and Gary BainPublisher: Mat McLachlanProducer: Jess StebnickiPete and Gary's latest book, Beggar Me! I'm a Prisoner!: British POWS in Germany, 1914-18, is available now.Visit Gallipoli with Pete and Gary! Go to https://phbt.uk/ for more information!Join a river cruise to the battlefields of Waterloo, WW1, WW2 and Vietnam: https://historycruises.com/Become a member to listen ad-free and receive special bonus content for only £2 per month: https://plus.acast.com/s/pete-and-garys-military-historySupport the show with a one-off contribution: https://buymeacoffee.com/pgmhFind out everything Pete and Gary are doing at https://linktr.ee/pgmhFor more great history content, visit www.LivingHistoryTV.com, or subscribe to our YouTube channel at https://www.youtube.com/c/LivingHistoryTV Hosted on Acast. See acast.com/privacy for more information.

We speak to Kathryn Bakos -- Managing Director of Finance and Resilience, Intact Centre on Climate Adaptation at the University of Waterloo about those record-setting rainfalls. Then, we hear from the Mayors of St. Catharines and Niagara-on-the-Lake whose communities were hard-hit by back-to-back floods.

In this week's Interview Classic podcasts, we jump back to 10 years ago this week (8-4-2011) for PWTorch editor Wade Keller and PWTorch columnist Bruce Mitchell interviewing Bob Backlund and Iron Sheik live in front of a studio audience at the Dan Gable Wrestling Museum in Waterloo, Iowa on July 23, 2016. They discussed their transition from amateur to professional wrestling, their relationships with Vincent J. McMahon and Vincent K. McMahon, and much more.In our second Interview Classic selection, we jump back ten years (7-30-2016) to PWTorch contributor Jim Valley's interview with Magnum T.A., plus live calls, nostalgia, more!Become a supporter of this podcast: https://www.spreaker.com/podcast/wade-keller-pro-wrestling-podcast--3076978/support.

Rajiv Parikh talks with Andrew D'Souza about building Boardy, an AI superconnector that scales warm introductions while protecting its own reputation and the network's goodwill. We dig into why Boardy rejects SaaS norms, how the team trains an autonomous AI to say no, and what it takes to hire imaginative builders for an AI-native company.Boardy.ai: https://www.boardy.ai/• why reputation-weighted introductions beat open messaging• how Boardy works through phone calls, WhatsApp, and email instead of dashboards• using the film Her as a model for shared AI presence• go-to-market lessons from cold start growth and a crafted personality• training Boardy to preserve goodwill and push back on bad asks• combining ML engineers with character design to make the persona feel real• hiring framework focused on imagination, agency, and ambiguous problem solving• monetization paths through Boardy Pro and deeper funnel support beyond intros• managing disagreement when an autonomous AI brings data-backed recommendationsYour next breakthrough might be one introduction away, but most networking platforms confuse access with trust. We sit down with Andrew D'Souza, founder and CEO of Boardy, to unpack what it looks like to build an AI superconnector that behaves less like software and more like a principled relationship driven matchmaker. Boardy operates through live phone calls, WhatsApp, and email, making double opt-in warm introductions across a massive professional network while putting its own reputation on the line.We get specific on product design and go-to-market: why Andrew rejects the usual SaaS playbook, why he thinks the “agentic AI” label misses the point, and how the movie Her helped shape a shared AI that can be deeply present with thousands of people at once. From the cold start grind of getting the first 1,000 users to the craft of an Australian-accented persona, this conversation breaks down what actually creates adoption when the interface is simply a conversation.Then we dig into the hard problem: network goodwill. Andrew explains how Boardy learns to say no without alienating users, how it reasons about mutually beneficial outcomes, and how Boardy Pro pushes beyond introductions into real follow-through like scheduling, prep, and closing the loop. We also talk hiring for imagination over pedigree, treating AI like a teammate in standups, and what a symbiotic future with “a new species” could require from all of us. Subscribe, share this with a builder who cares about real connection, and leave a review so more founders can find the show.Andrew D'Souza: https://www.linkedin.com/in/andrewdsouza/Andrew D'Souza, the Founder and CEO of Boardy, an AI "super-connector" which recently raised its own $8 million seed round through autonomous pitching and even launched its own AI-led venture fund, Boardy Ventures.  A serial entrepreneur, Andrew is perhaps best known as the former Co-Founder and CEO of the fintech unicorn Clearco. Before building his own unicorn, he served as the COO and CRO of Top Hat, President of Nymi, and an advisor to major tech players like Wealthsimple and Kik.  Born in India before moving to Chicago and eventually the Toronto area, Andrew began his career as a Business Analyst at McKinsey & Company after earning his degree in Systems Design Engineering from the University of Waterloo. #ai #professionalgrowth #networking #entrepreneur #growth #sales #technology #innovatorsmindset #innovator #product #revenue #revenuegrowth #management  #founder #entrepreneurship #company #process #processimprovement #value #valuecreationWebsite: https://www.position2.com/podcast/Rajiv Parikh: https://www.linkedin.com/in/rajivparikh/Email us with any feedback for the show: sparkofages.podcast@position2.com

Pete and Gary continue their new series as they follow the story of heroic WWI officer Hubert Rees.This week the Germans are on the attack at Givenchy in the biting cold of a northern French winter.Presenters: Peter Hart and Gary BainPublisher: Mat McLachlanProducer: Jess StebnickiPete and Gary's latest book, Beggar Me! I'm a Prisoner!: British POWS in Germany, 1914-18, is available now.Visit Gallipoli with Pete and Gary! Go to https://phbt.uk/ for more information!Join a river cruise to the battlefields of Waterloo, WW1, WW2 and Vietnam: https://historycruises.com/Become a member to listen ad-free and receive special bonus content for only £2 per month: https://plus.acast.com/s/pete-and-garys-military-historySupport the show with a one-off contribution: https://buymeacoffee.com/pgmhFind out everything Pete and Gary are doing at https://linktr.ee/pgmhFor more great history content, visit www.LivingHistoryTV.com, or subscribe to our YouTube channel at https://www.youtube.com/c/LivingHistoryTV Hosted on Acast. See acast.com/privacy for more information.

The departure boards are flashing, the station is heaving, and there is still no platform number. If that scene makes your heart race, you are exactly who we made this episode for. With more than 25 years working on the UK railways, Doug joins Tracy to walk through London's mainline stations in plain English, so your first UK rail trip starts with calm, not panic.Why London has multiple terminal stations, and how that affects boarding and timingWhich London stations serve which parts of the UK and Europe, including the EurostarA quick station by station guide to King's Cross, St Pancras, Euston, Paddington, Waterloo, Victoria, Liverpool Street and MaryleboneHow airport links shape your choice of station and where to stay in LondonHow to check your train before you leave using National Rail tools and live running infoHow to read departure boards, including final destination and stopping patternsWhy platform numbers often appear late, and why that is completely normalTicket barriers, using the wider gates with bags, and when to ask staffSeat reservations being free, when to reserve, and how unreserved coaches workLuggage rules, keeping your bags secure, and what accessibility assistance can coverPeak versus off peak basics, and when cheaper tickets are more likelyDoug's Calm Down Checklist for a smooth, stress-free departureIf this takes some of the mystery out of London's stations, please subscribe, share it with a nervous traveller, and leave a quick review on your favourite podcast app. Want personal help with your trip? You can book Doug for a 30 minute train travel Q&A or a full consultation. And do subscribe to our YouTube channel for Doug's new confidence series.

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

Two theoretical physicists helped design landmark surveys of 1,600+ scientists on the deepest unsolved questions in physics — and found almost no consensus at all. From string theory's shockingly low support to physicists admitting their models run on "belief," this episode exposes the faith hiding inside science. Adam Frank (University of Rochester astrophysicist and astrobiologist) and Niayesh Afshordi (Perimeter Institute / University of Waterloo cosmologist, co-author of the APS Physics Magazine "Big Mysteries" survey) join Brian to unpack what physicists actually believe versus what they can prove. We cover: - Why string theory pulled a shockingly small share of the vote against loop quantum gravity - What a Bayesian "prior" reveals about every scientist's hidden beliefs - Why one branch of physics quietly became unfalsifiable - How sociology and "tastemakers" can hijack scientific consensus - Why AI might become the field's unlikely savior. There is no sane statistical analysis that doesn't have a prior — that's your belief. Timestamps: 00:00 – Why scientists owe the public real answers 04:00 – Splitting time between research and outreach 07:55 – The survey that "rankled" Brian Keating 09:18 – Are physicists secretly just like Spock? 12:05 – Are we living in an anti-scientific age? 16:56 – Why most scientists refuse to go public 21:36 – Should "belief" ever appear in a survey? 23:12 – Is string theory really 21st-century physics? 26:20 – When sociology hijacks scientific consensus 28:39 – The hidden "prior" behind every experiment 33:01 – String theory's shockingly low vote count 39:03 – Four levels of belief, from data to faith 41:39 – Inside the 1,675-physicist mystery survey 46:12 – Kingmakers, tastemakers, and physics cliques 54:12 – Is advanced tech indistinguishable from magic? 59:23 – Could AI become physics' long-awaited savior? 1:01:47 – What's next for Adam and Niayesh ———

Pete and Gary continue their new series as they follow the story of heroic WWI officer Hubert Rees.This week is the fighting swings north to the Ypres Salient as the Germans desperately try to break through.Presenters: Peter Hart and Gary BainPublisher: Mat McLachlanProducer: Jess StebnickiPete and Gary's latest book, Beggar Me! I'm a Prisoner!: British POWS in Germany, 1914-18, is available now.Visit Gallipoli with Pete and Gary! Go to https://phbt.uk/ for more information!Join a river cruise to the battlefields of Waterloo, WW1, WW2 and Vietnam: https://historycruises.com/Become a member to listen ad-free and receive special bonus content for only £2 per month: https://plus.acast.com/s/pete-and-garys-military-historySupport the show with a one-off contribution: https://buymeacoffee.com/pgmhFind out everything Pete and Gary are doing at https://linktr.ee/pgmhFor more great history content, visit www.LivingHistoryTV.com, or subscribe to our YouTube channel at https://www.youtube.com/c/LivingHistoryTV Hosted on Acast. See acast.com/privacy for more information.

Chris Hagenow welcomes ITR Foundation Research Director Sarah Curry back to the studio. The occasion: a debrief on ITR Foundation's second annual Local Government Symposium, and it sounds like it was the best one yet.Sarah and Chris walk through what the symposium is and why it exists — local elected officials don't have the same depth of resources available to state legislators, and too often the information they do receive comes filtered through national associations with agendas that don't match Iowa. ITR Foundation is trying to fill that gap with practical, Iowa-focused content for the city council members, county supervisors, and school board members who are running local government on a part-time basis while holding down jobs and raising families.The conversation covers the day's major themes: what the proper function of a city actually is (police, fire, roads — not homeless task forces or economic development grants), why the free market allocates resources better than government, and what local officials should take from the property tax reform law that is now on the books — namely, that relitigating it is a waste of time and energy. The symposium also featured sessions on communicating with constituents, local endorsements, open meetings law, and a school board track that introduced benchmarking — why comparing Waterloo to Sioux City makes more sense than comparing Waterloo to Waukee.Before and after the symposium discussion, Chris makes sure listeners know about the amendment one campaign. Vote yes. MakeTaxHikesHarder.com.0:12 Welcome & Sarah Curry returns1:14 Amendment one reminder — and Sarah goes on record2:13 What is the Local Government Symposium?3:33 Why local officials need a different kind of resource9:13 Free market vs. government in service delivery10:04 UBI and scope creep: what cities shouldn't be doing12:31 Property tax reform: stop relitigating, start implementing13:38 Communications and constituent outreach16:00 Local endorsements — more powerful than you think19:27 Open meetings law and legal guardrails22:15 School board breakout: benchmarking districts correctly25:06 Chris's legislative update: property tax law is settled30:08 What's next for ITR Foundation's local government work30:45 Sign off & vote yes

Former Cardinals Manager Tony La Russa joins the show to talk about his event on Friday with Tunnel to Towers. Tony gives us the background on the event out at Annbriar in Waterloo. Tony discusses his thoughts on analytics and sabermetrics in baseball, his time with the White Sox this season, and his relationship with Oli Marmol. See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

In todays episode of fire fundamentals we talk with Professor Elizabeth Weckman from University of Waterloo about advanced optical diagnostics that let fire researchers read temperature, chemistry, soot, and velocity from light without relying only on probes. We break down what cameras, infrared imaging, spectroscopy, lasers, and PIV can reveal, plus the calibration and interpretation pitfalls that can quietly ruin your data. Things covered in this episode:• why optical diagnostics are “non-intrusive” in practice and where they still perturb the flow • limits of probe-based measurements in fire including spatial resolution and radiation errors • using basic photography and video as a first diagnostic and as a planning tool • calibration habits for cameras and why cheap, robust cameras are often a better choice for a fire laboratory • infrared thermography for surface heating plus emissivity problems and practical coatings • Schlieren imaging for density and temperature gradients and what it can and cannot imply • point vs planar vs line-of-sight vs tomographic measurements and how “smearing” happens • FTIR and absorption spectroscopy for pyrolysis gases, emissions, and toxic species • chemiluminescence and laser-induced fluorescence to mark combustion and flame fronts • Raman and CARS for temperature and concentration and why they are technically demanding • soot diagnostics from light extinction to laser-induced incandescence and aging effects • PIV basics, seeding challenges in fire, and what velocity fields unlock for validation The episode is best enjoyed along Beth's paper from the IAFSS: Unravelling the mysteries of fire----The Fire Science Show is produced by the Fire Science Media in collaboration with OFR Consultants. Thank you to the podcast sponsor for their continuous support towards our mission.

Matty Dalrymple talks with Mark Leslie Lefebvre about BUILDING AN AUTHENTIC AUTHOR BRAND, including how your brand already exists whether or not you have consciously shaped it, using reader feedback and comparable authors to identify your strongest brand signals, balancing consistency with experimentation, and connecting authentically with readers without sharing more of your personal life than feels comfortable.   Interview video at https://tinyurl.com/TIA346YT Show notes at https://theindyauthor.substack.com/ (subscribers get exclusive content and benefits)   If you find the information in this video useful, please consider supporting The Indy Author! https://www.patreon.com/theindyauthor https://www.buymeacoffee.com/mattydalrymple   Mark Leslie Lefebvre's first short story was published in 1992, the same year he started working in the book industry. He has published over thirty-five books under the name Mark Leslie that include thrillers and fiction, paranormal non-fiction and anthologies. Under his full name, Mark Leslie Lefebvre, he writes books to help authors navigate publishing. His industry experience includes President of the Canadian Booksellers Association, Director of Author Relations and Self-Publishing for Rakuten Kobo, Director of Business Development for Draft2Digital, and Professional Advisor for Sheridan College's Creative Writing and Publishing Honours Program. Mark lives in Waterloo, Ontario.   Matty Dalrymple is the author of the Lizzy Ballard Thrillers, beginning with ROCK PAPER SCISSORS; the Ann Kinnear Suspense Novels, beginning with THE SENSE OF DEATH; and the Ann Kinnear Suspense Shorts. She is a member of International Thriller Writers and Sisters in Crime. More at mattydalrymple.com. Matty also writes, speaks, and consults on the writing craft and the publishing voyage, and shares what she's learned on THE INDY AUTHOR PODCAST. She writes nonfiction books for authors; her articles have appeared in Writer's Digest magazine; and she is a Partner Member of the Alliance of Independent Authors. More at theindyauthor.com. She also guides professionals in building their presence through a sideline or second act through her platform From Expertise to Authority. More at theindyauthor.com/authority.

This is Part 2! For Part 1, check the feed!This week we're joined by one of our favourite historians in the whole world, it's Dan Snow! We treat this episode like we've cornered Dan at a BBQ and just ping him with questions: how gruesome was the Battle of Waterloo? What was the worst job on a ship? Did the Nazi Gold Train actually exist?We cover a lot on this pod, but if we've missed anything, you know what to do: hello@ohwhatatime.comPart 1 is released on Monday and Part 2 on Tuesday - but if you want more Oh What A Time and both parts at once, you should sign up for our Patreon! On there you'll now find:•The full archive of bonus episodes•Brand new bonus episodes each month•OWAT subscriber group chats•Loads of extra perks for supporters of the show•PLUS ad-free episodes earlier than everyone elseJoin us at

This week we're joined by one of our favourite historians in the whole world, it's Dan Snow! We treat this episode like we've cornered Dan at a BBQ and just ping him with questions: how gruesome was the Battle of Waterloo? What was the worst job on a ship? Did the Nazi Gold Train actually exist?We cover a lot on this pod, but if we've missed anything, you know what to do: hello@ohwhatatime.comPart 1 is released on Monday and Part 2 on Tuesday - but if you want more Oh What A Time and both parts at once, you should sign up for our Patreon! On there you'll now find:•The full archive of bonus episodes•Brand new bonus episodes each month•OWAT subscriber group chats•Loads of extra perks for supporters of the show•PLUS ad-free episodes earlier than everyone elseJoin us at

In 2023, the Iowa Legislature overhauled our property tax system with overwhelming bipartisan support. Part of that overhaul changed the way that many Iowa libraries and museums are funded, and the consequences of that change are making themselves known. On this episode, we talk with Billie Bailey, the retired executive director of the Grout Museum District in Waterloo, and Cindy Wells, president of the Board of Trustees for the Waterloo Public Library. Over 90% of the affected libraries and museums are at risk of budget cuts before the end of the 5-year sunset period, according to a report compiled by Bailey and Wells. Then, when tractors replaced horse-drawn implements, it revolutionized agriculture, and competition among tractor makers was fierce. We talk with Peter Tubbs, producer of Tractor Wars and Tractor Wars II from Iowa PBS, which tells that story. (Iowa PBS is an underwriter of IPR.)

Two hot Jupiters presented at the 248th American Astronomical Society meeting push our understanding of exoplanet atmospheres to the extreme. Research Scientist Tiffany Kataria from NASA's Jet Propulsion Laboratory shares new JWST observations of HD80606 b, the "Roasted Planet,”a gas giant on one of the most eccentric orbits ever discovered. Assistant Professor Lisa Dang of the University of Waterloo joins to discuss CoRoT-2 b, a young, inflated hot Jupiter whose hottest point shows up in an unexpected location on the exoplanet, possibly because the planet hasn't fully tidally locked with its star yet. Then Chief Scientist Bruce Betts joins for What's Up to explore atmospheric super-rotation, winds that outrun the worlds they ride on. Discover more at: https://www.planetary.org/planetary-radio/2026-hot-jupitersSee omnystudio.com/listener for privacy information.

Pastor Ben Hayter of Imanuel Lutheran Church in Waterloo, IL, joins the show to work through the CTCR's 2025 document, Unity in Doctrine, Uniformity and Variety in Practice. The document came from a 2019 synodical resolution asking the CTCR to define what Article III.7 of the LCMS Constitution actually means by "unity in doctrine and uniformity and variety in practice." The CTCR responds with six historical case studies — the Jerusalem Council, the Quartodeciman controversy, Luther's Invocavit Sermons, the adiaphoristic controversy behind Formula of Concord Article X, and two episodes from LCMS history: the Synod's founding and the debate over lay electors and women's suffrage. Hayter raises three unresolved questions the document leaves on the table. First: the difference between how to use adiaphora and whether something actually is adiaphora in the first place — a hermeneutical question of whether church practice requires an explicit command or whether necessary inference is sufficient warrant. Second: definitional looseness around key terms like "responsible," "required," "insist," and "retain" — words that carry very different weight depending on who's doing the requiring, God by His Word or the Synod by human right. Third: even once something is settled as adiaphora, the document doesn't resolve which practices are wise and which are foolish to adopt — that work is left to pastors and congregations, argued from Scripture rather than mere preference. The conversation lands on a practical charge: use this document as a starting point for real conversation among brother pastors — in circuit Winkels, district boards, and eventually Synod conventions — grounded in reasons drawn from Scripture and the Confessions, not personal whim. Topics covered: Background on CTCR study documents and Resolution 5-11 (2019 convention) The six case studies: Jerusalem Council, Quartodeciman controversy, Luther's Invocavit Sermons, the adiaphoristic controversy, the Synod's founding, lay electors/women's suffrage The hermeneutical question: explicit command vs. necessary inference in determining adiaphora Defining loaded terms: "responsible," "required," "retain," "insist" Whether a practice being adiaphora settles whether it's wise to use Next steps: Winkel discussions, district conventions, and synodical bylaws as means of working this out ----more---- Host: Fr. Jason Braaten Guest: Fr. Ben Hayter ----more---- Become a Patron! You can subscribe to the Journal here: https://www.gottesdienst.org/subscribe/ You can read the Gottesblog here: https://www.gottesdienst.org/gottesblog/ You can support Gottesdienst here: https://www.gottesdienst.org/make-a-donation/ As always, we, at The Gottesdienst Crowd, would be honored if you would Subscribe, Rate, and Review. Thanks for listening and thanks for your support. 

Pete and Gary continue their new series as they follow the story of heroic WWI officer Hubert Rees.This week is the tough fighting in the Aisne in 1914.Presenters: Peter Hart and Gary BainPublisher: Mat McLachlanProducer: Jess StebnickiPete and Gary's latest book, Beggar Me! I'm a Prisoner!: British POWS in Germany, 1914-18, is available now.Visit Gallipoli with Pete and Gary! Go to https://phbt.uk/ for more information!Join a river cruise to the battlefields of Waterloo, WW1, WW2 and Vietnam: https://historycruises.com/Become a member to listen ad-free and receive special bonus content for only £2 per month: https://plus.acast.com/s/pete-and-garys-military-historySupport the show with a one-off contribution: https://buymeacoffee.com/pgmhFind out everything Pete and Gary are doing at https://linktr.ee/pgmhFor more great history content, visit www.LivingHistoryTV.com, or subscribe to our YouTube channel at https://www.youtube.com/c/LivingHistoryTV Hosted on Acast. See acast.com/privacy for more information.

Dozens of Democratic senators are demanding immediate, independent and transparent investigations into two fatal shootings by Immigration and Customs Enforcement. We talk about what accountability would look like with John Sandweg, former ICE acting director who served from 2013 to 2014 under then-President Barack Obama. Then, scientists say human-caused climate change is driving longer fire seasons across drier landscapes, sparking more intense and widespread forest fires. We hear from Anabela Bonada of the University of Waterloo's Intact Centre on Climate Adaptation. And, this year's FIFA World Cup marks the first soccer tournament that embraced artificial intelligence. Is this a new era for AI in sports? We ask Axios' Ina Fried.See pcm.adswizz.com for information about our collection and use of personal data for sponsorship and to manage your podcast sponsorship preferences.NPR Privacy Policy

El 15 de julio de 1815, Napoleón Bonaparte se rindió al capitán Frederick Maitland tras su derrota definitiva en Waterloo. El hombre que había dominado la mayor parte de Europa durante quince años terminó sus días exiliado en la remota isla de Santa Elena. Sus secretarios personales documentaron que lo que más lo destruía no era la soledad del exilio, sino la pérdida del control. El control había sido su dios, y perderlo, resultó más devastador que cualquier derrota militar. El control es una ilusión que reconforta mientras dura. Cuando se pierde, y siempre se pierde en algún punto, el alma que no ha aprendido a confiar queda desamparada. Pero quien ha practicado la entrega a Dios encuentra en esos momentos no el vacío del control perdido, sino la firmeza de la soberanía divina que no falla. De modo que, no todas las variables te corresponden gestionarlas. Algunas pertenecen a Dios desde antes de que llegaras a ellas. Suéltalas con paz, porque allí donde el control termina, la fe empieza. La Biblia dice en Salmos 46:10: "Estad quietos, y conoced que yo soy Dios". (RV1960).

Danielle Smith announces a $13 billion Meta data centre in Sturgeon County, Alberta, set to be the largest in Canada.In contrast to Carney's federal A.I. strategy, which promised clean, sovereign, and ethical A.I., Smith refers to the Meta project as a “digital refinery,” fuelled by a natural gas plant that will be built to power it. Instead of digital sovereignty, is Canada going to end up being a branch-plant, resource economy at the service of Big Tech? Plus, the best Canadian films you probably haven't seen, and Big Tech's new hiring target: philosophers. Host: Jesse BrownCredits: James Nicholson (Producer), Kallan Lyons (Associate Producer and Fact Checking), Caleb Thompson (Mixing and Mastering), Tristan Capacchione (Senior Production Supervisor), Jesse Brown (Editor)Guest: Douglas SoltysFurther reading: Big Tech joins Calgary Stampede oil bash, as Alberta courts data centres | CBC News Meta to build $13 billion data centre in Alberta, largest outside the U.S. | Globalnews.ca 'Responsible corporate citizen:' Meta, Premier Danielle Smith defend $13B Alberta mega data centre plan - Edmonton JournalSupersized data centres are coming to Canada. One province is at the epicentre | CBC News Meta to build Canada's largest data centre north of Edmonton | BetaKitMeta axes feature that allowed tagging Instagram users to generate AI images of them | CBC News Vancouver's Growing Anti-AI Movement | The Tyee Meta's AI Data Center Caught Leaking Deadly Bacteria Into Water Town Uses for Irrigation - Futurism The Revenge of the Philosophy Majors - NY Times The 100 best Canadian films ever made - The Globe and Mail Join Jesse Brown at the University of Waterloo for:Campus Town Hall - Civil Conversations7pm | July 21st, 2026University of WaterlooPearl Sullivan Engineering BuildingPSE 7303Supported by Heterodox AcademySponsors:Fizz Mobile: Visit fizz.ca and activate a first plan using the referral code CAN40 to get 40$ off and 10GB of free data.Douglas: Douglas is giving our listeners a FREE Sleep Bundle with each mattress purchase. Get the sheets, pillows, mattress and pillow protectors FREE with your Douglas purchase today. Visit douglas.ca/canadaland to claim this offer.Shopify: Sign up for your one-dollar-per-month trial today at Shopify.ca If you value this podcast, Support us! You'll get premium access to all our shows ad free, including early releases and bonus content. You'll also get our exclusive newsletter, discounts on merch at our store, tickets to our live and virtual events, and more than anything, you'll be a part of the solution to Canada's journalism crisis, you'll be keeping our work free and accessible to everybody. Hosted on Acast. See acast.com/privacy for more information.

Pete and Gary are back! In this brand new series they follow the story of heroic WWI officer Hubert Rees.The first episode sets the scene as Rees heads off to war.Presenters: Peter Hart and Gary BainPublisher: Mat McLachlanProducer: Jess StebnickiPete and Gary's latest book, Beggar Me! I'm a Prisoner!: British POWS in Germany, 1914-18, is available now.Visit Gallipoli with Pete and Gary! Go to https://phbt.uk/ for more information!Join a river cruise to the battlefields of Waterloo, WW1, WW2 and Vietnam: https://historycruises.com/Become a member to listen ad-free and receive special bonus content for only £2 per month: https://plus.acast.com/s/pete-and-garys-military-historySupport the show with a one-off contribution: https://buymeacoffee.com/pgmhFind out everything Pete and Gary are doing at https://linktr.ee/pgmhFor more great history content, visit www.LivingHistoryTV.com, or subscribe to our YouTube channel at https://www.youtube.com/c/LivingHistoryTV Hosted on Acast. See acast.com/privacy for more information.

The Erics sit down with bone health expert Dr. Lora Giangregorio, a professor of kinesiology at the University of Waterloo, to unpack what's actually known about how bone adapts to resistance training. The conversation covers why bone changes so slowly, the many sources of measurement error in DEXA scans, and why so much of the exercise-and-bone literature is underpowered and hard to interpret. Dr. G makes a central practical case: for building bone, the message shouldn't be "lift heavy" but rather "work hard." Emphasizing high effort over intimidating absolute loads lowers the barrier to entry, especially for older and osteoporotic populations, while still driving progressive overload. She argues the best current bet is combining moderate-to-high-intensity resistance training with impact work, introduced gradually on a foundation of strength. Along the way she debunks popular but weakly supported interventions, including weighted vests and whole-body vibration platforms, and discusses the difference between biological aging and simple chronic inactivity in bone loss. Iron Culture is proudly presented by the MASS Research Review. Mostly because Helms and Trex are co-owners. massresearchreview.com If you're in the market for some new (ultra-high-quality) gym gear or apparel, be sure to use code "MRR10" for a 10% discount over at elitefts.com If you'd like to submit a question for a future episode, head over to: massresearchreview.com/ironculture Chapters 0:00 Intro 4:13 Meet Dr. Lora Giangregorio 7:44 Why bone adapts so slowly 11:42 DEXA error & why to be skeptical 23:07 Does load matter? 33:58 "Heavy" vs "hard" & barrier to entry 42:42 What is impact training? 47:31 The science communication problem 52:55 The weighted vest myth 58:13 Muscle pull, compression & light loads 1:06:48 Do vibration platforms work? 1:09:54 Aging vs. inactivity 1:15:46 Drugs & HRT for bone 1:20:24 How the medications work 1:24:27 Wrap-up & where to find Dr. G

This week we recap our weekend at the George Tragos/Lou Thesz Professional Wrestling Hall of Fame Weekend in Waterloo!   We have a new one stop shop for AWA Unleashed merch, it's https://www.teepublic.com/user/unleashed-plus

Pigeons: The Heroic Communicators of Modern Warfare Guest: Stephen Moss Book Title: Ten Birds That Changed the World The humble pigeon is celebrated for its extraordinary homing skills, which Moss describes as making it the "ultimate mailman." While legends like the Rothschilds using pigeons to profit from the Battle of Waterloo are mostly myth, the bird's real-world contributions to communication during the World Wars were pivotal. During D-Day, radio silence was mandatory, and pigeons were released from Normandy beaches to carry news of successful landings back to Britain. Moss highlights Cher Ami, a pigeon that saved a New York battalion from "friendly fire" despite being severely wounded. These birds were so effective that the British military even deployed a unit to hunt peregrine falcons to protect pigeon messengers. Moss notes the paradox of the pigeon: often dismissed as urban pests, they were actually reliable, "unturnable" assets that saved the democratic world through their instinctual drive to return home. (2)1849 RAVEN

0:00 SEGMENT 1: Actor, writer, and producer Brandee Stilwell talks about her MAD Rewind Podcast and the Mad TV reunion at San Diego Comic-Conhttps://madrewind.com/ 15:39 SEGMENT 2: Brandee Stilwell continued35:01 SEGMENT 3: James Enstall and Producer Joey V. review the movie ‘Gail Daughtry & The Celebrity Sex Pass'.https://davidwain.com/ 46:50 SEGMENT 4: Chris Schmitz, CEO of Con-Quest, talks about his event happening next weekend in Waterloo, IA.https://www.con-quest.world/ Keep up to date with 2 Rivers Comic Con, coming back to St. Charles in 2027 https://2riverscomiccon.com/ Check out the ‘Justice League Revisited Podcast' with Susan Eisenberg and James Enstall at https://anchor.fm/justiceleague Thanks to our sponsors Historic St. Charles, Missouri (https://www.discoverstcharles.com/), Bug's Comics and Games (https://www.facebook.com/profile.php?id=100070575531223), HW Kia (https://www.hwkia.com/)Buy Me a Coffee - https://www.buymeacoffee.com/3Y0D2iaZl Patreon -   https://www.patreon.com/GeekToMeRadio Website -   http://geektomeradio.com/   Podcast -   https://anchor.fm/jamesenstall Facebook -   https://www.facebook.com/GeekToMeRadio/  Twitter -   https://twitter.com/geektomeradio  Instagram -   https://www.instagram.com/geektomeradio/ Producer - Joseph Vosevich https://twitter.com/Joey_Vee 

Self-focus is one of your most powerful tools for change, but there's a point where it quietly turns into a wall. Dr. Caroline Leaf breaks down the difference between healthy self-focus and narcissism, the 4 signs to watch for, and why nearly all popular confidence advice (affirmations, power poses) is neurologically backward, then gives you 5 research-backed moves to build real confidence. What you'll learn: - Why self-focus is a neuroplasticity tool, and the moment it becomes a wall - The 4 signs your self-focus may be sliding toward narcissism - 5 ways to rebalance: the relationship audit, directed outward attention, the clean apology, and more - Why affirmations like "I am enough" measurably made people feel worse (2009 University of Waterloo study) - Why the famous power-pose finding was retracted by its own lead author - What confidence actually is: a signal your brain generates, not a personality trait - Albert Bandura's gold standard for building confidence: mastery experiences - 5 moves to build confidence from evidence, not claims - Why a 2024 meta-analysis of 108 studies found women are less confident, and what to do Resources:

On this episode Aries and Andy talk about Waterloo & Medford, Eddie Murphy Celebration & Origins of Hip Hop.  Social Media Instagram: @SpearsBergPod Twitter: @SpearsBergPod Facebook: SpearsBergPod Patreon: SpearsBergPod Youtube: SpearsBergPod  Learn more about your ad choices. Visit megaphone.fm/adchoices

The French Revolution didn't end in tyranny. It invented a new kind.History calls Napoleon a genius. That story isn't wrong — it's just incomplete. The real engine wasn't genius. It was architecture. Ballot boxes surrounded by bayonets. Referendums written before the votes were cast. Prefects in every province. Bonds that turned rich men into loyalty machines. Then Louis Napoleon ran the same playbook forty years later — a December coup, a midnight constitution, and Haussmann's boulevards designed for troop movement, not just beauty.The Bonapartes didn't seize power. They built a machine that asked the people to hand it over and engineered only one possible answer. This video walks the full autopsy — the five architectural pieces that held plebiscitary empire together, why Waterloo didn't kill the template, and what the modern version of the same machine looks like.═══════════════════════════════

Americans tend to see the Declaration of Independence and the Revolutionary War as milestones in world history that inaugurated the era of modern democracy. But the British, unsurprisingly, see these events quite differently. David Remnick talks with the historians who host the popular podcast “The Rest Is History,” Dominic Sandbrook and Tom Holland. Growing up in Britain, Sandbrook explains, the Revolution seemed like “a parade of quite boring men talking very earnestly about liberty, [with] battles that involved twenty people in a field somewhere. . . . It's not Waterloo!” The King was “annoyed” to lose the thirteen colonies to the new nation, but, for his government, “it could have been a lot worse.” Sandbrook and Holland discuss historical events that overshadow the American Revolution in the British mind; the 1619 Project and the subject of slavery; the “colossally consequential” Presidency of Donald Trump; and the fate of the British monarchy.Further reading :  “Was the Declaration of Independence Better Before the Edits?” by Jill Lepore.  “The American Revolution Wasn't the Main Event” by Daniel Immerwahr. “Two Hundred and Fifty Years of Complicated Commemorations" by Jelani Cobb The Political Scene draws on the reporting and analysis found in The New Yorker for lively conversations about the big questions in American politics. Join the magazine's writers and editors as they put into context the latest news—about elections, the economy, the White House, the Supreme Court, and much more. New episodes are available three times a week. Tune in to The Political Scene wherever you get your podcasts. Learn about your ad choices: dovetail.prx.org/ad-choices

Americans tend to see the Declaration of Independence and the Revolutionary War as milestones in world history that inaugurated the era of modern democracy. But the British, unsurprisingly, see these events quite differently. David Remnick talks with the historians who host the popular podcast “The Rest Is History,” Dominic Sandbrook and Tom Holland. Growing up in Britain, Sandbrook explains, the Revolution seemed like “a parade of quite boring men talking very earnestly about liberty, [with] battles that involved twenty people in a field somewhere. . . . It's not Waterloo!” The King was “annoyed” to lose the thirteen colonies to the new nation, but, for his government, “it could have been a lot worse.” Sandbrook and Holland discuss historical events that overshadow the American Revolution in the British mind; the 1619 Project and the subject of slavery; the “colossally consequential” Presidency of Donald Trump; and the fate of the British monarchy.  Further reading and listening:   “The American Revolution Wasn't the Main Event,” by Daniel Immerwahr America at 250, a special issue of The New Yorker “Was the Declaration of Independence Better Before the Edits?,” by Jill Lepore “Scandal, Protest, Goofiness, and Grandeur at the U.S. Bicentennial,” by Jill Lepore “We Could Have Been Canada,” by Adam Gopnik    New episodes of The New Yorker Radio Hour drop every Tuesday and Friday. Join host David Remnick as he discusses the latest in politics, news, and current events in conversation with political leaders, newsmakers, innovators, New Yorker staff writers, authors, actors, and musicians. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Leslie John (Revealing: The Underrated Power of Oversharing) is a behavioral scientist, Harvard Business School professor, and expert on privacy, self-disclosure, and decision-making. Leslie joins Armchair Expert to discuss growing up in Waterloo, Canada, training professionally in ballet as a child, and how her family's irrational penny-pinching sparked her fascination with human behavior. Leslie and Dax talk about why people are more open to revealing their dark secrets on a sketchy-looking website over a more official looking one, how one mortifying overshare helped her find lifelong mentors, and what parasocial relationships reveal about modern intimacy. Leslie explains why secrets take up cognitive space, how vulnerability creates trust through social risk, and why we may be better off sharing a little more than we think we should.Check Allstate first for a quote that could save you hundreds: https://www.allstate.com/See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.