Podcasts about R1

Opening Round is here, or as we like to call it, Cheat Code Round! It means we get a look at some potential AFL Fantasy players who we are considering and know what score will be in there to affect the first price change. Calvin, Roy and Warnie chat through who is on the watchlist across the five matches and answer plenty of your questions! Head to fantasy.afl.com.au or download the app to start picking your team today. Episode guide 0:00 - Key info ahead of R1 5:30 - Sydney v Carlton Watchlist 8:20 - Gold Coast v Geelong Watchlist 12:55 - GWS v Hawthorn Watchlist 15:20 - Brisbane v Western Bulldogs Watchlist 19:45 - St Kilda v Collingwood Watchlist 22:00 - Questions from social media - follow @AFLFantasy on X, @aflfantasy on Instagram and like the Official AFL Fantasy facebook page. 37:05 - Last minute tips. - - - - Find more from Roy, Calvin and Warnie. Head to afl.com.au/fantasy for more content from The Traders. Like AFL Fantasy on Facebook. Follow @AFLFantasy on Instagram. Follow @AFLFantasy on X.See omnystudio.com/listener for privacy information.

Public Works Minister Dean Macpherson is facing a Public Protector investigation over a R1 million taxpayer-funded trip to Brazil with his partner. ActionSA alleges possible misuse of funds. Elvis Presslin spoke to ActionSA MP, Malebo Kobe

Africa Melane speaks to Ann Coetzee, CEO of the Little Eden Society, about the 2026 CEO Wheelchair Challenge, launching at Eastgate Mall. Business leaders will spend a workday in a wheelchair to raise R1.5 million for lifelong care supporting 300 residents with profound intellectual disabilities. Early Breakfast with Africa Melane is 702’s and CapeTalk’s early morning talk show. Experienced broadcaster Africa Melane brings you the early morning news, sports, business, and interviews politicians and analysts to help make sense of the world. He also enjoys chatting to guests in the lifestyle sphere and the Arts. All the interviews are podcasted for you to catch-up and listen. Thank you for listening to this podcast from Early Breakfast with Africa Melane For more about the show click https://buff.ly/XHry7eQ and find all the catch-up podcasts here https://buff.ly/XJ10LBU Listen live on weekdays between 04:00 and 06:00 (SA Time) to the Early Breakfast with Africa Melane broadcast on 702 https://buff.ly/gk3y0Kj and CapeTalk https://buff.ly/NnFM3N Subscribe to the 702 and CapeTalk daily and weekly newsletters https://buff.ly/v5mfetc Follow us on social media: 702 on Facebook: https://www.facebook.com/TalkRadio702 702 on TikTok: https://www.tiktok.com/@talkradio702 702 on Instagram: https://www.instagram.com/talkradio702/ 702 on X: https://x.com/Radio702 702 on YouTube: https://www.youtube.com/@radio702 CapeTalk on Facebook: https://www.facebook.com/CapeTalk CapeTalk on TikTok: https://www.tiktok.com/@capetalk CapeTalk on Instagram: https://www.instagram.com/ CapeTalk on X: https://x.com/CapeTalk CapeTalk on YouTube: https://www.youtube.com/@CapeTalk567 See omnystudio.com/listener for privacy information.

AI is already displacing workers in targeted ways - entry-level knowledge workers are being quietly erased from hiring pipelines, freelancers are getting crushed, and the career ladder is being sawed off at the bottom rungs. Yet ML engineer demand has surged 89% with a 3.2:1 talent deficit and $187K median salary. Covers the real displacement data, lessons from the artist bloodbath, the trades escape hatch, the orchestrator treadmill, expert disagreements on timelines, and concrete short- and long-term career moves for ML engineers. Links Notes and resources at ocdevel.com/mlg/mla-4 Try a walking desk - stay healthy & sharp while you learn & code Generate a podcast - use my voice to listen to any AI generated content you want Market Metrics and Displacement Dynamics ML Market: H1 2025 demand rose 89% with a 3.2 to 1 talent deficit. Median salary is $187,500, while Generative AI specialists earn a 40 to 60 percent premium. The "Quiet" Decline: Macro data shows only 4.5% of total layoffs are AI-attributed, but entry-level hiring is collapsing. Stanford/ADP data shows a 13 to 16 percent employment drop for workers aged 22 to 25 in AI-exposed roles since late 2022. UK graduate job postings fell 67%. Corporate Attrition: Salesforce cut 4,000 roles after AI absorbed 30 to 50 percent of workloads. Microsoft cut 15,000 roles as AI began generating 30% of its code. Amazon cut 30,000 jobs while spending $100 billion on AI infrastructure. Sector Analysis: Creative and Trades Illustrators: Jobs in China's gaming sector fell 70% in one year. Clients accept "good enough" work (80% quality) at 5% of the cost. Western freelance graphic design and writing jobs fell 18.5% and 30% respectively within eight months of ChatGPT's launch. Manual Labor: The U.S. construction industry lacks 1.7 million workers annually, but apprenticeships take five years. Humanoid robotics are advancing, with Unitree's R1 priced at $5,900 and Figure AI robots completing 1,250 runtime hours at BMW. Full automation is 10 to 15 years away, but partial displacement via smaller crews is closer. The Orchestration Treadmill Obsolescence Speed: Prompt engineering roles went from $375,000 salaries to obsolescence in 24 months. AI coding agents like Claude Code now resolve 72% of medium-complexity GitHub issues autonomously. Fragile Expertise: Replacing junior workers with AI prevents the development of future senior talent. New engineers risk "fragile expertise," directed by tools they cannot debug during novel failure modes. Economic and Expert Outlook Macro Risks: Daron Acemoglu warns of "so-so automation" that cuts costs without raising productivity, predicting only 0.66% growth over ten years. "Ghost GDP" describes AI-inflated accounts that fail to circulate because machines do not consume. Expert Camps: Accelerationists (Anthropic, OpenAI) predict human-level AI by 2027. Skeptics (LeCun, Marcus) argue LLMs are a dead end lacking world models. Pragmatists (Andrew Ng) suggest shifting from implementation to specification as the cost of code nears zero. Tactical Adaptation for ML Engineers Immediate Skills: Master production ML systems, MLOps, LLM evaluation, and safety engineering. Ability to manage deployment risks and hallucination detection is the primary hiring differentiator. Long-term Moats: Focus on "Small AI" (on-device, private), mechanistic interpretability, and deep domain knowledge in healthcare, logistics, or climate science. The Playbook: Optimize for the current three to five year window. Move from being a model builder to a product-focused engineer who understands business tradeoffs and regulatory compliance.

Stephen Grootes speaks to Lebogang Luvuno, B-BBEE Executive at Microsoft South Africa, about the launch of Microsoft’s Mission Next Equity Equivalent & Investment Programme (EEIP), a R1.3 billion commitment aimed at accelerating high-impact, black-owned SMMEs into globally competitive frontier companies. The Money Show is a podcast hosted by well-known journalist and radio presenter, Stephen Grootes. He explores the latest economic trends, business developments, investment opportunities, and personal finance strategies. Each episode features engaging conversations with top newsmakers, industry experts, financial advisors, entrepreneurs, and politicians, offering you thought-provoking insights to navigate the ever-changing financial landscape.    Thank you for listening to a podcast from The Money Show Listen live Primedia+ weekdays from 18:00 and 20:00 (SA Time) to The Money Show with Stephen Grootes broadcast on 702 https://buff.ly/gk3y0Kj and CapeTalk https://buff.ly/NnFM3Nk For more from the show, go to https://buff.ly/7QpH0jY or find all the catch-up podcasts here https://buff.ly/PlhvUVe Subscribe to The Money Show Daily Newsletter and the Weekly Business Wrap here https://buff.ly/v5mfetc The Money Show is brought to you by Absa     Follow us on social media   702 on Facebook: https://www.facebook.com/TalkRadio702 702 on TikTok: https://www.tiktok.com/@talkradio702 702 on Instagram: https://www.instagram.com/talkradio702/ 702 on X: https://x.com/CapeTalk 702 on YouTube: https://www.youtube.com/@radio702   CapeTalk on Facebook: https://www.facebook.com/CapeTalk CapeTalk on TikTok: https://www.tiktok.com/@capetalk CapeTalk on Instagram: https://www.instagram.com/ CapeTalk on X: https://x.com/Radio702 CapeTalk on YouTube: https://www.youtube.com/@CapeTalk567 See omnystudio.com/listener for privacy information.

Se Stanislavem Fortem o nástupu AI agentů, limitech a rizicích umělé inteligenci, ochraně a opravování softwarových katedrál a budování kyberbezpečnostního startupu Aisle v Praze. Moderuje Štěpán Sedláček.Pravděpodobně prožíváme technologickou revoluci, jejíž rychlost, rozsah a potenciálních dopady na lidský život a práci nemají obdoby, ať už skončí jakkoli. Nástup velkých jazykových modelů a generativní AI je čím dál patrnější napříč různými sférami lidské činnosti od programování po umění. Otázky, které dříve řešila poměrné malá skupina lidí spojených s výzkumem a vývojem umělé inteligence nebo science fiction, jsou dnes často ve středu zájmu celospolečenské debaty, byť by si možná zasloužily ještě více pozornosti a to i ze strany států. Otázku po tom, jestli někdy bude k dispozici umělá inteligence, která předčí člověka, dnes spíše přebíjí otázka, jestli nás od ní dělí rok, několik let nebo víc času. Stanislav Fort je matematik, fyzik a expert na umělou inteligenci a velké jazykové modely (LLM), který dříve působil v předních světových společnostech v oboru Google DeepMind nebo Anthropic. Jak vidí letošní rok na poli AI?„Myslím, že letos si většina lidí uvědomí, že AI funguje a je schopná dělat užitečnou intelektuální práci. V roce 2025 se staly mainstreamem přemýšlecí (tzv. reasoning) modely zejména v souvislosti s nástupem modelu R1 od společnosti DeepSeek. Během té doby se modely extrémně zlepšily a začaly být schopné řešit dlouhé a obtížné intelektuální úkoly napříč obory u nichž je třeba koordinovat přemýšlení přes dlouhé časové horizonty. A ty se měsíc po měsíci prodlužovaly rapidním tempem. Dnes si už většina lidí v programování i softwarovém inženýrství a odvětvích, která silně závisejí na využití počítačů, uvědomuje, že jsme na hraně toho, kdy tyto věci dokáží pracovat na podobných věcech jako elitní lidé a nepotřebují příliš supervize. Rok 2026 bude rokem, kdy AI agenti a přemýšlecí modely, které je pohánějí, začnou fungovat v reálných ekonomicky důležitých činnostech,“ říká expert Stanislav Fort, který společně s Ondřejem Vlčkem a Jayou Baloo založil firmu Aisle, kde působí jako hlavní vědec.Podařilo se jim vytvořit autonomní AI nástroj, který umí rychle nacházet a opravovat bezpečnostní chyby ve složitých softwarových systémech jako je protokol OpenSSL, který šifruje většinu komunikace na webu. Jaké mají po roce fungování na poli kybernetické bezpečnosti cíle? Jaký zásadní problém se jim podařilo vyřešit? Co říká na nástup AI agentů dění kolem sítě Moltbook? Vidí nějaké fundamentální limity ve vývoji umělé inteligence? Co si myslí o AI bublině na trzích? Jak by se měla Evropa postavit k aktuálním závodům ve vývoji AI? A jaká úskalí má zakládání kyberbezpečnostního startupu v Praze? Nejen na to se ptá v podcastu Zeitgeist Štěpán Sedláček.

Mandy Wiener speaks to Ekurhuleni MMC for Finance and Strategy, Jongizizwe Dlabathi about the R1.6 billion debt owed by Tembisa residents that has resulted in the water and electricity cuts in the area. The Midday Report with Mandy Wiener is 702 and CapeTalk’s flagship news show, your hour of essential news radio. The show is podcasted every weekday, allowing you to catch up with a 60-minute weekday wrap of the day's main news. It's packed with fast-paced interviews with the day’s newsmakers, as well as those who can make sense of the news and explain what's happening in your world. All the interviews are podcasted for you to catch up and listen to. Thank you for listening to this podcast of The Midday Report Listen live on weekdays between 12:00 and 13:00 (SA Time) to The Midday Report broadcast on 702 https://buff.ly/gk3y0Kj and on CapeTalk https://buff.ly/NnFM3Nk For more from The Midday Report go to https://buff.ly/BTGmL9H and find all the catch-up podcasts here https://buff.ly/LcbDdFI Subscribe to the 702 and CapeTalk daily and weekly newsletters https://buff.ly/v5mfetc Follow us on social media: 702 on Facebook: https://www.facebook.com/TalkRadio702 702 on TikTok: https://www.tiktok.com/@talkradio702 702 on Instagram: https://www.instagram.com/talkradio702/ 702 on X: https://x.com/Radio702 702 on YouTube: https://www.youtube.com/@radio702 CapeTalk on Facebook: https://www.facebook.com/CapeTalk CapeTalk on TikTok: https://www.tiktok.com/@capetalk CapeTalk on Instagram: https://www.instagram.com/ CapeTalk on X: https://x.com/CapeTalk CapeTalk on YouTube: https://www.youtube.com/@CapeTalk567See omnystudio.com/listener for privacy information.

In this episode of the Crazy Wisdom Podcast, host Stewart Alsop sits down with Jake Hamilton, founder of Groundwire and Nockbox, to explore zero-knowledge proofs, Bitcoin identity systems, and the intersection of privacy-preserving cryptography with AI and blockchain technology. They discuss how ZK proofs could offer an alternative to invasive identity verification systems being rolled out by governments worldwide, the potential for continual learning AI models to shift the balance between centralized and open-source development, and why building secure, auditable computing infrastructure on platforms like Urbit matters more than ever as we face an explosion of AI agents and automated systems. Jake also explains Nockchain's approach to creating a global repository of cryptographically verified facts that can power trustless programmable systems, and how these technologies might converge to solve problems around supply chain security, personal data sovereignty, and resistance to censorship.Timestamps00:00 Introduction to Groundwire and Knockbox02:48 Understanding Zero-Knowledge Proofs06:04 Government Adoption of ZK Proofs08:55 The Future of Identity Verification11:52 AI and ZK Proofs: A New Era14:54 The Role of Urbit in Technology18:03 The Impact of COVID on Trust20:51 The Evolution of AI and Data Privacy23:47 The Future of AI Models26:54 The Need for Local AI Solutions29:51 Interoperability of Knockchain and BitcoinKey Insights1. Zero-Knowledge Proofs Enable Privacy-Preserving Verification: Jake explains that ZK proofs allow you to prove computational outcomes without revealing the underlying data. For example, you could prove you're over 18 without exposing your full identity or driver's license information. The proof demonstrates that a specific program ran through certain steps and reached a particular conclusion, and validating this proof is fast and compact. This technology has profound implications for age verification, identity systems, and protecting privacy while maintaining necessary compliance, potentially offering a middle path between surveillance states and complete anonymity.2. Government Adoption of Privacy Technology Remains Uncertain: There are three competing motivations driving government identity verification systems: genuine surveillance desires, bureaucratic efficiency seeking, and legitimate child protection concerns. Jake believes these groups can be separated, with some officials potentially supporting ZK-based solutions if positioned correctly. He notes the EU is exploring ZK identity verification, and UK officials have shown interest. The key is framing privacy-preserving technology as protection against "the swamp" rather than just abstract privacy benefits, which could resonate with certain political constituencies.3. The COVID Era Destroyed Institutional Trust at Unprecedented Scale: The conversation identifies COVID as potentially the largest institutional trust-burning event in human history, with numerous institutions simultaneously losing credibility with large portions of the population. This represents a dramatic shift from the boomer generation's default trust in authority figures and mainstream media. This collapse is compounded by the incoming AI revolution, creating a perfect storm where established bureaucracies cannot adapt quickly enough to manage rapidly evolving technology, leaving society in fundamentally unmanageable territory.4. Centralized AI Models Create Dangerous Dependencies: Both speakers acknowledge growing dependence on centralized AI services like Claude, with some users spending thousands monthly on tokens. This dependency creates vulnerability to price increases and service disruptions. Jake advocates for local AI deployment using models like DeepSeek R1, running on personal hardware to maintain control and privacy. The shift toward continuous learning models will fundamentally change the AI landscape, making personal data harvesting even more valuable and raising urgent questions about compensation and consent for training data contribution.5. High-Quality Training Data Is Becoming the Primary AI Bottleneck: Stewart argues that AI development is now limited more by high-quality training data than by compute power. The industry has exhausted easily accessible internet data and body-shop-style data labeling. Companies are now using specialized boutique services with techniques like head-mounted cameras for live-streaming world model training. This scarcity is subtly driving price increases across AI services and will fundamentally reshape the economics of AI development, with implications for who controls these increasingly powerful systems.6. Urbit Offers a Foundation for Trustworthy Computing: Jake positions Urbit as essential infrastructure for the AI age because its 30,000-line codebase (versus Unix's three million lines) can be understood by individual humans. Its deterministic, purely functional, and strictly typed design aims for eventual ossification—software that doesn't require constant security patches. This "tiny and diamond perfect" approach addresses the fundamental insecurity of systems requiring monthly vulnerability patches. In an era of AI agents and potential prompt injection attacks, having verifiable, comprehensible computing infrastructure becomes existentially important rather than merely desirable.7. Nockchain Creates a Global Repository of Provable Truth: Jake's vision for Nockchain combines ZK proofs with blockchain technology to create a globally available "truth repository" where verified facts can be programmatically accessed together. This enables smart contracts or programs gated on combinations of proven facts—such as temperature readings from secure devices, supply chain events, and payment confirmations. By using Nock's abstract, simple design optimized for ZK proof generation, the system can validate complex real-world conditions without exposing underlying data, creating infrastructure for coordinating action based on verifiable private information at global scale.

https://www.tadeclinicagem.com.br/materiais/evento-org-dzero-na-residencia-2026/

Guest: Wayne Duvenage | CEO at Organisation Undoing Tax Abuse (OUTA) OUTA CEO Wayne Duvenage joins Africa Melane to unpack concerns over the RTIA’s proposed R1.2bn private tender to roll out Aarto. Early Breakfast with Africa Melane is 702’s and CapeTalk’s early morning talk show. Experienced broadcaster Africa Melane brings you the early morning news, sports, business, and interviews politicians and analysts to help make sense of the world. He also enjoys chatting to guests in the lifestyle sphere and the Arts. All the interviews are podcasted for you to catch-up and listen.Thank you for listening to this podcast from Early Breakfast with Africa Melane For more about the show click https://buff.ly/XHry7eQ and find all the catch-up podcasts here https://buff.ly/XJ10LBUListen live on weekdays between 04:00 and 06:00 (SA Time) to the Early Breakfast with Africa Melane broadcast on 702 https://buff.ly/gk3y0Kj and CapeTalk https://buff.ly/NnFM3NSubscribe to the 702 and CapeTalk daily and weekly newsletters https://buff.ly/v5mfetcFollow us on social media:702 on Facebook: https://www.facebook.com/TalkRadio702702 on TikTok: https://www.tiktok.com/@talkradio702702 on Instagram: https://www.instagram.com/talkradio702/702 on X: https://x.com/Radio702702 on YouTube: https://www.youtube.com/@radio702 CapeTalk on Facebook: https://www.facebook.com/CapeTalkCapeTalk on TikTok: https://www.tiktok.com/@capetalkCapeTalk on Instagram: https://www.instagram.com/CapeTalk on X: https://x.com/CapeTalkCapeTalk on YouTube: https://www.youtube.com/@CapeTalk567See omnystudio.com/listener for privacy information.

Bonjour à toutes et tous,Le podcast est de retour avec un nouvel épisode… et pas des moindres !

De slimme bril van Even Realities maakte vorig jaar indruk in De Schaal van Hebben. De G1 was opvallend omdat hij géén camera had, maar wel een schermpje in het glas met groene tekst voor navigatie, notificaties en AI. Een jaar later ligt de opvolger op tafel: de G2. En nieuw in het ecosysteem is de R1, een slimme ring die de bediening moet aanvullen. De basis is hetzelfde gebleven. Ook de G2 projecteert alleen groene tekst in het glas. Wat wel is verbeterd, is het display zelf. Het schermoppervlak is groter, de helderheid hoger en de tekst scherper. In binnenruimtes is het goed leesbaar, buiten in fel zonlicht blijft het contrast beperkt. De bril is dunner en lichter dan zijn voorganger en voelt minder experimenteel aan. Ook de software reageert sneller. Transcriptie en realtime vertalingen verschijnen vrijwel direct in beeld. Nieuw is de functie ‘Conversate’, waarbij de bril gesprekken live ondertitelt en via AI aanvullende informatie suggereert. Dat werkt het best in rustige omgevingen. In rumoerige settings neemt de nauwkeurigheid af, zoals bij veel spraakherkenning het geval is. De R1-ring fungeert als afstandsbediening. Via een aanraakgevoelig oppervlak kun je scrollen en tikken om door menu’s te navigeren. Dat werkt, maar is soms gevoelig of onnauwkeurig. De ring meet daarnaast hartslag, slaap, stappen en zuurstofsaturatie. Opvallend is dat deze data alleen binnen de Even Realities-app blijft; koppeling met bijvoorbeeld Apple Health ontbreekt. De stapmeting wijkt in de praktijk af van een smartwatch, de slaapmeting lijkt consistenter. Er zijn ook kanttekeningen. Het display zit iets hoger in het glas, waardoor je de bril soms verder op je neus moet schuiven om goed te kunnen lezen. En waar concurrenten inzetten op audio of bredere app-integratie, blijft deze bril beperkt tot tekst. De G2 is duidelijk volwassener dan de eerste generatie en technisch een van de betere slimme brillen zonder camera. Maar de lat ligt inmiddels hoger. Het is geen onmisbare gadget, wel een product dat laat zien waar deze categorie naartoe kan. Luister de podcast voor het eindoordeel. See omnystudio.com/listener for privacy information.

https://www.tadeclinicagem.com.br/materiais/evento-org-dzero-na-residencia-2026/O começo do R1 de Clínica Médica pode ser desafiador. Mas você não precisa passar por isso sozinho(a).Descubra como tornar o R1 mais leve e produtivo com o D0 na Residência, evento on-line e gratuito do TdC.Garanta agora mesmo sua vaga e receba grátis no seu e-mail o Ebook: Sala de Emergência - Temas Essenciais.Do manejo de antibióticos à organização da rotina com IA, confira os temas que abordaremos ao vivo Sábado 28/02 no youtube.João Urbano, nosso Joca, Raphael Coelho e José Marcos se reunem para falar sobre Transtorno do Déficit de Atenção e Hiperatividade (TDAH) no adulto, focando na abordagem no adulto.Referências em breve.

In South Africa, high earners face a 45% marginal rate with limited deductions, making Section 12B of the Income Tax Act a powerful tool to ease the tax burden while investing in renewable energy. In an interview with Biznews, Mitchel Fieldgate, wealth manager and alternative investment lead at Pangea Wealth, revealed they have facilitated R1 billion in deductions over the last two years. He noted that while this provides relief for high earners, it has also mobilised billions into commercial solar projects, reducing reliance on Eskom and alleviated load shedding. Fieldgate explained who might not benefit from the tax relief, clarified that it is not a tax loophole, and stressed that the 12B opportunities are not a one-size-fits-all solution. He suggested the government introduce similar tax breaks for repairing other infrastructure, such as water systems or roads, “because it mobilises money and there's no leakage.”

Dr. Christie Langenberg reviews the No. 6 article of 2024, titled “Clinical Efficacy of Multiple Intra-Articular Injection for Hip Osteoarthritis,” which was originally published in The Bone and Joint Journal in June 2024. Dr. Jeremy Schroeder serves as the series host. Dr. Langenberg is a member of the Top Articles Subcommittee, and this episode is part of an ongoing mini journal club series highlighting each of the Top Articles in Sports Medicine from 2024, as selected for the 2025 AMSSM Annual Meeting. Clinical Efficacy of Multiple Intra-Articular Injection for Hip Osteoarthritis: https://boneandjoint.org.uk/Article/10.1302/0301-620X.106B6.BJJ-2023-1272.R1

In today's episode of BizNews Daybreak, President Cyril Ramaphosa delivers a "game of two halves" State of the Nation Address—promising a R1 trillion infrastructure drive while admitting water has replaced electricity as South Africa's newest crisis. We dissect his branding of mining as a "sunrise industry" that just continues sliding into darkness despite the country sitting on R40 trillion in mineral reserves. Plus: Political Earthquake: The Patriotic Alliance surges in George, snatching a key stronghold by giving the Democratic Alliance another bruising, taking its head-to-head score to four from four in the by-elections. School Scandal: An emotional David Shapiro weighs in on Roedean School's refusal to play tennis against its Jewish counterpart, King David, sparking fierce debate about antisemitism and politics in sport. Market Moves: Gold breaks $5,000, the Rand strengthens, and Sasol jumps 7%.

Welcome to a Wide World of Sports update. A snapshot of the latest sport stories from the 9News team including: Maori coach Blair wants All Stars game to remain in pre-season WA and Victoria train ahead of State of Origin clash Leishman tied for lead after R1 at LIV Adelaide The biggest sport stories in less than 5 minutes delivered twice a day, with reports from the 9News team across Australia and overseas. Subscribe now to make it part of your daily news diet. See omnystudio.com/listener for privacy information.

Analitzem el futur de la l

This episode recorded live at the Becker's 13th Annual CEO + CFO Roundtable features Peter D. Banko,  President & CEO of Baystate Health. Here, he discusses how Baystate Health is strengthening its commitment to people-centric care while navigating the realities of cost management and the evolving landscape of AI governance. He shares insights into building resilient, future-ready health systems that balance innovation with operational discipline.In collaboration with R1.

This episode recorded live at the Becker's 13th Annual CEO + CFO Roundtable features John Mallia, Interim Chief Financial Officer at Methodist Le Bonheur Healthcare. Here, he discusses how Methodist Le Bonheur Healthcare is navigating revenue cycle and payer dynamics while adapting financial strategies to meet evolving challenges. He shares insights into managing patient expectations and addressing concerns around data usage to create a more transparent and effective healthcare experience.In collaboration with R1.

From Palantir and Two Sigma to building Goodfire into the poster-child for actionable mechanistic interpretability, Mark Bissell (Member of Technical Staff) and Myra Deng (Head of Product) are trying to turn “peeking inside the model” into a repeatable production workflow by shipping APIs, landing real enterprise deployments, and now scaling the bet with a recent $150M Series B funding round at a $1.25B valuation.In this episode, we go far beyond the usual “SAEs are cool” take. We talk about Goodfire's core bet: that the AI lifecycle is still fundamentally broken because the only reliable control we have is data and we post-train, RLHF, and fine-tune by “slurping supervision through a straw,” hoping the model picks up the right behaviors while quietly absorbing the wrong ones. Goodfire's answer is to build a bi-directional interface between humans and models: read what's happening inside, edit it surgically, and eventually use interpretability during training so customization isn't just brute-force guesswork.Mark and Myra walk through what that looks like when you stop treating interpretability like a lab demo and start treating it like infrastructure: lightweight probes that add near-zero latency, token-level safety filters that can run at inference time, and interpretability workflows that survive messy constraints (multilingual inputs, synthetic→real transfer, regulated domains, no access to sensitive data). We also get a live window into what “frontier-scale interp” means operationally (i.e. steering a trillion-parameter model in real time by targeting internal features) plus why the same tooling generalizes cleanly from language models to genomics, medical imaging, and “pixel-space” world models.We discuss:* Myra + Mark's path: Palantir (health systems, forward-deployed engineering) → Goodfire early team; Two Sigma → Head of Product, translating frontier interpretability research into a platform and real-world deployments* What “interpretability” actually means in practice: not just post-hoc poking, but a broader “science of deep learning” approach across the full AI lifecycle (data curation → post-training → internal representations → model design)* Why post-training is the first big wedge: “surgical edits” for unintended behaviors likereward hacking, sycophancy, noise learned during customization plus the dream of targeted unlearning and bias removal without wrecking capabilities* SAEs vs probes in the real world: why SAE feature spaces sometimes underperform classifiers trained on raw activations for downstream detection tasks (hallucination, harmful intent, PII), and what that implies about “clean concept spaces”* Rakuten in production: deploying interpretability-based token-level PII detection at inference time to prevent routing private data to downstream providers plus the gnarly constraints: no training on real customer PII, synthetic→real transfer, English + Japanese, and tokenization quirks* Why interp can be operationally cheaper than LLM-judge guardrails: probes are lightweight, low-latency, and don't require hosting a second large model in the loop* Real-time steering at frontier scale: a demo of steering Kimi K2 (~1T params) live and finding features via SAE pipelines, auto-labeling via LLMs, and toggling a “Gen-Z slang” feature across multiple layers without breaking tool use* Hallucinations as an internal signal: the case that models have latent uncertainty / “user-pleasing” circuitry you can detect and potentially mitigate more directly than black-box methods* Steering vs prompting: the emerging view that activation steering and in-context learning are more closely connected than people think, including work mapping between the two (even for jailbreak-style behaviors)* Interpretability for science: using the same tooling across domains (genomics, medical imaging, materials) to debug spurious correlations and extract new knowledge up to and including early biomarker discovery work with major partners* World models + “pixel-space” interpretability: why vision/video models make concepts easier to see, how that accelerates the feedback loop, and why robotics/world-model partners are especially interesting design partners* The north star: moving from “data in, weights out” to intentional model design where experts can impart goals and constraints directly, not just via reward signals and brute-force post-training—Goodfire AI* Website: https://goodfire.ai* LinkedIn: https://www.linkedin.com/company/goodfire-ai/* X: https://x.com/GoodfireAIMyra Deng* Website: https://myradeng.com/* LinkedIn: https://www.linkedin.com/in/myra-deng/* X: https://x.com/myra_dengMark Bissell* LinkedIn: https://www.linkedin.com/in/mark-bissell/* X: https://x.com/MarkMBissellFull Video EpisodeTimestamps00:00:00 Introduction00:00:05 Introduction to the Latent Space Podcast and Guests from Goodfire00:00:29 What is Goodfire? Mission and Focus on Interpretability00:01:01 Goodfire's Practical Approach to Interpretability00:01:37 Goodfire's Series B Fundraise Announcement00:02:04 Backgrounds of Mark and Myra from Goodfire00:02:51 Team Structure and Roles at Goodfire00:05:13 What is Interpretability? Definitions and Techniques00:05:30 Understanding Errors00:07:29 Post-training vs. Pre-training Interpretability Applications00:08:51 Using Interpretability to Remove Unwanted Behaviors00:10:09 Grokking, Double Descent, and Generalization in Models00:10:15 404 Not Found Explained00:12:06 Subliminal Learning and Hidden Biases in Models00:14:07 How Goodfire Chooses Research Directions and Projects00:15:00 Troubleshooting Errors00:16:04 Limitations of SAEs and Probes in Interpretability00:18:14 Rakuten Case Study: Production Deployment of Interpretability00:20:45 Conclusion00:21:12 Efficiency Benefits of Interpretability Techniques00:21:26 Live Demo: Real-Time Steering in a Trillion Parameter Model00:25:15 How Steering Features are Identified and Labeled00:26:51 Detecting and Mitigating Hallucinations Using Interpretability00:31:20 Equivalence of Activation Steering and Prompting00:34:06 Comparing Steering with Fine-Tuning and LoRA Techniques00:36:04 Model Design and the Future of Intentional AI Development00:38:09 Getting Started in Mechinterp: Resources, Programs, and Open Problems00:40:51 Industry Applications and the Rise of Mechinterp in Practice00:41:39 Interpretability for Code Models and Real-World Usage00:43:07 Making Steering Useful for More Than Stylistic Edits00:46:17 Applying Interpretability to Healthcare and Scientific Discovery00:49:15 Why Interpretability is Crucial in High-Stakes Domains like Healthcare00:52:03 Call for Design Partners Across Domains00:54:18 Interest in World Models and Visual Interpretability00:57:22 Sci-Fi Inspiration: Ted Chiang and Interpretability01:00:14 Interpretability, Safety, and Alignment Perspectives01:04:27 Weak-to-Strong Generalization and Future Alignment Challenges01:05:38 Final Thoughts and Hiring/Collaboration Opportunities at GoodfireTranscriptShawn Wang [00:00:05]: So welcome to the Latent Space pod. We're back in the studio with our special MechInterp co-host, Vibhu. Welcome. Mochi, Mochi's special co-host. And Mochi, the mechanistic interpretability doggo. We have with us Mark and Myra from Goodfire. Welcome. Thanks for having us on. Maybe we can sort of introduce Goodfire and then introduce you guys. How do you introduce Goodfire today?Myra Deng [00:00:29]: Yeah, it's a great question. So Goodfire, we like to say, is an AI research lab that focuses on using interpretability to understand, learn from, and design AI models. And we really believe that interpretability will unlock the new generation, next frontier of safe and powerful AI models. That's our description right now, and I'm excited to dive more into the work we're doing to make that happen.Shawn Wang [00:00:55]: Yeah. And there's always like the official description. Is there an understatement? Is there an unofficial one that sort of resonates more with a different audience?Mark Bissell [00:01:01]: Well, being an AI research lab that's focused on interpretability, there's obviously a lot of people have a lot that they think about when they think of interpretability. And I think we have a pretty broad definition of what that means and the types of places that can be applied. And in particular, applying it in production scenarios, in high stakes industries, and really taking it sort of from the research world into the real world. Which, you know. It's a new field, so that hasn't been done all that much. And we're excited about actually seeing that sort of put into practice.Shawn Wang [00:01:37]: Yeah, I would say it wasn't too long ago that Anthopic was like still putting out like toy models or superposition and that kind of stuff. And I wouldn't have pegged it to be this far along. When you and I talked at NeurIPS, you were talking a little bit about your production use cases and your customers. And then not to bury the lead, today we're also announcing the fundraise, your Series B. $150 million. $150 million at a 1.25B valuation. Congrats, Unicorn.Mark Bissell [00:02:02]: Thank you. Yeah, no, things move fast.Shawn Wang [00:02:04]: We were talking to you in December and already some big updates since then. Let's dive, I guess, into a bit of your backgrounds as well. Mark, you were at Palantir working on health stuff, which is really interesting because the Goodfire has some interesting like health use cases. I don't know how related they are in practice.Mark Bissell [00:02:22]: Yeah, not super related, but I don't know. It was helpful context to know what it's like. Just to work. Just to work with health systems and generally in that domain. Yeah.Shawn Wang [00:02:32]: And Mara, you were at Two Sigma, which actually I was also at Two Sigma back in the day. Wow, nice.Myra Deng [00:02:37]: Did we overlap at all?Shawn Wang [00:02:38]: No, this is when I was briefly a software engineer before I became a sort of developer relations person. And now you're head of product. What are your sort of respective roles, just to introduce people to like what all gets done in Goodfire?Mark Bissell [00:02:51]: Yeah, prior to Goodfire, I was at Palantir for about three years as a forward deployed engineer, now a hot term. Wasn't always that way. And as a technical lead on the health care team and at Goodfire, I'm a member of the technical staff. And honestly, that I think is about as specific as like as as I could describe myself because I've worked on a range of things. And, you know, it's it's a fun time to be at a team that's still reasonably small. I think when I joined one of the first like ten employees, now we're above 40, but still, it looks like there's always a mix of research and engineering and product and all of the above. That needs to get done. And I think everyone across the team is, you know, pretty, pretty switch hitter in the roles they do. So I think you've seen some of the stuff that I worked on related to image models, which was sort of like a research demo. More recently, I've been working on our scientific discovery team with some of our life sciences partners, but then also building out our core platform for more of like flexing some of the kind of MLE and developer skills as well.Shawn Wang [00:03:53]: Very generalist. And you also had like a very like a founding engineer type role.Myra Deng [00:03:58]: Yeah, yeah.Shawn Wang [00:03:59]: So I also started as I still am a member of technical staff, did a wide range of things from the very beginning, including like finding our office space and all of this, which is we both we both visited when you had that open house thing. It was really nice.Myra Deng [00:04:13]: Thank you. Thank you. Yeah. Plug to come visit our office.Shawn Wang [00:04:15]: It looked like it was like 200 people. It has room for 200 people. But you guys are like 10.Myra Deng [00:04:22]: For a while, it was very empty. But yeah, like like Mark, I spend. A lot of my time as as head of product, I think product is a bit of a weird role these days, but a lot of it is thinking about how do we take our frontier research and really apply it to the most important real world problems and how does that then translate into a platform that's repeatable or a product and working across, you know, the engineering and research teams to make that happen and also communicating to the world? Like, what is interpretability? What is it used for? What is it good for? Why is it so important? All of these things are part of my day-to-day as well.Shawn Wang [00:05:01]: I love like what is things because that's a very crisp like starting point for people like coming to a field. They all do a fun thing. Vibhu, why don't you want to try tackling what is interpretability and then they can correct us.Vibhu Sapra [00:05:13]: Okay, great. So I think like one, just to kick off, it's a very interesting role to be head of product, right? Because you guys, at least as a lab, you're more of an applied interp lab, right? Which is pretty different than just normal interp, like a lot of background research. But yeah. You guys actually ship an API to try these things. You have Ember, you have products around it, which not many do. Okay. What is interp? So basically you're trying to have an understanding of what's going on in model, like in the model, in the internal. So different approaches to do that. You can do probing, SAEs, transcoders, all this stuff. But basically you have an, you have a hypothesis. You have something that you want to learn about what's happening in a model internals. And then you're trying to solve that from there. You can do stuff like you can, you know, you can do activation mapping. You can try to do steering. There's a lot of stuff that you can do, but the key question is, you know, from input to output, we want to have a better understanding of what's happening and, you know, how can we, how can we adjust what's happening on the model internals? How'd I do?Mark Bissell [00:06:12]: That was really good. I think that was great. I think it's also a, it's kind of a minefield of a, if you ask 50 people who quote unquote work in interp, like what is interpretability, you'll probably get 50 different answers. And. Yeah. To some extent also like where, where good fire sits in the space. I think that we're an AI research company above all else. And interpretability is a, is a set of methods that we think are really useful and worth kind of specializing in, in order to accomplish the goals we want to accomplish. But I think we also sort of see some of the goals as even more broader as, as almost like the science of deep learning and just taking a not black box approach to kind of any part of the like AI development life cycle, whether that. That means using interp for like data curation while you're training your model or for understanding what happened during post-training or for the, you know, understanding activations and sort of internal representations, what is in there semantically. And then a lot of sort of exciting updates that were, you know, are sort of also part of the, the fundraise around bringing interpretability to training, which I don't think has been done all that much before. A lot of this stuff is sort of post-talk poking at models as opposed to. To actually using this to intentionally design them.Shawn Wang [00:07:29]: Is this post-training or pre-training or is that not a useful.Myra Deng [00:07:33]: Currently focused on post-training, but there's no reason the techniques wouldn't also work in pre-training.Shawn Wang [00:07:38]: Yeah. It seems like it would be more active, applicable post-training because basically I'm thinking like rollouts or like, you know, having different variations of a model that you can tweak with the, with your steering. Yeah.Myra Deng [00:07:50]: And I think in a lot of the news that you've seen in, in, on like Twitter or whatever, you've seen a lot of unintended. Side effects come out of post-training processes, you know, overly sycophantic models or models that exhibit strange reward hacking behavior. I think these are like extreme examples. There's also, you know, very, uh, mundane, more mundane, like enterprise use cases where, you know, they try to customize or post-train a model to do something and it learns some noise or it doesn't appropriately learn the target task. And a big question that we've always had is like, how do you use your understanding of what the model knows and what it's doing to actually guide the learning process?Shawn Wang [00:08:26]: Yeah, I mean, uh, you know, just to anchor this for people, uh, one of the biggest controversies of last year was 4.0 GlazeGate. I've never heard of GlazeGate. I didn't know that was what it was called. The other one, they called it that on the blog post and I was like, well, how did OpenAI call it? Like officially use that term. And I'm like, that's funny, but like, yeah, I guess it's the pitch that if they had worked a good fire, they wouldn't have avoided it. Like, you know what I'm saying?Myra Deng [00:08:51]: I think so. Yeah. Yeah.Mark Bissell [00:08:53]: I think that's certainly one of the use cases. I think. Yeah. Yeah. I think the reason why post-training is a place where this makes a lot of sense is a lot of what we're talking about is surgical edits. You know, you want to be able to have expert feedback, very surgically change how your model is doing, whether that is, you know, removing a certain behavior that it has. So, you know, one of the things that we've been looking at or is, is another like common area where you would want to make a somewhat surgical edit is some of the models that have say political bias. Like you look at Quen or, um, R1 and they have sort of like this CCP bias.Shawn Wang [00:09:27]: Is there a CCP vector?Mark Bissell [00:09:29]: Well, there's, there are certainly internal, yeah. Parts of the representation space where you can sort of see where that lives. Yeah. Um, and you want to kind of, you know, extract that piece out.Shawn Wang [00:09:40]: Well, I always say, you know, whenever you find a vector, a fun exercise is just like, make it very negative to see what the opposite of CCP is.Mark Bissell [00:09:47]: The super America, bald eagles flying everywhere. But yeah. So in general, like lots of post-training tasks where you'd want to be able to, to do that. Whether it's unlearning a certain behavior or, you know, some of the other kind of cases where this comes up is, are you familiar with like the, the grokking behavior? I mean, I know the machine learning term of grokking.Shawn Wang [00:10:09]: Yeah.Mark Bissell [00:10:09]: Sort of this like double descent idea of, of having a model that is able to learn a generalizing, a generalizing solution, as opposed to even if memorization of some task would suffice, you want it to learn the more general way of doing a thing. And so, you know, another. A way that you can think about having surgical access to a model's internals would be learn from this data, but learn in the right way. If there are many possible, you know, ways to, to do that. Can make interp solve the double descent problem?Shawn Wang [00:10:41]: Depends, I guess, on how you. Okay. So I, I, I viewed that double descent as a problem because then you're like, well, if the loss curves level out, then you're done, but maybe you're not done. Right. Right. But like, if you actually can interpret what is a generalizing or what you're doing. What is, what is still changing, even though the loss is not changing, then maybe you, you can actually not view it as a double descent problem. And actually you're just sort of translating the space in which you view loss and like, and then you have a smooth curve. Yeah.Mark Bissell [00:11:11]: I think that's certainly like the domain of, of problems that we're, that we're looking to get.Shawn Wang [00:11:15]: Yeah. To me, like double descent is like the biggest thing to like ML research where like, if you believe in scaling, then you don't need, you need to know where to scale. And. But if you believe in double descent, then you don't, you don't believe in anything where like anything levels off, like.Vibhu Sapra [00:11:30]: I mean, also tendentially there's like, okay, when you talk about the China vector, right. There's the subliminal learning work. It was from the anthropic fellows program where basically you can have hidden biases in a model. And as you distill down or, you know, as you train on distilled data, those biases always show up, even if like you explicitly try to not train on them. So, you know, it's just like another use case of. Okay. If we can interpret what's happening in post-training, you know, can we clear some of this? Can we even determine what's there? Because yeah, it's just like some worrying research that's out there that shows, you know, we really don't know what's going on.Mark Bissell [00:12:06]: That is. Yeah. I think that's the biggest sentiment that we're sort of hoping to tackle. Nobody knows what's going on. Right. Like subliminal learning is just an insane concept when you think about it. Right. Train a model on not even the logits, literally the output text of a bunch of random numbers. And now your model loves owls. And you see behaviors like that, that are just, they defy, they defy intuition. And, and there are mathematical explanations that you can get into, but. I mean.Shawn Wang [00:12:34]: It feels so early days. Objectively, there are a sequence of numbers that are more owl-like than others. There, there should be.Mark Bissell [00:12:40]: According to, according to certain models. Right. It's interesting. I think it only applies to models that were initialized from the same starting Z. Usually, yes.Shawn Wang [00:12:49]: But I mean, I think that's a, that's a cheat code because there's not enough compute. But like if you believe in like platonic representation, like probably it will transfer across different models as well. Oh, you think so?Mark Bissell [00:13:00]: I think of it more as a statistical artifact of models initialized from the same seed sort of. There's something that is like path dependent from that seed that might cause certain overlaps in the latent space and then sort of doing this distillation. Yeah. Like it pushes it towards having certain other tendencies.Vibhu Sapra [00:13:24]: Got it. I think there's like a bunch of these open-ended questions, right? Like you can't train in new stuff during the RL phase, right? RL only reorganizes weights and you can only do stuff that's somewhat there in your base model. You're not learning new stuff. You're just reordering chains and stuff. But okay. My broader question is when you guys work at an interp lab, how do you decide what to work on and what's kind of the thought process? Right. Because we can ramble for hours. Okay. I want to know this. I want to know that. But like, how do you concretely like, you know, what's the workflow? Okay. There's like approaches towards solving a problem, right? I can try prompting. I can look at chain of thought. I can train probes, SAEs. But how do you determine, you know, like, okay, is this going anywhere? Like, do we have set stuff? Just, you know, if you can help me with all that. Yeah.Myra Deng [00:14:07]: It's a really good question. I feel like we've always at the very beginning of the company thought about like, let's go and try to learn what isn't working in machine learning today. Whether that's talking to customers or talking to researchers at other labs, trying to understand both where the frontier is going and where things are really not falling apart today. And then developing a perspective on how we can push the frontier using interpretability methods. And so, you know, even our chief scientist, Tom, spends a lot of time talking to customers and trying to understand what real world problems are and then taking that back and trying to apply the current state of the art to those problems and then seeing where they fall down basically. And then using those failures or those shortcomings to understand what hills to climb when it comes to interpretability research. So like on the fundamental side, for instance, when we have done some work applying SAEs and probes, we've encountered, you know, some shortcomings in SAEs that we found a little bit surprising. And so have gone back to the drawing board and done work on that. And then, you know, we've done some work on better foundational interpreter models. And a lot of our team's research is focused on what is the next evolution beyond SAEs, for instance. And then when it comes to like control and design of models, you know, we tried steering with our first API and realized that it still fell short of black box techniques like prompting or fine tuning. And so went back to the drawing board and we're like, how do we make that not the case and how do we improve it beyond that? And one of our researchers, Ekdeep, who just joined is actually Ekdeep and Atticus are like steering experts and have spent a lot of time trying to figure out like, what is the research that enables us to actually do this in a much more powerful, robust way? So yeah, the answer is like, look at real world problems, try to translate that into a research agenda and then like hill climb on both of those at the same time.Shawn Wang [00:16:04]: Yeah. Mark has the steering CLI demo queued up, which we're going to go into in a sec. But I always want to double click on when you drop hints, like we found some problems with SAEs. Okay. What are they? You know, and then we can go into the demo. Yeah.Myra Deng [00:16:19]: I mean, I'm curious if you have more thoughts here as well, because you've done it in the healthcare domain. But I think like, for instance, when we do things like trying to detect behaviors within models that are harmful or like behaviors that a user might not want to have in their model. So hallucinations, for instance, harmful intent, PII, all of these things. We first tried using SAE probes for a lot of these tasks. So taking the feature activation space from SAEs and then training classifiers on top of that, and then seeing how well we can detect the properties that we might want to detect in model behavior. And we've seen in many cases that probes just trained on raw activations seem to perform better than SAE probes, which is a bit surprising if you think that SAEs are actually also capturing the concepts that you would want to capture cleanly and more surgically. And so that is an interesting observation. I don't think that is like, I'm not down on SAEs at all. I think there are many, many things they're useful for, but we have definitely run into cases where I think the concept space described by SAEs is not as clean and accurate as we would expect it to be for actual like real world downstream performance metrics.Mark Bissell [00:17:34]: Fair enough. Yeah. It's the blessing and the curse of unsupervised methods where you get to peek into the AI's mind. But sometimes you wish that you saw other things when you walked inside there. Although in the PII instance, I think weren't an SAE based approach actually did prove to be the most generalizable?Myra Deng [00:17:53]: It did work well in the case that we published with Rakuten. And I think a lot of the reasons it worked well was because we had a noisier data set. And so actually the blessing of unsupervised learning is that we actually got to get more meaningful, generalizable signal from SAEs when the data was noisy. But in other cases where we've had like good data sets, it hasn't been the case.Shawn Wang [00:18:14]: And just because you named Rakuten and I don't know if we'll get it another chance, like what is the overall, like what is Rakuten's usage or production usage? Yeah.Myra Deng [00:18:25]: So they are using us to essentially guardrail and inference time monitor their language model usage and their agent usage to detect things like PII so that they don't route private user information.Myra Deng [00:18:41]: And so that's, you know, going through all of their user queries every day. And that's something that we deployed with them a few months ago. And now we are actually exploring very early partnerships, not just with Rakuten, but with other people around how we can help with potentially training and customization use cases as well. Yeah.Shawn Wang [00:19:03]: And for those who don't know, like it's Rakuten is like, I think number one or number two e-commerce store in Japan. Yes. Yeah.Mark Bissell [00:19:10]: And I think that use case actually highlights a lot of like what it looks like to deploy things in practice that you don't always think about when you're doing sort of research tasks. So when you think about some of the stuff that came up there that's more complex than your idealized version of a problem, they were encountering things like synthetic to real transfer of methods. So they couldn't train probes, classifiers, things like that on actual customer data of PII. So what they had to do is use synthetic data sets. And then hope that that transfer is out of domain to real data sets. And so we can evaluate performance on the real data sets, but not train on customer PII. So that right off the bat is like a big challenge. You have multilingual requirements. So this needed to work for both English and Japanese text. Japanese text has all sorts of quirks, including tokenization behaviors that caused lots of bugs that caused us to be pulling our hair out. And then also a lot of tasks you'll see. You might make simplifying assumptions if you're sort of treating it as like the easiest version of the problem to just sort of get like general results where maybe you say you're classifying a sentence to say, does this contain PII? But the need that Rakuten had was token level classification so that you could precisely scrub out the PII. So as we learned more about the problem, you're sort of speaking about what that looks like in practice. Yeah. A lot of assumptions end up breaking. And that was just one instance where you. A problem that seems simple right off the bat ends up being more complex as you keep diving into it.Vibhu Sapra [00:20:41]: Excellent. One of the things that's also interesting with Interp is a lot of these methods are very efficient, right? So where you're just looking at a model's internals itself compared to a separate like guardrail, LLM as a judge, a separate model. One, you have to host it. Two, there's like a whole latency. So if you use like a big model, you have a second call. Some of the work around like self detection of hallucination, it's also deployed for efficiency, right? So if you have someone like Rakuten doing it in production live, you know, that's just another thing people should consider.Mark Bissell [00:21:12]: Yeah. And something like a probe is super lightweight. Yeah. It's no extra latency really. Excellent.Shawn Wang [00:21:17]: You have the steering demos lined up. So we were just kind of see what you got. I don't, I don't actually know if this is like the latest, latest or like alpha thing.Mark Bissell [00:21:26]: No, this is a pretty hacky demo from from a presentation that someone else on the team recently gave. So this will give a sense for, for technology. So you can see the steering and action. Honestly, I think the biggest thing that this highlights is that as we've been growing as a company and taking on kind of more and more ambitious versions of interpretability related problems, a lot of that comes to scaling up in various different forms. And so here you're going to see steering on a 1 trillion parameter model. This is Kimi K2. And so it's sort of fun that in addition to the research challenges, there are engineering challenges that we're now tackling. Cause for any of this to be sort of useful in production, you need to be thinking about what it looks like when you're using these methods on frontier models as opposed to sort of like toy kind of model organisms. So yeah, this was thrown together hastily, pretty fragile behind the scenes, but I think it's quite a fun demo. So screen sharing is on. So I've got two terminal sessions pulled up here. On the left is a forked version that we have of the Kimi CLI that we've got running to point at our custom hosted Kimi model. And then on the right is a set up that will allow us to steer on certain concepts. So I should be able to chat with Kimi over here. Tell it hello. This is running locally. So the CLI is running locally, but the Kimi server is running back to the office. Well, hopefully should be, um, that's too much to run on that Mac. Yeah. I think it's, uh, it takes a full, like each 100 node. I think it's like, you can. You can run it on eight GPUs, eight 100. So, so yeah, Kimi's running. We can ask it a prompt. It's got a forked version of our, uh, of the SG line code base that we've been working on. So I'm going to tell it, Hey, this SG line code base is slow. I think there's a bug. Can you try to figure it out? There's a big code base, so it'll, it'll spend some time doing this. And then on the right here, I'm going to initialize in real time. Some steering. Let's see here.Mark Bissell [00:23:33]: searching for any. Bugs. Feature ID 43205.Shawn Wang [00:23:38]: Yeah.Mark Bissell [00:23:38]: 20, 30, 40. So let me, uh, this is basically a feature that we found that inside Kimi seems to cause it to speak in Gen Z slang. And so on the left, it's still sort of thinking normally it might take, I don't know, 15 seconds for this to kick in, but then we're going to start hopefully seeing him do this code base is massive for real. So we're going to start. We're going to start seeing Kimi transition as the steering kicks in from normal Kimi to Gen Z Kimi and both in its chain of thought and its actual outputs.Mark Bissell [00:24:19]: And interestingly, you can see, you know, it's still able to call tools, uh, and stuff. It's um, it's purely sort of it's it's demeanor. And there are other features that we found for interesting things like concision. So that's more of a practical one. You can make it more concise. Um, the types of programs, uh, programming languages that uses, but yeah, as we're seeing it come in. Pretty good. Outputs.Shawn Wang [00:24:43]: Scheduler code is actually wild.Vibhu Sapra [00:24:46]: Yo, this code is actually insane, bro.Vibhu Sapra [00:24:53]: What's the process of training in SAE on this, or, you know, how do you label features? I know you guys put out a pretty cool blog post about, um, finding this like autonomous interp. Um, something. Something about how agents for interp is different than like coding agents. I don't know while this is spewing up, but how, how do we find feature 43, two Oh five. Yeah.Mark Bissell [00:25:15]: So in this case, um, we, our platform that we've been building out for a long time now supports all the sort of classic out of the box interp techniques that you might want to have like SAE training, probing things of that kind, I'd say the techniques for like vanilla SAEs are pretty well established now where. You take your model that you're interpreting, run a whole bunch of data through it, gather activations, and then yeah, pretty straightforward pipeline to train an SAE. There are a lot of different varieties. There's top KSAEs, batch top KSAEs, um, normal ReLU SAEs. And then once you have your sparse features to your point, assigning labels to them to actually understand that this is a gen Z feature, that's actually where a lot of the kind of magic happens. Yeah. And the most basic standard technique is look at all of your d input data set examples that cause this feature to fire most highly. And then you can usually pick out a pattern. So for this feature, If I've run a diverse enough data set through my model feature 43, two Oh five. Probably tends to fire on all the tokens that sounds like gen Z slang. You know, that's the, that's the time of year to be like, Oh, I'm in this, I'm in this Um, and, um, so, you know, you could have a human go through all 43,000 concepts andVibhu Sapra [00:26:34]: And I've got to ask the basic question, you know, can we get examples where it hallucinates, pass it through, see what feature activates for hallucinations? Can I just, you know, turn hallucination down?Myra Deng [00:26:51]: Oh, wow. You really predicted a project we're already working on right now, which is detecting hallucinations using interpretability techniques. And this is interesting because hallucinations is something that's very hard to detect. And it's like a kind of a hairy problem and something that black box methods really struggle with. Whereas like Gen Z, you could always train a simple classifier to detect that hallucinations is harder. But we've seen that models internally have some... Awareness of like uncertainty or some sort of like user pleasing behavior that leads to hallucinatory behavior. And so, yeah, we have a project that's trying to detect that accurately. And then also working on mitigating the hallucinatory behavior in the model itself as well.Shawn Wang [00:27:39]: Yeah, I would say most people are still at the level of like, oh, I would just turn temperature to zero and that turns off hallucination. And I'm like, well, that's a fundamental misunderstanding of how this works. Yeah.Mark Bissell [00:27:51]: Although, so part of what I like about that question is you, there are SAE based approaches that might like help you get at that. But oftentimes the beauty of SAEs and like we said, the curse is that they're unsupervised. So when you have a behavior that you deliberately would like to remove, and that's more of like a supervised task, often it is better to use something like probes and specifically target the thing that you're interested in reducing as opposed to sort of like hoping that when you fragment the latent space, one of the vectors that pops out.Vibhu Sapra [00:28:20]: And as much as we're training an autoencoder to be sparse, we're not like for sure certain that, you know, we will get something that just correlates to hallucination. You'll probably split that up into 20 other things and who knows what they'll be.Mark Bissell [00:28:36]: Of course. Right. Yeah. So there's no sort of problems with like feature splitting and feature absorption. And then there's the off target effects, right? Ideally, you would want to be very precise where if you reduce the hallucination feature, suddenly maybe your model can't write. Creatively anymore. And maybe you don't like that, but you want to still stop it from hallucinating facts and figures.Shawn Wang [00:28:55]: Good. So Vibhu has a paper to recommend there that we'll put in the show notes. But yeah, I mean, I guess just because your demo is done, any any other things that you want to highlight or any other interesting features you want to show?Mark Bissell [00:29:07]: I don't think so. Yeah. Like I said, this is a pretty small snippet. I think the main sort of point here that I think is exciting is that there's not a whole lot of inter being applied to models quite at this scale. You know, Anthropic certainly has some some. Research and yeah, other other teams as well. But it's it's nice to see these techniques, you know, being put into practice. I think not that long ago, the idea of real time steering of a trillion parameter model would have sounded.Shawn Wang [00:29:33]: Yeah. The fact that it's real time, like you started the thing and then you edited the steering vector.Vibhu Sapra [00:29:38]: I think it's it's an interesting one TBD of what the actual like production use case would be on that, like the real time editing. It's like that's the fun part of the demo, right? You can kind of see how this could be served behind an API, right? Like, yes, you're you only have so many knobs and you can just tweak it a bit more. And I don't know how it plays in. Like people haven't done that much with like, how does this work with or without prompting? Right. How does this work with fine tuning? Like, there's a whole hype of continual learning, right? So there's just so much to see. Like, is this another parameter? Like, is it like parameter? We just kind of leave it as a default. We don't use it. So I don't know. Maybe someone here wants to put out a guide on like how to use this with prompting when to do what?Mark Bissell [00:30:18]: Oh, well, I have a paper recommendation. I think you would love from Act Deep on our team, who is an amazing researcher, just can't say enough amazing things about Act Deep. But he actually has a paper that as well as some others from the team and elsewhere that go into the essentially equivalence of activation steering and in context learning and how those are from a he thinks of everything in a cognitive neuroscience Bayesian framework, but basically how you can precisely show how. Prompting in context, learning and steering exhibit similar behaviors and even like get quantitative about the like magnitude of steering you would need to do to induce a certain amount of behavior similar to certain prompting, even for things like jailbreaks and stuff. It's a really cool paper. Are you saying steering is less powerful than prompting? More like you can almost write a formula that tells you how to convert between the two of them.Myra Deng [00:31:20]: And so like formally equivalent actually in the in the limit. Right.Mark Bissell [00:31:24]: So like one case study of this is for jailbreaks there. I don't know. Have you seen the stuff where you can do like many shot jailbreaking? You like flood the context with examples of the behavior. And the topic put out that paper.Shawn Wang [00:31:38]: A lot of people were like, yeah, we've been doing this, guys.Mark Bissell [00:31:40]: Like, yeah, what's in this in context learning and activation steering equivalence paper is you can like predict the number. Number of examples that you will need to put in there in order to jailbreak the model. That's cool. By doing steering experiments and using this sort of like equivalence mapping. That's cool. That's really cool. It's very neat. Yeah.Shawn Wang [00:32:02]: I was going to say, like, you know, I can like back rationalize that this makes sense because, you know, what context is, is basically just, you know, it updates the KV cache kind of and like and then every next token inference is still like, you know, the sheer sum of everything all the way. It's plus all the context. It's up to date. And you could, I guess, theoretically steer that with you probably replace that with your steering. The only problem is steering typically is on one layer, maybe three layers like like you did. So it's like not exactly equivalent.Mark Bissell [00:32:33]: Right, right. There's sort of you need to get precise about, yeah, like how you sort of define steering and like what how you're modeling the setup. But yeah, I've got the paper pulled up here. Belief dynamics reveal the dual nature. Yeah. The title is Belief Dynamics Reveal the Dual Nature of Incompetence. And it's an exhibition of the practical context learning and activation steering. So Eric Bigelow, Dan Urgraft on the who are doing fellowships at Goodfire, Ekt Deep's the final author there.Myra Deng [00:32:59]: I think actually to your question of like, what is the production use case of steering? I think maybe if you just think like one level beyond steering as it is today. Like imagine if you could adapt your model to be, you know, an expert legal reasoner. Like in almost real time, like very quickly. efficiently using human feedback or using like your semantic understanding of what the model knows and where it knows that behavior. I think that while it's not clear what the product is at the end of the day, it's clearly very valuable. Thinking about like what's the next interface for model customization and adaptation is a really interesting problem for us. Like we have heard a lot of people actually interested in fine-tuning an RL for open weight models in production. And so people are using things like Tinker or kind of like open source libraries to do that, but it's still very difficult to get models fine-tuned and RL'd for exactly what you want them to do unless you're an expert at model training. And so that's like something we'reShawn Wang [00:34:06]: looking into. Yeah. I never thought so. Tinker from Thinking Machines famously uses rank one LoRa. Is that basically the same as steering? Like, you know, what's the comparison there?Mark Bissell [00:34:19]: Well, so in that case, you are still applying updates to the parameters, right?Shawn Wang [00:34:25]: Yeah. You're not touching a base model. You're touching an adapter. It's kind of, yeah.Mark Bissell [00:34:30]: Right. But I guess it still is like more in parameter space then. I guess it's maybe like, are you modifying the pipes or are you modifying the water flowing through the pipes to get what you're after? Yeah. Just maybe one way.Mark Bissell [00:34:44]: I like that analogy. That's my mental map of it at least, but it gets at this idea of model design and intentional design, which is something that we're, that we're very focused on. And just the fact that like, I hope that we look back at how we're currently training models and post-training models and just think what a primitive way of doing that right now. Like there's no intentionalityShawn Wang [00:35:06]: really in... It's just data, right? The only thing in control is what data we feed in.Mark Bissell [00:35:11]: So, so Dan from Goodfire likes to use this analogy of, you know, he has a couple of young kids and he talks about like, what if I could only teach my kids how to be good people by giving them cookies or like, you know, giving them a slap on the wrist if they do something wrong, like not telling them why it was wrong or like what they should have done differently or something like that. Just figure it out. Right. Exactly. So that's RL. Yeah. Right. And, and, you know, it's sample inefficient. There's, you know, what do they say? It's like slurping feedback. It's like, slurping supervision. Right. And so you'd like to get to the point where you can have experts giving feedback to their models that are, uh, internalized and, and, you know, steering is an inference time way of sort of getting that idea. But ideally you're moving to a world whereVibhu Sapra [00:36:04]: it is much more intentional design in perpetuity for these models. Okay. This is one of the questions we asked Emmanuel from Anthropic on the podcast a few months ago. Basically the question, was you're at a research lab that does model training, foundation models, and you're on an interp team. How does it tie back? Right? Like, does this, do ideas come from the pre-training team? Do they go back? Um, you know, so for those interested, you can, you can watch that. There wasn't too much of a connect there, but it's still something, you know, it's something they want toMark Bissell [00:36:33]: push for down the line. It can be useful for all of the above. Like there are certainly post-hocVibhu Sapra [00:36:39]: use cases where it doesn't need to touch that. I think the other thing a lot of people forget is this stuff isn't too computationally expensive, right? Like I would say, if you're interested in getting into research, MechInterp is one of the most approachable fields, right? A lot of this train an essay, train a probe, this stuff, like the budget for this one, there's already a lot done. There's a lot of open source work. You guys have done some too. Um, you know,Shawn Wang [00:37:04]: There's like notebooks from the Gemini team for Neil Nanda or like, this is how you do it. Just step through the notebook.Vibhu Sapra [00:37:09]: Even if you're like, not even technical with any of this, you can still make like progress. There, you can look at different activations, but, uh, if you do want to get into training, you know, training this stuff, correct me if I'm wrong is like in the thousands of dollars, not even like, it's not that high scale. And then same with like, you know, applying it, doing it for post-training or all this stuff is fairly cheap in scale of, okay. I want to get into like model training. I don't have compute for like, you know, pre-training stuff. So it's, it's a very nice field to get into. And also there's a lot of like open questions, right? Um, some of them have to go with, okay, I want a product. I want to solve this. Like there's also just a lot of open-ended stuff that people could work on. That's interesting. Right. I don't know if you guys have any calls for like, what's open questions, what's open work that you either open collaboration with, or like, you'd just like to see solved or just, you know, for people listening that want to get into McInturk because people always talk about it. What are, what are the things they should check out? Start, of course, you know, join you guys as well. I'm sure you're hiring.Myra Deng [00:38:09]: There's a paper, I think from, was it Lee, uh, Sharky? It's open problems and, uh, it's, it's a bit of interpretability, which I recommend everyone who's interested in the field. Read. I'm just like a really comprehensive overview of what are the things that experts in the field think are the most important problems to be solved. I also think to your point, it's been really, really inspiring to see, I think a lot of young people getting interested in interpretability, actually not just young people also like scientists to have been, you know, experts in physics for many years and in biology or things like this, um, transitioning into interp, because the barrier of, of what's now interp. So it's really cool to see a number to entry is, you know, in some ways low and there's a lot of information out there and ways to get started. There's this anecdote of like professors at universities saying that all of a sudden every incoming PhD student wants to study interpretability, which was not the case a few years ago. So it just goes to show how, I guess, like exciting the field is, how fast it's moving, how quick it is to get started and things like that.Mark Bissell [00:39:10]: And also just a very welcoming community. You know, there's an open source McInturk Slack channel. There are people are always posting questions and just folks in the space are always responsive if you ask things on various forums and stuff. But yeah, the open paper, open problems paper is a really good one.Myra Deng [00:39:28]: For other people who want to get started, I think, you know, MATS is a great program. What's the acronym for? Machine Learning and Alignment Theory Scholars? It's like the...Vibhu Sapra [00:39:40]: Normally summer internship style.Myra Deng [00:39:42]: Yeah, but they've been doing it year round now. And actually a lot of our full-time staff have come through that program or gone through that program. And it's great for anyone who is transitioning into interpretability. There's a couple other fellows programs. We do one as well as Anthropic. And so those are great places to get started if anyone is interested.Mark Bissell [00:40:03]: Also, I think been seen as a research field for a very long time. But I think engineering... I think engineers are sorely wanted for interpretability as well, especially at Goodfire, but elsewhere, as it does scale up.Shawn Wang [00:40:18]: I should mention that Lee actually works with you guys, right? And in the London office and I'm adding our first ever McInturk track at AI Europe because I see this industry applications now emerging. And I'm pretty excited to, you know, help push that along. Yeah, I was looking forward to that. It'll effectively be the first industry McInturk conference. Yeah. I'm so glad you added that. You know, it's still a little bit of a bet. It's not that widespread, but I can definitely see this is the time to really get into it. We want to be early on things.Mark Bissell [00:40:51]: For sure. And I think the field understands this, right? So at ICML, I think the title of the McInturk workshop this year was actionable interpretability. And there was a lot of discussion around bringing it to various domains. Everyone's adding pragmatic, actionable, whatever.Shawn Wang [00:41:10]: It's like, okay, well, we weren't actionable before, I guess. I don't know.Vibhu Sapra [00:41:13]: And I mean, like, just, you know, being in Europe, you see the Interp room. One, like old school conferences, like, I think they had a very tiny room till they got lucky and they got it doubled. But there's definitely a lot of interest, a lot of niche research. So you see a lot of research coming out of universities, students. We covered the paper last week. It's like two unknown authors, not many citations. But, you know, you can make a lot of meaningful work there. Yeah. Yeah. Yeah.Shawn Wang [00:41:39]: Yeah. I think people haven't really mentioned this yet. It's just Interp for code. I think it's like an abnormally important field. We haven't mentioned this yet. The conspiracy theory last two years ago was when the first SAE work came out of Anthropic was they would do like, oh, we just used SAEs to turn the bad code vector down and then turn up the good code. And I think like, isn't that the dream? Like, you know, like, but basically, I guess maybe, why is it funny? Like, it's... If it was realistic, it would not be funny. It would be like, no, actually, we should do this. But it's funny because we know there's like, we feel there's some limitations to what steering can do. And I think a lot of the public image of steering is like the Gen Z stuff. Like, oh, you can make it really love the Golden Gate Bridge, or you can make it speak like Gen Z. To like be a legal reasoner seems like a huge stretch. Yeah. And I don't know if that will get there this way. Yeah.Myra Deng [00:42:36]: I think, um, I will say we are announcing. Something very soon that I will not speak too much about. Um, but I think, yeah, this is like what we've run into again and again is like, we, we don't want to be in the world where steering is only useful for like stylistic things. That's definitely not, not what we're aiming for. But I think the types of interventions that you need to do to get to things like legal reasoning, um, are much more sophisticated and require breakthroughs in, in learning algorithms. And that's, um...Shawn Wang [00:43:07]: And is this an emergent property of scale as well?Myra Deng [00:43:10]: I think so. Yeah. I mean, I think scale definitely helps. I think scale allows you to learn a lot of information and, and reduce noise across, you know, large amounts of data. But I also think we think that there's ways to do things much more effectively, um, even, even at scale. So like actually learning exactly what you want from the data and not learning things that you do that you don't want exhibited in the data. So we're not like anti-scale, but we are also realizing that scale is not going to get us anywhere. It's not going to get us to the type of AI development that we want to be at in, in the future as these models get more powerful and get deployed in all these sorts of like mission critical contexts. Current life cycle of training and deploying and evaluations is, is to us like deeply broken and has opportunities to, to improve. So, um, more to come on that very, very soon.Mark Bissell [00:44:02]: And I think that that's a use basically, or maybe just like a proof point that these concepts do exist. Like if you can manipulate them in the precise best way, you can get the ideal combination of them that you desire. And steering is maybe the most coarse grained sort of peek at what that looks like. But I think it's evocative of what you could do if you had total surgical control over every concept, every parameter. Yeah, exactly.Myra Deng [00:44:30]: There were like bad code features. I've got it pulled up.Vibhu Sapra [00:44:33]: Yeah. Just coincidentally, as you guys are talking.Shawn Wang [00:44:35]: This is like, this is exactly.Vibhu Sapra [00:44:38]: There's like specifically a code error feature that activates and they show, you know, it's not, it's not typo detection. It's like, it's, it's typos in code. It's not typical typos. And, you know, you can, you can see it clearly activates where there's something wrong in code. And they have like malicious code, code error. They have a whole bunch of sub, you know, sub broken down little grain features. Yeah.Shawn Wang [00:45:02]: Yeah. So, so the, the rough intuition for me, the, why I talked about post-training was that, well, you just, you know, have a few different rollouts with all these things turned off and on and whatever. And then, you know, you can, that's, that's synthetic data you can kind of post-train on. Yeah.Vibhu Sapra [00:45:13]: And I think we make it sound easier than it is just saying, you know, they do the real hard work.Myra Deng [00:45:19]: I mean, you guys, you guys have the right idea. Exactly. Yeah. We replicated a lot of these features in, in our Lama models as well. I remember there was like.Vibhu Sapra [00:45:26]: And I think a lot of this stuff is open, right? Like, yeah, you guys opened yours. DeepMind has opened a lot of essays on Gemma. Even Anthropic has opened a lot of this. There's, there's a lot of resources that, you know, we can probably share of people that want to get involved.Shawn Wang [00:45:41]: Yeah. And special shout out to like Neuronpedia as well. Yes. Like, yeah, amazing piece of work to visualize those things.Myra Deng [00:45:49]: Yeah, exactly.Shawn Wang [00:45:50]: I guess I wanted to pivot a little bit on, onto the healthcare side, because I think that's a big use case for you guys. We haven't really talked about it yet. This is a bit of a crossover for me because we are, we are, we do have a separate science pod that we're starting up for AI, for AI for science, just because like, it's such a huge investment category and also I'm like less qualified to do it, but we actually have bio PhDs to cover that, which is great, but I need to just kind of recover, recap your work, maybe on the evil two stuff, but then, and then building forward.Mark Bissell [00:46:17]: Yeah, for sure. And maybe to frame up the conversation, I think another kind of interesting just lens on interpretability in general is a lot of the techniques that were described. are ways to solve the AI human interface problem. And it's sort of like bidirectional communication is the goal there. So what we've been talking about with intentional design of models and, you know, steering, but also more advanced techniques is having humans impart our desires and control into models and over models. And the reverse is also very interesting, especially as you get to superhuman models, whether that's narrow superintelligence, like these scientific models that work on genomics, data, medical imaging, things like that. But down the line, you know, superintelligence of other forms as well. What knowledge can the AIs teach us as sort of that, that the other direction in that? And so some of our life science work to date has been getting at exactly that question, which is, well, some of it does look like debugging these various life sciences models, understanding if they're actually performing well, on tasks, or if they're picking up on spurious correlations, for instance, genomics models, you would like to know whether they are sort of focusing on the biologically relevant things that you care about, or if it's using some simpler correlate, like the ancestry of the person that it's looking at. But then also in the instances where they are superhuman, and maybe they are understanding elements of the human genome that we don't have names for or specific, you know, yeah, discoveries that they've made that that we don't know about, that's, that's a big goal. And so we're already seeing that, right, we are partnered with organizations like Mayo Clinic, leading research health system in the United States, our Institute, as well as a startup called Prima Menta, which focuses on neurodegenerative disease. And in our partnership with them, we've used foundation models, they've been training and applied our interpretability techniques to find novel biomarkers for Alzheimer's disease. So I think this is just the tip of the iceberg. But it's, that's like a flavor of some of the things that we're working on.Shawn Wang [00:48:36]: Yeah, I think that's really fantastic. Obviously, we did the Chad Zuckerberg pod last year as well. And like, there's a plethora of these models coming out, because there's so much potential and research. And it's like, very interesting how it's basically the same as language models, but just with a different underlying data set. But it's like, it's the same exact techniques. Like, there's no change, basically.Mark Bissell [00:48:59]: Yeah. Well, and even in like other domains, right? Like, you know, robotics, I know, like a lot of the companies just use Gemma as like the like backbone, and then they like make it into a VLA that like takes these actions. It's, it's, it's transformers all the way down. So yeah.Vibhu Sapra [00:49:15]: Like we have Med Gemma now, right? Like this week, even there was Med Gemma 1.5. And they're training it on this stuff, like 3d scans, medical domain knowledge, and all that stuff, too. So there's a push from both sides. But I think the thing that, you know, one of the things about McInturpp is like, you're a little bit more cautious in some domains, right? So healthcare, mainly being one, like guardrails, understanding, you know, we're more risk adverse to something going wrong there. So even just from a basic understanding, like, if we're trusting these systems to make claims, we want to know why and what's going on.Myra Deng [00:49:51]: Yeah, I think there's totally a kind of like deployment bottleneck to actually using. foundation models for real patient usage or things like that. Like, say you're using a model for rare disease prediction, you probably want some explanation as to why your model predicted a certain outcome, and an interpretable explanation at that. So that's definitely a use case. But I also think like, being able to extract scientific information that no human knows to accelerate drug discovery and disease treatment and things like that actually is a really, really big unlock for science, like scientific discovery. And you've seen a lot of startups, like say that they're going to accelerate scientific discovery. And I feel like we actually are doing that through our interp techniques. And kind of like, almost by accident, like, I think we got reached out to very, very early on from these healthcare institutions. And none of us had healthcare.Shawn Wang [00:50:49]: How did they even hear of you? A podcast.Myra Deng [00:50:51]: Oh, okay. Yeah, podcast.Vibhu Sapra [00:50:53]: Okay, well, now's that time, you know.Myra Deng [00:50:55]: Everyone can call us.Shawn Wang [00:50:56]: Podcasts are the most important thing. Everyone should listen to podcasts.Myra Deng [00:50:59]: Yeah, they reached out. They were like, you know, we have these really smart models that we've trained, and we want to know what they're doing. And we were like, really early that time, like three months old, and it was a few of us. And we were like, oh, my God, we've never used these models. Let's figure it out. But it's also like, great proof that interp techniques scale pretty well across domains. We didn't really have to learn too much about.Shawn Wang [00:51:21]: Interp is a machine learning technique, machine learning skills everywhere, right? Yeah. And it's obviously, it's just like a general insight. Yeah. Probably to finance too, I think, which would be fun for our history. I don't know if you have anything to say there.Mark Bissell [00:51:34]: Yeah, well, just across the science. Like, we've also done work on material science. Yeah, it really runs the gamut.Vibhu Sapra [00:51:40]: Yeah. Awesome. And, you know, for those that should reach out, like, you're obviously experts in this, but like, is there a call out for people that you're looking to partner with, design partners, people to use your stuff outside of just, you know, the general developer that wants to. Plug and play steering stuff, like on the research side more so, like, are there ideal design partners, customers, stuff like that?Myra Deng [00:52:03]: Yeah, I can talk about maybe non-life sciences, and then I'm curious to hear from you on the life sciences side. But we're looking for design partners across many domains, language, anyone who's customizing language models or trying to push the frontier of code or reasoning models is really interesting to us. And then also interested in the frontier of modeling. There's a lot of models that work in, like, pixel space, as we call it. So if you're doing world models, video models, even robotics, where there's not a very clean natural language interface to interact with, I think we think that Interp can really help and are looking for a few partners in that space.Shawn Wang [00:52:43]: Just because you mentioned the keyword

The Department of Employment and Labour has announced the latest updates to South Africa's National Minimum Wage, affecting domestic workers, farm workers, and other applicable employees. Following recommendations from the National Minimum Wage Council, the rate has been increased by 5%, moving from R28.79 to R30.23 per ordinary hour worked. For a standard 45-hour work week, the increase translates to a weekly wage of R1,264.85. When calculated over a month of 4.3 weeks (or 195 hours), the minimum becomes R5,894.40. For households employing domestic workers on a more typical 160-hour month, the minimum monthly wage rises to R4,834, up from R4,606 in 2025. Households must also note that South Africa's minimum wage laws require employers to pay for at least four hours of work each day, regardless of the actual hours worked. This sets the true daily minimum for domestic workers at approximately R121, an increase from R115 last year. Domestic workers have been fully covered by the National Minimum Wage since 2022. However, data suggests that many workers are still earning below the legal minimum. According to BusinessTech, median salaries reported by Stats SA show that domestic workers earn around R2,350 per month, equivalent to R14.69 per hour for a 160-hour month. This is less than half of the 2026 minimum wage.

In this episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, Dave Newman discusses how Sanford Health is expanding access by innovating virtual care—meeting patients where they are, including by phone. He shares insight into preventive strategies for chronic kidney disease and emphasizes how collaboration across teams and technologies serves as a powerful catalyst for progress in modern care delivery.In collaboration with R1.

In this episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, Abha Agrawal discusses how NorthStar Hospitals is reshaping rural healthcare by advancing the technology agenda and elevating the patient experience. She shares insights into building sustainable, tech-forward systems that improve access and outcomes for communities often left behind.In collaboration with R1.

Jannie Stockenström, beleggingshoof van WIPHOLD, gesels oor die R1,4 miljard Zaad-groep verkryging. Volg RSG Geldsake op Twitter

In this episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, Michael Mutterer discusses leading growth in an increasingly competitive healthcare landscape and the strategies Silver Cross Hospital is using to elevate patient satisfaction. He also highlights how technology is helping strengthen payment and revenue cycle performance, creating a more sustainable and efficient operational model.In collaboration with R1.

In this episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable Michael Charlton discusses the importance of serving underserved communities while supporting caregiver satisfaction. He shares how AtlantiCare is investing in technology across both the payer and provider landscape, and how these advancements are shaping the future workforce. He also explores the evolving role of AI and its potential impact on staffing and care delivery.In collaboration with R1.

The world is changing at an exponential rate. By 2030, two billion jobs will be transformed or displaced. Are you prepared to bridge the gap between innovation and adoption, or will you fall into it?In this "Golden Nuggets" highlight, we're extracting the high-level leadership blueprint from John Rathje, VP and CIO of Kent State University. John oversees the tech strategy for a top-tier R1 research powerhouse—the same place that birthed the liquid crystal technology in your smartphone.From writing code on a borrowed computer to leading global tech shifts, John shares why "looking up" is the most important skill an entrepreneur can have in the next decade.Key Takeaways:(1) Stop Hunting, Start Solving: Don't build a product and look for a market; find a massive pain point and obsess over it.(2) Environment is Everything: You are the architect of your team's creativity. Build an "Innovation Hub," not just a workspace.(3) Human Impact to Code: Technology is built by people, for people. If you lose the human element, you lose your competitive edge.Listen To The Full Podcast Episode Here: https://www.youtube.com/watch?v=2vr-_ybTrFI#FutureOfWork #EntrepreneurMindset #LeadershipTips #TechInnovation #GoldenNuggets #JohnRathje #2030Vision #BusinessGrowth #StartupAdvice

This episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, features Dr. Shondra Williams, President & Chief Executive Officer at InclusivCare, as she shares how her organization is navigating Medicaid uncertainty, financial pressure, and patient access challenges. Dr. Williams also discusses leadership, culture, and the role of technology and AI in strengthening community health centers heading into 2026.In collaboration with R1.

In this episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, Dr. Mike Guertin discusses how AI is creating new opportunities to advance patient care, particularly in perioperative settings. He highlights how emerging technologies are improving surgical efficiency by streamlining information gathering, reporting, and clinical decision support.In collaboration with R1.

Rodalies reabre las líneas R1, R2 y parte de la R4. La investigación apunta a la soldadura como causa del descarrilamiento en Adamuz. El Gobierno pacta la regularización de medio millón de inmigrantes irregulares, con requisitos de cinco meses de estancia y documentación; PP y Vox alertan de un efecto llamada. Europa investiga a X por su IA Grok, que genera fotos explícitas de menores. Francia prohíbe redes sociales a menores de 15 años y planea vetar móviles en institutos. En '¡Buenos días, Javi y Mar!', hablan de chapuzas domésticas y de la preferencia por una buena distribución de grasa corporal sobre la musculatura en hombres. CADENA 100 ofrece la 'Encuesta Absurda' y la mejor música.

La investigación del accidente de tren de Adamuz confirma que la causa es una vía rota, indetectable por alarmas debido a su leve fractura, lo que provoca que políticos pidan responsabilidades. Adif revisa tramos similares y reduce la velocidad en la línea Madrid-Valladolid. La borrasca Ingrid azota el noreste peninsular con nieve, afectando a Galicia con olas de diez metros, restricciones de tráfico y mil camiones embolsados. En Cataluña, los Rodalies reanudan servicio tras un accidente, pero un nuevo desprendimiento corta la línea R1. La Fiscalía archiva denuncias contra Julio Iglesias por falta de competencia. España y Europa enfrentan una crisis demográfica por la inestabilidad juvenil. La patronal propone que las empresas no paguen cotizaciones por bajas para frenar el absentismo, que ha aumentado un 50% y cuesta 33.000 millones. Madrid sufre bajas temperaturas y posibles nevadas; pide a la UE no activar el acuerdo con Mercosur para proteger a agricultores.

El accidente ferroviario de Adamuz se atribuye a una fractura en la vía según CIAF. Adif revisa carriles por fallos y el ministro Puente admite que sensores no detectan anomalías. Pedro Sánchez solicita comparecer en el Congreso por la situación ferroviaria. Se destaca la labor de profesionales y vecinos en la emergencia. La línea R1 de Rodalies sufre un corte. La Audiencia de A Coruña confirma la condena al maquinista del accidente de Angrois. El temporal de frío y nieve afecta a 80 carreteras y a un millar de camiones, con Galicia en alerta roja por olas y nieve. Internacionalmente, Ucrania, Rusia y Estados Unidos celebran su primera reunión trilateral en Emiratos Árabes, centrada en el Donbass. La Fiscalía archiva la denuncia contra Julio Iglesias. El juicio al hermano de Pedro Sánchez, David Sánchez, por prevaricación y tráfico de influencias, se realizará en mayo y junio. Se investiga a Xavier García Albiol por desalojo de migrantes en Badalona. Feijóo comparece el 2 de ...

In this episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, Mark Behl, President & CEO of NorthBay Health, discusses investing in new technology, collaborating with payers to enhance patient experience, creating tech solutions to improve efficiency, and advancing value-based care initiatives.In collaboration with R1.

In this episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, Matthew Love, President and CEO of Nicklaus Children's Health System, discusses initiatives to develop a strong pediatric cancer program in Florida, navigating cost and reimbursement pressures, and handling AI governance, including insights from the “Ask Nick” program.In collaboration with R1.

In this episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, Garrick Stoldt, VP Finance and Chief Financial Officer at Saint Peter's Healthcare System, discusses the effects of the Big Beautiful Bill, the rapid expansion of automation across healthcare finance, and why maintaining a strong human role remains essential as organizations modernize.In collaboration with R1.

Cracked Racquets Editor-in-Chief Alex Gruskin recaps Day 3 of the 2026 Australian Open. He reflects upon a relatively simple opening round for many of the tournament's top title contenders. He also breaks down a marvelous R1 of results for men with college tennis ties, looks at those players continuing to trend upwards, previews Day 4's action, plus SO much more!! Don't forget to give a 5 star review on your favorite podcast app! In addition, add your twitter/instagram handle to the review for a chance to win some FREE CR gear!! Episode Bookmarks Biggest Storylines - 6:00 Title Contenders cruise through R1: 6:25 Men w/College Tennis Ties: 22:20 Continuing to trend up: 37:53 Upsets - 46:24 Going the Distance - 48:25 Other Women's Results - 51:43 Other Men's Results - 53:05 American Update - 54:35 Players w/College Ties Update - 58:20 Forecast heading into R2/Day 4 Preview - 59:25 _____ Laurel Springs Ranked among the best online private schools in the United States, Laurel Springs stands out when it comes to support, personalization, community, and college prep. They give their K-12 students the resources, guidance, and learning opportunities they need at each grade level to reach their full potential. Find Cracked Racquets Website: https://www.crackedracquets.com Instagram: https://instagram.com/crackedracquets Twitter: https://twitter.com/crackedracquets Facebook: https://Facebook.com/crackedracquets YouTube: https://www.youtube.com/c/crackedracquets Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

On this one Zac and Joe talk about what is new at Donathen RC, the new B7.1, The new R1 buggy, and the discussion about having refs at large, expensive races.

In this episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, Brian Peters, CEO of the Michigan Health & Hospital Association, discusses the push for fair and sustainable funding, evolving cybersecurity efforts across hospitals, and how leaders are navigating today's financial pressures with resilience and strategy.In collaboration with R1.

Episode 273-Warning: Critical Gun Law Alert  Also Available OnSearchable Podcast Transcript Gun Lawyer — Episode Transcript Gun Lawyer — Episode 273 Transcript SUMMARY KEYWORDS New Jersey gun laws, accidental discharge, criminalization, reckless discharge, felony consequences, gun ownership rights, self-defense, insurance coverage, Second Amendment, gun safety, gun dealers, international disarmament, gun control, gun owner education, legal challenges. SPEAKERS Speaker 2, Evan Nappen, Teddy Nappen Evan Nappen 00:17 I’m Evan Nappen. Teddy Nappen 00:19 And I’m Teddy Nappen. Evan Nappen 00:20 And welcome to Gun Lawyer. Well, folks, the New Jersey legislature has done it again. They have passed some atrocious gun laws, and I need to make all of you aware of one, particularly, that is very much a threat. It is something that’s going to affect many, many gun owners, and it is not being talked about in the general media, of course, because of how they write these laws in such a sneaky, underhanded way. But this law is going to impact all of us, frankly. And the potential is there, under this law, to not only take away gun owners’ rights to have guns, but to turn us all into felons at any time, simply based on an accident. That’s right, an accident. Evan Nappen 01:31 Because what New Jersey legislature’s both houses have passed, and I expect, very shortly, the governor will sign, is Assembly Bill, 4976. (https://pub.njleg.gov/Bills/2024/A5000/4976_R2.PDF) And what this bill does is it criminalizes Accidental Discharges (ADs). Now, an accidental discharge is when your gun goes off, accidentally, either by what some folks call an uncommanded discharge or an accidental discharge. But it is something that can happen, and although we have to always be very careful, circumstances can be such that a mistake can be made. I mean, we’re all human, and mistakes can happen. And unfortunately, you know, I see it in the practice, and I get accidental discharge cases all the time where individuals make a mistake and a gun goes off unintended. It happens. Now sometimes it happens because of the actual mechanical flaws to a firearm and that can be because of a gun’s design. It can even be due to circumstances where a firearm can go off from the slightest touch. Evan Nappen 03:08 Now you may not be aware of this, but years and years ago, I know of a case where an individual had a shotgun that this fellow had kept loaded. One of those single shot, top-breaker type shotguns. You know, like the old kind of like the toppers, the H and R Toppers, and what have you, similar to that. It Page – 1 – of 11may even have been one. I don’t know. But it’s one of those old single shot shotguns. And for probably 50 years, that gun had remained loaded with a shell in it. At one point, there were folks that were causing all kinds of problems in this guy’s yard, and he wanted to scare them off. He didn’t want to shoot them, and he put the gun out of, pointed the gun out the window, and boom. It went off, and he never pulled the trigger. He absolutely never pulled the trigger. There was no hit to the primer of the shell when it went off. And what has happened was, in this particular design of the gun, the firing pin had been pushing against the primer because the hammer was down and it didn’t have a firing pin block. And for like 50 years, this gun sat there, sensitizing the primer so that the slightest touch, you know, just the right jolt, without having to actually pull the trigger, made it go off. So, a gun can actually even do that under those extraordinary circumstances. Evan Nappen 04:57 But normally, an accidental discharge or uncommanded discharge that we encounter is because somebody believed, honestly believed, their gun was unloaded. And it ends up, of course, that it wasn’t. Now this can happen because somebody thought they checked it and maybe even did check. But then, with a magazine in and the slide going forward, a round loads, but they didn’t realize that it loaded, because they checked that it was unloaded. And sure enough, there’s a round there. I mean, I’ve seen every combination of error that could happen and a discharge can occur. And, of course, we know the rules, always point in a safe direction, etc. Make sure your gun is unloaded. Double, triple check to make sure that the chamber is empty. That there’s no magazine, and there’s no live ammo. I mean, all those things that we do. But accidents can happen, just like in a motor vehicle. We drive as safe as we possibly can, but people still have accidents. And what New Jersey has done in this bill is essentially criminalize an accident so that individuals will be looking at what is, in all likelihood, felony level charges. And they structured this bill in just a sneaky, evil, devious way. It’s going to have great impact, and it’s going to create, I think, unintended consequences. Evan Nappen 06:40 Now, as gun owners, we have to understand how we have to behave if any of us ever are so unfortunate as to have an uncommanded or accidental discharge. So, the law talks about “recklessly” having a discharge. “Reckless” in criminal law means, you know, with a conscious disregard of a known risk, okay? Criminal laws can have recklessly or reckless as a standard, as opposed to something being intentional, right? So, if you intentionally meant to pull the trigger, that’s intentional. Reckless could still be you didn’t intend to do it. But if there was that conscious disregard of that known risk and it ended up discharged, then you could argue that it’s reckless. So, reckless is kind of a standard where it’s not that traditional mens rea, your mental and your mental state of having that intention to have the gun fire. Reckless has been in our criminal law for a long time, and reckless conduct is something that’s out there, like reckless driving. I’m sure that you have heard of that. Evan Nappen 08:05 But what they’re doing here is even more devious by using the word “reckless”. So, what now is being prohibited? And I’m going to read this to you so you can see how they’ve done this. It says, a person commits a disorderly person’s offense. Now that sounds like, okay. A disorderly persons offense in New Jersey is equivalent to a misdemeanor. It’s not a felony. So, you’re saying, well, first of all, this is not creating a felony. It’s creating a disorderly persons offense, right? It sounds like it’s, you know, Page – 2 – of 11apparently, trying to be reasonable. But trust me, folks, it isn’t. I’m going to show you why. “A person commits a disorderly persons offense by recklessly discharging a firearm.” Okay, so at this point they’re saying, well, it’s just a low level offense, and it’s for recklessly discharge. You know, we’ve conscientiously disregarded a known risk. Okay, so it started out sounding, you know, not great, but okay. It’s not. It shouldn’t affect a lot of folks, and luckily, if it does, it’s still a misdemeanor. And, of course, it requires that recklessness. So, that sounds all good. Evan Nappen 09:22 Let me start again and read you, but wait until you hear the rest of it. A person commits a disorderly persons offense by recklessly discharging a firearm “using live ammunition rounds”. Well, okay, that’s good to know. A blank gun isn’t a reckless discharge, but you know you’re firing a blank. No live ammo. Okay. So, if I’m firing dead ammo or ammo that’s not live, then that’s not a reckless discharge. Well, good. How do I get a discharge with ammo that’s not live? I don’t know how that’s even going to happen. But okay, they throw that in, probably more as subterfuge and, you know, smoke and mirrors. But again, here we go. “A person commits a disorderly persons offense by recklessly discharging a firearm using live ammunition rounds unlawfully . . .” Okay, unlawfully. So, you’re unlawful. “. . . or without a lawful purpose.” Whoops, wait a minute. “Without a lawful purpose.” You commit a disorderly persons offense by recklessly discharging a firearm without lawful purpose. Evan Nappen 10:35 Except that a second conviction for such an offense constitutes a crime of the fourth degree. Well, a crime is a felony, and that’s a fourth degree. It’s a year and a half in jail. And a third or subsequent conviction is a third degree and that’s five years in State Prison. Okay. So, you may even read this part and say, well, it’s still arguably, weirdly reckless, maybe. But it’s for discharging a firearm without lawful purpose, but at least it’s a disorderly persons offense. And I, boy, if we do it once, I sure wouldn’t think I’d do it again. So, why is this such a problem, you know. Evan Nappen 11:09 But oh, well, wait, wait, wait. We’re not done yet. Because then it says, if a person commits a violation under this section, you’re charged with a crime one degree higher than what ordinarily would be charged for such an offense when the violation occurs within 100 yards, 100 yards, folks. Not 100 feet. A football field’s worth of distance of an occupied structure. Oh, what’s an occupied structure? Any building, room, ship, vessel, car, vehicle, or airplane, or a place adopted for overnight accommodations of persons or for carrying on business therein. Wait a minute, wait a minute. Wait a minute! An occupied structure includes a car or vehicle, and it doesn’t even mean it has to be occupied. It means even a vehicle or a building or a room, and it has to be within 100 yards, a football field, of a car. If there’s a car driving by within 100 yards where the accidental discharge takes place. If you’re in your own home? I mean, this is basically every accidental discharge. You will probably be within 100 yards of a car or a building or a room, or hotel or whatever, or an airplane. Man, even if the airplane is flying over the sky, I don’t know. I mean, this is nuts. Evan Nappen 12:55 So, if the violation occurs within 100 yards of a “structure”, guess what? It’s no longer that disorderly persons offense. It’s bumped instantly to the fourth degree, felony level offense. Up to a year and a half Page – 3 – of 11in State Prison, and now you’re going to be a convicted felon. That’s if your gun discharged for not having, without a lawful purpose. Oh, you mean like an accident? Yeah! Like an accident. An accident because you didn’t have a lawful purpose. Did you lawfully have a purposeful accident? No, that’s silliness in a nutshell. So, what it means now is essentially any accidental discharge is a felony in New Jersey, and you can face State Prison time of at least a year and a half, unless it’s going to be enhanced even more based on these other factors. And as a felon, you lose your gun rights for the entire United States. Evan Nappen 14:12 And even if it’s kept at the misdemeanor, a so-called disorderly persons level, they’re still going to go after your gun license and your gun rights. They’ll claim, under Chapter 58-3 of the licensing law, that you’re somehow a danger to public health, safety, welfare. You think if you’re going to have a criminal charge, a criminal offense charged here of accidental discharge, where they’re classifying it as reckless because it went for a “an unlawful purpose”. Like I said, I don’t know how you have a lawful purpose accident. And it was somehow within 100 yards of any car or room, which made it originally a felony even, right? Felony level in New Jersey. You’re getting your license and your guns confiscated and taken and face prosecution over this insane law. Evan Nappen 15:17 Now, this is the consequences of this bill, right? But that’s just the consequences in the law itself, like the penalty you may face and licensing problems. But what it also means is that upon any accidental discharge, folks, any, you immediately, now, immediately, have a Fifth Amendment right against self- incrimination, and you’re going to have to stand by that. Because I know in many of the cases we’ve seen, someone had an accidental discharge, and it may have gone through their wall. It may have gone to a neighbor’s house. It may have not whatever. But if you react, if you call the police, if you try to find out what happened, any type, you’re getting criminally charged. You have a right to say nothing. You have an absolute right, a Fifth Amendment right to remain silent, because you will end up incriminating yourself. This is going to mean that any New Jersey gun owner who has an AD or an uncommanded accidental discharge needs to immediately take the Fifth and seek counsel, the Sixth Amendment. Just call your attorney and don’t say anything to anyone. Do not make any statements to law enforcement or anybody. And, you know, this is a shame. Because what if that round actually caused injury to somebody? Teddy Nappen 16:59 Actually take it a step further. Evan Nappen 17:01 Think about it. You’re gonna incriminate yourself. You gotta absolutely. Go ahead, Teddy. Teddy Nappen 17:07 Take it a step further. Imagine instead of “gun”, this was “car”. I asked. I was in. I got into a car accident. So, therefore, all car accidents are felonies, where there is nowhere. Were you back? Were you 100 feet from your driveway? Was there a car driving by? Did you back into that car? Felony! You are now a felon because of that. And don’t tell me it’s the firearm versus the car! Because the car is a Page – 4 – of 11two ton steel death machine that kills more people than firearms do. So, it’s that level of argument, the utter draconian insanity that they have created here. Where from an accident, an actual accident, God forbid. Evan Nappen 17:54 An accident. That is right. Teddy Nappen 17:56 You are guilty until proven innocent. Evan Nappen 17:59 And wait. Let me say this. This has been put out there as a possible problem for self-defenders. And that’s actually not completely accurate, because there’s an exemption here that says it’s an affirmative defense, if you fired your gun in self-defense. Okay. Affirmative defense means the burden is on you to prove that you acted in self-defense. Then they’ll say, okay, that wasn’t a reckless discharge. But even the fact that the legislature has to put in there that if you act in self-defense, it’s an affirmative defense. Well, wait a minute. Why is it an affirmative defense? Because it wouldn’t have been reckless if it was intentional. Why do we even need that? So, in other words, the legislature itself knows that they’ve manipulated this law to simply be discharge for unlawful purpose, period. If you didn’t have a lawful purpose when your gun went off, it’s felony level if it’s within 100 yards of a car, or a room, or a building. Insanity. Evan Nappen 19:05 And as you say, Teddy, it would be like making every car accident, any fender bender that you have, you become charged with a felony. New Jersey has done that to gun owners now. Any accident, any accidental discharge, you’re going to face these criminal charges. This is going to, you’re going to end up in the system. If you have an AD, you’re getting charged. And now we’re going to have to fight this out on an offense that is essentially strict liability. That is the way they’ve set it up. Couching it and hiding it under so-called reckless, recklessly. But when they actually write it, they put the recklessly with the little bonus of having “without a lawful purpose”. This is nuts. Nuts. Nuts. Evan Nappen 19:58 I’m telling you right now the cases we get, it’s going to be crazy, crazy and a problem. So, folks, be extra careful. This is bad news. It has passed both houses, and the Governor, I’m sure, will sign it very shortly, if he hasn’t signed it already. And now gun owners are at extreme risk under this law. Teddy Nappen 20:24 I just thought of another one, too. What if you’re a first time shooter and you go to a range course, you’re in a range, a gun range learning, and the gun accidentally goes off because you’re brand new to firearms? You’re now a convicted felon. No discretion. Evan Nappen 20:44 Oh, well. It was near a room. That’s right, no discretion, and anybody that has that AD. So, again, it’s designed to disenfranchise gun owners of Second Amendment rights. And by the way, you may not be Page – 5 – of 11able to then get even insurance coverage. Because if it’s criminalized over what you did, it’s not anymore. Now, you’re talking about behavior where they can claim it’s a criminal act. It’s a criminal act, okay? And again, you may depending on your policy, depending on what actually the injuries and damage, you may not even have coverage. The insurance companies will use it to deny you coverage. I’m sure of that. That’s their job, as it normally is, anyway. To try to figure out how to deny coverage. Well, they’ve just given them that ability on the civil side to further make it harder for you. It’s jeopardizing gun rights, and it’s looking at creating incarceration at felony level for gun owners. It’s outrageous, and it really is something that I’m sure we’re going to see major, major impact. And then that’s not the only fun. Go ahead, Teddy. Teddy Nappen 22:04 I was just curious on the constitutionality of it? Because they’ve made, there must be some avenue. Because it’s very, like they’re giving no discretion? And just saying. Evan Nappen 22:16 Nope. Teddy Nappen 22:16 It’s automatic. There’s no constitutional challenge. Evan Nappen 22:20 Well, I guess there could always be a constitutional challenge. But what’s going to happen is it’s going to have to be the fight. The real fight is going to be over, maybe an argument of, was there a conscious, conscientious disregard, or conscious disregard of a known risk. But the other side will argue that as soon as you have a gun with ammo, you have a known risk. I mean, a firearm, and that’s their entire anti-Second Amendment strategy. Teddy Nappen 22:43 When you deal with guns, you do so at your peril. Evan Nappen 22:53 Right! And that’s New Jersey case law, right there. So, they’re saying, hey, you do it at your peril. You took a known risk because you possessed a gun, even. You can well see a New Jersey jury buying that argument. This is nuts, and gun owners, beware, beware, beware, beware. And like I said, this isn’t the only shenanigan that occurred in Trenton. They also signed S1425. (https://pub.njleg.gov/Bills/2024/S1500/1425_R1.PDF) Now, this is actually law. This law, real quick, specifically applies just to dealers. Just to New Jersey dealers. How nice. They have their own very special law now. This law says, “A licensed dealer who sells or transfers a firearm to a person when the dealer knows or reasonably should know that person intends to sell, transfer, assign, or otherwise dispose of that firearm to a person who is disqualified from possessing a firearm under State or federal law is guilty of a crime of the second degree.” That means up to 10 years in State Prison. They have a minimum mandatory period of three and a half years, and they made it a second degree. This is insane. Page – 6 – of 11Evan Nappen 24:03 If you’re a dealer in New Jersey, they can claim that you reasonably should have known that a gun you transferred to somebody was going to be transferred to somebody who was disqualified from possessing. Let me give you an example. You sell a Red Rider BB gun. That’s a firearm under New Jersey law. And if you reasonably should have known that that person was going to let their kid have that BB gun, you’re looking at a second degree charge here, Dealers. Yeah for that BB gun. Because as long as the state can show you reasonably should know that, that the person intended to transfer it to someone who was disqualified, who would be arguably that minor, unless it’s under a strict exemption. I mean, this is the kind of pathways being cut here. How do you know or reasonably should know? What is that reasonably should have known nonsense? Evan Nappen 25:03 I mean, that’s again, 12 people on a jury are the ones who’s going to decide whether reasonably you should know. All the law says, “. . . ‘reasonably should know’ means that a person reasonably should know a fact when, under the circumstances, a person of reasonable prudence and competence would ascertain or know that fact.” Oh, that’s a that’s so crystal clear. Huh? Real, crystal clear. Now what it means is 12 people who aren’t smart enough to avoid jury duty are going to decide whether the dealer should have known on that gun sale. And if they decide otherwise, the dealer is looking at a minimum mandatory sentence on a second degree crime, which carries up to 10 years in State Prison. Okay? That’s what they’re doing. Focused on New Jersey dealers. Do you think they want to put every dealer in New Jersey out of business? I do. And that’s the other bonus law that’s actually signed into law. It’s ripe for abuse, folks. Beware. It is just atrocious what’s going on in New Jersey. Evan Nappen 26:07 Let me tell you about our fight. You know, we are in this fight. We constantly, we’ve tried to fight these things. New Jersey is an extremely tough environment. We’re going to see court challenges, even more court challenges, and it’s our state Association that’s going to be heading the fight. I’m sure we’re going to see a constitutional challenge to this so-called Accidental Discharge bill and the same over what they’re looking to do to dealers. And it’s ANJRPC, the Association of New Jersey Rifle & Pistol Clubs at the forefront, fighting for our rights. They’re the umbrella organization of gun clubs in New Jersey, and you can join as an individual member. You really need to. You’ll be sent email alerts, and you’ll be told what’s going on. And you know, we’re able to get changes made with pressure, but most importantly, our salvation seems to be in the judicial fight in the courts. The Association is there as we speak. This is an extremely tough environment in New Jersey, the toughest in America, where the oppression of Second Amendment rights is second to none. New Jersey wins the prize for Second Amendment oppression, and it’s the Association there at the forefront. You need to be a member. Go to anjrpc.org and join today. Be part of the solution. It’s really important that you do that. Evan Nappen 27:43 I’d also like to talk about our good friends at WeShoot. WeShoot is an indoor range in Lakewood, easily accessible, off the Parkway. It’s where Teddy and I both shoot, and we both qualified. It’s where we got our CCARE and where we get our training. We love WeShoot. That’s the place to shoot. It’s a place you can shoot. They have a wonderful facility, a great pro shop, and great instructors. You’ve got to check out their website, magnificent photography there. And they run all kinds of great deals and Page – 7 – of 11specials, and they have all the top state of the art equipment. Check out weshootusa.com. weshootusa.com. You’ll be glad you did. It is a great resource for us to have a range right there in Central New Jersey that is as professional and modern as WeShoot. Go to weshootusa.com and check them out. You will be thrilled, just like Teddy and I. Well, that’s where we shoot. It’s what we love. You’ll love it too. Evan Nappen 29:00 Let me also mention my book, New Jersey Gun Law. It’s the Bible of New Jersey gun law. I’m working on the update from what I just told you today. So, the free update will be coming out, including the 2026 Comprehensive Update. We’re going to look at and add in all the new laws that’ll be coming out shortly. So, if you have the book, make sure you scan the QR cover. The QR code on the cover. Join my free private subscriber base, and you’ll get notice of the updates that are forthcoming. You can buy the book at EvanNappen.com. That’s right, www.EvanNappen.com. Go to EvanNappen.com and get the big orange book today. You’ll be glad you did. It’s over 500 pages, 120 topics, all Question and Answer, designed to make it as user friendly as possible. I try to make it so you can navigate these treacherous waters of Second Amendment oppression in New Jersey. So, go to EvanNappen.com and get your book. Teddy, what do you have for us today in Press Checks? Teddy Nappen 30:15 Well, as you know, Press Checks are always free. While you’re talking about the utter insanity that is New Jersey, there’s one positive bit of news. It’s kind of been, you know, from the entire news cycle of everything they try to cover. There’s one thing that kind of slipped under the cracks that some people did pick up on. And it caught my eye. I was like, wait a second, I remember this. So, President Trump has withdrawn from the UN Register of Conventional Arms. (https://gunrights.org/united-states- withdraws-from-united-nations-register-of-conventional-arms/) That treaty. Now, I remember growing up as a kid, Dad, you told me, always keep an eye out if there are blue helmets walking down the street. Evan Nappen 31:01 Yeah, that’s right, that blue helmet day came, if that ever was to come. Yep. Teddy Nappen 31:08 And oh, I remember you telling me about that treaty. And you know that stupid, you know, the UN has always been an anti-gun organization, with that stupid, bent revolver they have. Evan Nappen 31:20 Yeah, the revolver with a barrel and a pretzel knot. (https://dam.media.un.org/archive/Gift-of- Luxembourg-to-the-United-Nations-2AM9LOQORWK.html) I mean, look at folks. It’s a revolver, by the way. It’s not an AK, you know. It’s not an AR. It’s not in an “assault firearm”. No, no. It’s a freaking revolver with a barrel in a pretzel knot there. Gee, who are the primary possessors of revolvers? I wonder. Is that paramilitary organizations? No. Terrorist, radicalized wackos? No. A revolver. Let me see. Oh, you mean, like average citizens? Wow, hmm. Interesting. Page – 8 – of 11Teddy Nappen 32:02 But what I remember that being back, you know, where this was a big fear. Where it was the giant arms treaty, where they were trying, I think it wasn’t ratified by Obama, but that was that insane policy to try, even. The UN even actually has an Office of Disarmament. (https://disarmament.unoda.org/en/our- work/conventional-arms/legal-instruments/arms-trade-treaty) That’s actually their whole like deal. What they try to push for. Now, they cloak it in like militarily. If you actually go to the website, this was from the gunrights.org. (https://gunrights.org/united-states-withdraws-from-united-nations-register-of- conventional-arms/) The National Association of Gun Rights put out the article, and they provide the link where you can go on to the UN website. You can see their register of their whole charter on the UN, and it goes into they brag about it. We’ve recorded and captured 90% of the global arms trade. By the way, this was supposed to be about, you know, tanks, armored carriers. You know, stuff used in actual, like, large scale warfare. But then I love how they do this. In 2016 they adopted the international small arms and light weapons in parallel with the other seven categories, so we can keep track of all small arms. Hmm, 2016. What were they doing to try, what was the big anti-gun push to try to disarm us around that time? Thinking that they’re going to try to go around collecting our arms in the United States. Like it’s so disgusting. I love how they just cloak it. You actually can go on to their reports. I got bored. So, I clicked the arms report of 2023 and I was like, okay, armored carriers, all that . Small arms. I wanted to look and see who were like the top buyers. So, revolvers and self-loading pistols – Iraq. Apparently. Evan Nappen 33:57 Really? Teddy Nappen 33:58 Yeah, like 2,150 pistols from us to Iraq. Evan Nappen 34:03 Oh, from the U.S.? Teddy Nappen 34:05 Yeah, from the U.S. It keeps track of each country. Evan Nappen 34:07 Well, we’re making them. Teddy Nappen 34:09 Yeah. Evan Nappen 34:09 Of course. We’re a major industrial manufacturer. What we should be doing is making guns. Teddy Nappen 34:14 Yeah. And then rifles and carbines. They separate that from “assault firearms”. Rifles and carbines. 20,000 to Israel. So, there you go for that end. Page – 9 – of 11Evan Nappen 34:27 Yeah, Israel makes a lot of their own weapons, too, and they make really good ones. Teddy Nappen 34:32 Yeah, I know they have the Hebrew hammer. Evan Nappen 34:35 Oh, yeah! Teddy Nappen 34:35 The Tabor X95. (https://iwi.us/firearms/tavor-x95/) But with the sub-machine guns, Saudi Arabia, 550. Evan Nappen 34:41 This doesn’t even matter. This is so absurd, and it’s just trying to globalize Second Amendment oppression. You know, our country’s blessed with Second Amendment. And of course, New Jersey does everything it can to undermine it, but the majority of America doesn’t do that. But internationally, we, you know, they hate us. They hate our Constitution, and they want to see us disarmed. We are standing as a threat to their globalist intentions, right? Teddy Nappen 35:21 I mean, that was the famous line that Donald Trump said to the world. The world does not belong to globalists. And that’s a fact. And here, in their charter, they even say, such measures, as they’re describing the whole disarmament office, such measures can also encourage restraint in the transfer and production of armament and decelerate military build up. In words of, okay, we need to lower the amount of guns in the world and try to disarm the people. That’s the cover they run, but they dress it up. I will give the Left credit. Their ability to wordsmith their way into something else is crazy. Evan Nappen 36:06 Well, listen, man. It’s not every political group that can convince people, you know, that a man can be a woman. So, why can’t they convince the world about this with guns? Right? Teddy Nappen 36:17 Well, it’s the political group that has the. When they did the whole study on mental health of different groups, the vast majority of people that vote Democrat have mental illness. So, let that sink in. That was an actual study, and that was put out by, like, CNN! So. Evan Nappen 36:18 Really? Teddy Nappen 36:19 Yeah, they had to be like. No, I love it. If you are ever bored? Anyone who’s very bored, go on to CNN and catch Harry Enton, the statistics guy. He’s the golden retriever of CNN. He just talks about numbers, and he gets so excited. He’s like, oh my God, have you seen these numbers? I can’t believe Page – 10 – of 11it. He’s always, like, shocked every time. He sees like, you know, everyone keeps saying Trump’s numbers are going bad, but you go over to here. Six months ago, 84, and now, it’s 85. Oh, wow, amazing. Like, it’s just, it’s that energy. It’s crazy. Evan Nappen 37:13 Well, how old is he? Maybe he’s just trying to get excitement to statistics? Teddy Nappen 37:18 I know, but it’s just like, what are the numbers? Pretty good. He’s like, gad Zooks. He’s like, clapping. I know. It’s just like, what the heck is it? Like if anyone is bored? Just look up Harry Enten on CNN. He’s, it’s so fucking weird. Evan Nappen 37:37 Okay, I love it. All right, Teddy. Well, that is interesting to know, but I’m not surprised, not surprised at all. This is the moment, the moment when we discuss the GOFU, that is the Gun Owner Fuck Up. It is one of the most important aspects of what we do, because every day we deal with Gun Owner Fuck Ups. And when we can let the listeners know, you get to learn expensive lessons for free. And this week’s GOFU is real simple. It’s Accidental Discharge. Let me just make it real clear. Now, more than ever, more than ever, you’ve got to be extremely overly conscientious. You better triple check chambers. You’ve got to make sure. You cannot afford in any way to have any kind of Accidental Discharge in New Jersey, because you risk it all. You risk it all. You risk becoming a felon. You risk going to prison. You risk losing your gun rights for the entire United States. You risk not being covered, arguably, by insurance. It is an insane risk that New Jersey is imposing, and I’ve seen 80 cases throughout my entire practice. Unfortunately, they happen, and, you know, in hindsight, they’re all avoidable. But folks don’t be a GOFU. Please, please, please. Follow all the rules of safety, and make sure you treat every gun as loaded. Every gun, you treat as loaded. Do not for a second, not do that. It’s just that critical. They’re criminalizing those who make a simple mistake, and there is no tolerance. Evan Nappen 40:00 This is Evan Nappen and Teddy Nappen reminding you that gun laws don’t protect honest citizens from criminals. They protect criminals from honest citizens. Speaker 2 40:13 Gun Lawyer is a CounterThink Media production. The music used in this broadcast was managed by Cosmo Music, New York, New York. Reach us by emailing Evan@gun.lawyer. The information and opinions in this broadcast do not constitute legal advice. Consult a licensed attorney in your state. Page – 11 – of 11 Downloadable PDF TranscriptGun Lawyer S5 E273_Transcript About The HostEvan Nappen, Esq.Known as “America's Gun Lawyer,” Evan Nappen is above all a tireless defender of justice. Author of eight bestselling books and countless articles on firearms, knives, and weapons history and the law, a certified Firearms Instructor, and avid weapons collector and historian with a vast collection that spans almost five decades — it's no wonder he's become the trusted, go-to expert for local, industry and national media outlets. Regularly called on by radio, television and online news media for his commentary and expertise on breaking news Evan has appeared countless shows including Fox News – Judge Jeanine, CNN – Lou Dobbs, Court TV, Real Talk on WOR, It's Your Call with Lyn Doyle, Tom Gresham's Gun Talk, and Cam & Company/NRA News. As a creative arts consultant, he also lends his weapons law and historical expertise to an elite, discerning cadre of movie and television producers and directors, and novelists. He also provides expert testimony and consultations for defense attorneys across America. Email Evan Your Comments and Questions  talkback@gun.lawyer Join Evan's InnerCircleHere's your chance to join an elite group of the Savviest gun and knife owners in America.  Membership is totally FREE and Strictly CONFIDENTIAL.  Just enter your email to start receiving insider news, tips, and other valuable membership benefits.   Email (required) *First Name *Select list(s) to subscribe toInnerCircle Membership Yes, I would like to receive emails from Gun Lawyer Podcast. (You can unsubscribe anytime)Constant Contact Use. Please leave this field blank.var ajaxurl = "https://gun.lawyer/wp-admin/admin-ajax.php";

Cracked Racquets Editor-in-Chief Alex Gruskin previews the 2026 Australian Open Men's Singles Draw. He runs through each quarter of the draw and discusses the best R1 and potential matchups. He also offers his thoughts on the winners and losers of the draw configuration, shares his predictions for how he sees things unfolding, plus SO much more!! Don't forget to give a 5 star review on your favorite podcast app! In addition, add your twitter/instagram handle to the review for a chance to win some FREE CR gear!! Episode Bookmarks Alcaraz (1) Quarter - 4:45 Best R1 Matchups - 8:21 Best Prospective Matchups - 11:17 Sleepers in the section - 12:30 Projecting the best storylines to emerge - 14:25 Who likes this draw most - 17:01 Who likes this draw least - 17:35 Prediction - 19:20 Zverev (3) Quarter - 20:17 Best R1 Matchups - 21:55 Best Prospective Matchups - 24:33 Sleepers in the section - 26:03 Projecting the best storylines to emerge - 28:20 Who likes this draw most - 29:11 Who likes this draw least - 30:18 Prediction - 31:48 Djokovic (4) Quarter - 32:56 Best R1 Matchups - 34:22 Best Prospective Matchups - 37:15 Sleepers in the section - 39:20 Projecting the best storylines to emerge - 41:00 Who likes this draw most - 44:00 Who likes this draw least - 45:25 Prediction - 46:38 Sinner (2) Quarter - 48:09 Best R1 Matchups - 49:28 Best Prospective Matchups - 51:02 Sleepers in the section - 52:30 Projecting the best storylines to emerge - 53:30 Who likes this draw most - 54:15 Who likes this draw least - 56:28 Prediction - 56:51 Final predictions - 57:22 _____ Laurel Springs Ranked among the best online private schools in the United States, Laurel Springs stands out when it comes to support, personalization, community, and college prep. They give their K-12 students the resources, guidance, and learning opportunities they need at each grade level to reach their full potential. Find Cracked Racquets Website: https://www.crackedracquets.com Instagram: https://instagram.com/crackedracquets Twitter: https://twitter.com/crackedracquets Facebook: https://Facebook.com/crackedracquets YouTube: https://www.youtube.com/c/crackedracquets Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

In this episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, Eric J. Price, Chief Financial Officer at Schoolcraft Memorial Hospital, discusses navigating economic, political, and financial uncertainties, incorporating AI to improve efficiency while remaining cautious, the impact of AI on labor, and strategies for reducing inefficiencies in the healthcare system.In collaboration with R1

Cracked Racquets Editor-in-Chief Alex Gruskin previews the 2026 Australian Open Women's Singles Draw. He runs through each quarter of the draw and discusses the best R1 and potential matchups. He also offers his thoughts on the winners and losers of the draw configuration, shares his predictions for how he sees things unfolding, plus SO much more!! Don't forget to give a 5 star review on your favorite podcast app! In addition, add your twitter/instagram handle to the review for a chance to win some FREE CR gear!! Episode Bookmarks Sabalenka (1) Quarter - 5:01 Best R1 Matchups - 8:00 Best Prospective Matchups - 10:01 Sleepers in the section - 11:29 Projecting the best storylines to emerge - 14:25 Who likes this draw most - 15:38 Who likes this draw least - 17:06 Prediction - 17:47 Gauff (3) Quarter - 18:25 Best R1 Matchups - 20:06 Best Prospective Matchups - 22:44 Sleepers in the section - 25:00 Projecting the best storylines to emerge - 28:04 Who likes this draw most - 30:02 Who likes this draw least - 31:32 Prediction - 32:45 Anisimova (4) Quarter - 34:10 Best R1 Matchups - 36:28 Best Prospective Matchups - 37:52 Sleepers in the section - 39:04 Projecting the best storylines to emerge - 41:05 Who likes this draw most - 42:19 Who likes this draw least - 43:34 Prediction - 45:38 Swiatek (2) Quarter - 46:15 Best R1 Matchups - 48:00 Best Prospective Matchups - 49:12 Sleepers in the section - 50:20 Projecting the best storylines to emerge - 51:47 Who likes this draw most - 52:39 Who likes this draw least - 54:00 Prediction - 54:40 Final predictions - 55:20 _____ Laurel Springs Ranked among the best online private schools in the United States, Laurel Springs stands out when it comes to support, personalization, community, and college prep. They give their K-12 students the resources, guidance, and learning opportunities they need at each grade level to reach their full potential. Find Cracked Racquets Website: https://www.crackedracquets.com Instagram: https://instagram.com/crackedracquets Twitter: https://twitter.com/crackedracquets Facebook: https://Facebook.com/crackedracquets YouTube: https://www.youtube.com/c/crackedracquets Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

In this episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, Kevin M. Spiegel, President & CEO of HSA Florida Medical Center, discusses how healthcare organizations can leverage technological advancements to improve scheduling, streamline operations, and enhance patient experience. He dives into the governance of AI within hospital systems, sharing insights on balancing innovation with ethical and operational considerations.In collaboration with R1.

This episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, features Michelle Joy, President and CEO of Carson Tahoe Health. She shares strategies for expanding access to care in rural Nevada, enhancing workforce engagement, and implementing AI technology to improve clinical efficiency and revenue cycle operations while maintaining strong community trust.In collaboration with R1.

This episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, features Lynn Fulton, CEO of Maui Health System. She discusses the unique challenges of providing care across Maui and Lanai, including workforce recruitment, technology integration, and payer partnerships, while highlighting strategies to enhance operational efficiency and community-centered care.In collaboration with R1.

This episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, features Jochen Reiser, MD PhD, President and CEO, UTMB. He discusses how UTMB is embedding innovation and AI across research, education, and clinical care, creating new partnerships, advancing governance, and shaping the future of academic medicine.In collaboration with R1.

This episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, features James Heilsberg, CFO, Tri State Health. He discusses how his organization is using AI, automation, and strategic growth to strengthen care delivery, support clinicians, and sustain financial health in a rural community setting.In collaboration with R1.

This episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, features Amy Lee, MJ, MBA, MBHA, FACMPE, President and COO of Nantucket Cottage Hospital/Mass General Brigham. She discusses the unique challenges of delivering care on an island, from workforce housing to telehealth partnerships, and how innovation and community support sustain high-quality care in a remote setting.In collaboration with R1.

This episode, recorded live at the Becker's 13th Annual CEO + CFO Roundtable, features David Dunkle, CEO, Johnson Memorial Health. He discusses the financial challenges facing community hospitals, the struggle for fair reimbursement, and how strong culture and patient focus help sustain independent organizations in a difficult healthcare landscape.In collaboration with R1.