Podcasts about Inference

In this episode, the VENDO team is joined by Andrew Bell, Amazon Lead for NFPA, to explore Amazon Rufus and the future of AI-driven shopping. We unpack the evolution of Rufus, its impact on new brands, and the growing role of inference, visual search, and agentic evaluation. Tune in to learn how brands can adapt and thrive in the next era of intelligent commerce. Topics Covered: Amazon Rufus Context & Evolution (3:00) Does Rufus Affect New Brands? (9:30) The Importance of Inference (10:38) Brands Optimizing for Inference (15:32) AI Shopping (17:06) Lens AI and Visual Search (22:00) Agentic Evaluation (26:19) Speakers: Andrew Bell, Amazon Lead, NFPA Delaney Del Mundo, VP Account Strategy - Amazon & TikTok Shop, VENDO Want to stay up to date on topics like this? Subscribe to our Amazon & Walmart Growth #podcast for bi-weekly episodes every other Thursday! ➡️ YouTube: https://www.youtube.com/channel/UCr2VTsj1X3PRZWE97n-tDbA ➡️ Spotify: https://open.spotify.com/show/4HXz504VRToYzafHcAhzke?si=9d57599ed19e4362 ➡️ Apple: https://podcasts.apple.com/us/podcast/vendo-amazon-walmart-growth-experts/id1512362107

Nikola Borisov, CEO and co-founder of Deep Infra, joins the show to unpack the rapid evolution of AI inference, the hardware race powering it, and how startups can actually keep up without burning out. From open source breakthroughs to the business realities of model selection, Nikola shares why speed, efficiency, and strategic focus matter more than ever. If you're building in AI, this conversation will help you see the road ahead more clearly.Key Takeaways• Open source AI models are advancing at a pace that forces founders to choose focus over chasing every release.• First mover advantage in AI is real but plays out differently than in consumer tech because models are often black boxes to end users.• Infrastructure and hardware strategy can make or break AI product delivery, especially for startups.• Efficient inference may become more important than efficient training as AI usage scales.• Optimizing for specific customer needs can create significant performance and cost advantages.Timestamped Highlights[02:12] How far AI has come — and why we're still under 10% of its future potential[04:11] The challenge of keeping pace with constant model releases[08:12] Why differentiation between models still matters for builders[14:08] The hidden costs and strategies of AI hardware infrastructure[18:05] Why inference efficiency could eclipse training efficiency[21:46] Lessons from missed opportunities and unexpected shifts in model innovationQuote of the Episode“Being more efficient at inference is going to be way more important than being very efficient at training.” — Nikola BorisovResources MentionedDeepInfra — https://deepinfra.comNikola Borisov on LinkedIn — https://www.linkedin.com/in/nikolabCall to ActionIf you enjoyed this conversation, share it with someone building in AI and subscribe so you never miss an episode. Your next big idea might just come from the next one.

In this episode, we're joined by Lin Qiao, CEO and co-founder of Fireworks AI. Drawing on key lessons from her time building PyTorch, Lin shares her perspective on the modern generative AI development lifecycle. She explains why aligning training and inference systems is essential for creating a seamless, fast-moving production pipeline, preventing the friction that often stalls deployment. We explore the strategic shift from treating models as commodities to viewing them as core product assets. Lin details how post-training methods, like reinforcement fine-tuning (RFT), allow teams to leverage their own proprietary data to continuously improve these assets. Lin also breaks down the complex challenge of what she calls "3D optimization"—balancing cost, latency, and quality—and emphasizes the role of clear evaluation criteria to guide this process, moving beyond unreliable methods like "vibe checking." Finally, we discuss the path toward the future of AI development: designing a closed-loop system for automated model improvement, a vision made more attainable by the exciting convergence of open and closed-source model capabilities. The complete show notes for this episode can be found at https://twimlai.com/go/742.

Join us on Discord with Semiconductor Insider, sign up on our website: www.chipstockinvestor.com/membershipSupercharge your analysis with AI! Get 15% of your membership with our special link here: https://fiscal.ai/csi/Investors are excited about the prospects for AI inference, and potential "next Nvidia" investments. Is AMD really going to be the biggest winner from this market? There are actually other candidates, like Astera Labs (ALAB) and Credo Technology (CRDO) that need to be considered. And of course, a big data center AI inference winner is none other than Arista Networks (ANET). Chip Stock Investors Nick and Kasey break it down in this important video update.Sign Up For Our Newsletter: https://mailchi.mp/b1228c12f284/sign-up-landing-page-short-formPrevious Arista videos referenced:https://youtu.be/gyfRB8E0p6ohttps://youtu.be/OuMuLBVcb84Astera Labs Video:https://youtu.be/jZyHWqBXDo8********************************************************Affiliate links that are sprinkled in throughout this video. If something catches your eye and you decide to buy it, we might earn a little coffee money. Thanks for helping us (Kasey) fuel our caffeine addiction!Content in this video is for general information or entertainment only and is not specific or individual investment advice. Forecasts and information presented may not develop as predicted and there is no guarantee any strategies presented will be successful. All investing involves risk, and you could lose some or all of your principal. #amd #alab #anet #gpus #aiinference #arista #asteralabs #semiconductors #chips #investing #stocks #finance #financeeducation #silicon #artificialintelligence #ai #financeeducation #chipstocks #finance #stocks #investing #investor #financeeducation #stockmarket #chipstockinvestor #fablesschipdesign #chipmanufacturing #semiconductormanufacturing #semiconductorstocks Timestamps:(00:00) AMD's Recent Performance and Market Position(01:58) AMD's Financials and Future Prospects(07:17) Astera Labs: Networking Innovations(15:12) Credo: A Strategic Investment(17:12) Arista Networks: A Growth Story(26:36) Conclusion Nick and Kasey own shares of AMD, ANET

While AI training garners most of the spotlight — and investment — the demands ofAI inferenceare shaping up to be an even bigger challenge. In this episode ofThe New Stack Makers, Sid Sheth, founder and CEO of d-Matrix, argues that inference is anything but one-size-fits-all. Different use cases — from low-cost to high-interactivity or throughput-optimized — require tailored hardware, and existing GPU architectures aren't built to address all these needs simultaneously.“The world of inference is going to be truly heterogeneous,” Sheth said, meaning specialized hardware will be required to meet diverse performance profiles. A major bottleneck? The distance between memory and compute. Inference, especially in generative AI and agentic workflows, requires constant memory access, so minimizing the distance data must travel is key to improving performance and reducing cost.To address this, d-Matrix developed Corsair, a modular platform where memory and compute are vertically stacked — “like pancakes” — enabling faster, more efficient inference. The result is scalable, flexible AI infrastructure purpose-built for inference at scale.Learn more from The New Stack about inference compute and AIScaling AI Inference at the Edge with Distributed PostgreSQLDeep Infra Is Building an AI Inference Cloud for DevelopersJoin our community of newsletter subscribers to stay on top of the news and at the top of your game  

Live from the 2025 Nonference, Mike and Tim (In the same room) are joined in studio by Journey Church Pastors Suzie P. Lind and Sam Barnhart. What does it mean to truly deconstruct faith, and how can that journey lead to healing? In this heartfelt and thought-provoking conversation, the hosts tackle the complexities of "deconstruction," exploring disillusionment, doubt, discipleship, and ultimately, the pursuit of Jesus amidst cultural challenges. From addressing church hurt and systemic issues to reexamining theologies and navigating the intersection of faith and politics, this episode unpacks the role of the church in society and the personal journeys that shape our understanding of Christianity. Through themes of justice, cruciformity, and reimagining what it means to follow Jesus, the discussion dives deep into how cultural realities and historical practices influence our faith. The panel shares stories of heartbreak and hope, challenging the idea that questioning or rethinking faith is a departure from Jesus—instead, it's often a move toward deeper authenticity. Whether you're wrestling with theological questions, processing church trauma, or striving to navigate cultural issues as a follower of Jesus, this episode offers a space for reflection and community. Feel free to share your thoughts, send in your questions, or engage with us on Facebook and Instagram. Let's continue pursuing a faith marked by humility, curiosity, and justice together. CHAPTERS: 00:00 - Welcome to the Nonference 02:12 - The Tennessee Buzz 04:35 - Deconstruction: A Second Innocence 07:11 - The Six D's of Deconstruction 14:46 - Why People Are Disillusioned 18:18 - Did the Church Move or Did the Curtain Open 23:16 - Deconstruction as Repentance 28:32 - Discipleship in Deconstruction 29:41 - Understanding Deconversion 32:44 - Redefinition in Faith 34:58 - Navigating Doubt 38:50 - Biblical Foundations of Deconstruction 41:00 - Purpose of Inference 42:26 - Q&A: Insights from Stafford 49:49 - National Park Moments 51:09 - Experiencing Death and Grief 56:32 - Neuroscience of Belief 56:41 - Josh McDowell and the Talking Snake 1:02:40 - Embracing the Power of Weakness 1:03:12 - Thank You 1:04:08 - Credits As always, we encourage and would love discussion as we pursue. Feel free to email in questions to hello@voxpodcast.com, and to engage the conversation on Facebook and Instagram. We're on YouTube (if you're into that kinda thing): VOXOLOGY TV. Our Merch Store! ETSY Learn more about the Voxology Podcast Subscribe on iTunes or Spotify Support the Voxology Podcast on Patreon The Voxology Spotify channel can be found here: Voxology Radio Follow us on Instagram: @voxologypodcast and "like" us on Facebook Follow Mike on Twitter: www.twitter.com/mikeerre Music in this episode by Timothy John Stafford Instagram & Twitter: @GoneTimothy

If you've been experimenting with image, video, and audio models, the chances are you've been both blown away by how good they're becoming, and also a little perturbed by how long they can take to generate. If you've been using a platform like Fal, however, your experience on the latter point might be more positive.In this episode, Fal cofounder and CEO Burkay Gur and head of engineering Batuhan Taskaya join a16z general partner Jennifer Li to discuss how they built an inference platform — or, as they call it, a generative media cloud — that's optimized for speed, performance, and user experience. These are core features for a great product, yes, and also ones borne of necessity as the early team obsessively engineered around its meager GPU capacity at the height of the AI infrastructure crunch.But this is more than a story about infrastructure. As you'll hear, they also delve into sales and hiring strategy; the team's overall excitement over these emerging modalities; and the trends they're seeing as competition in the world of video models, especially, heats up.  Check out everything a16z is doing with artificial intelligence here, including articles, projects, and more podcasts.

Modal is a serverless compute platform that's specifically focused on AI workloads. The company's goal is to enable AI teams to quickly spin up GPU-enabled containers, and rapidly iterate and autoscale. It was founded by Erik Bernhardsson who was previously at Spotify for 7 years where he built the music recommendation system and the popular The post Modal and Scaling AI Inference with Erik Bernhardsson appeared first on Software Engineering Daily.

Modal is a serverless compute platform that’s specifically focused on AI workloads. The company's goal is to enable AI teams to quickly spin up GPU-enabled containers, and rapidly iterate and autoscale. It was founded by Erik Bernhardsson who was previously at Spotify for 7 years where he built the music recommendation system and the popular The post Modal and Scaling AI Inference with Erik Bernhardsson appeared first on Software Engineering Daily.

Many students graduate high school today without having read a book cover to cover. Many students struggle to learn to read at all. How did this happen? Listen as educator and author Doug Lemov talks with EconTalk's Russ Roberts about the failed fads in reading education, the mistaken emphasis on vocabulary as a skill, and the importance of background knowledge for thinking and reading comprehension. Lemov and Roberts also discuss their love of difficult-to-read authors, the power of reading in groups, the value of keeping a reading journal, and how even basketball can be more enjoyable when we have the right terminology.

Fal.ai, once focused on machine learning infrastructure, has evolved into a major player in generative media. In this episode of The New Stack Agents, hosts speak with Fal.ai CEO Burkay Gur and investor Glenn Solomon of Notable Capital. Originally aiming to optimize Python runtimes, Fal.ai shifted direction as generative AI exploded, driven by tools like DALL·E and ChatGPT. Today, Fal.ai hosts hundreds of models—from image to audio and video—and emphasizes fast, optimized inference to meet growing demand.Speed became Fal.ai's competitive edge, especially as newer generative models require GPU power not just for training but also for inference. Solomon noted that while optimization alone isn't a sustainable business model, Fal's value lies in speed and developer experience. Fal.ai offers both an easy-to-use web interface and developer-focused APIs, appealing to both technical and non-technical users.Gur also addressed generative AI's impact on creatives, arguing that while the cost of creation has plummeted, the cost of creativity remains—and may even increase as content becomes easier to produce.Learn more from The New Stack about AI's impact on creatives:AI Will Steal Developer Jobs (But Not How You Think) How AI Agents Will Change the Web for Users and Developers Join our community of newsletter subscribers to stay on top of the news and at the top of your game. 

Send us a textIn this episode of Embedded Insiders, Rich and I sit down with Sid Sheth, CEO and co-founder of d-Matrix, to explore the ongoing generative AI boom—why it's becoming increasingly unsustainable, and how d-Matrix is addressing the challenge with a chiplet-based compute architecture built specifically for AI inference.Next, Ken brings us up to speed on some of the week's top embedded industry headlines, with updates from ASUS IoT, LG, and Microelectronics UK.But first, Rich, Ken, and I share our thoughts on the state of generative AI and AI inference. For more information, visit embeddedcomputing.com

Series: Bible Class 2025 - AuthorityService: Bible Study - SundayType: Bible ClassSpeaker: Ralph Walker

Colab is cozy. But production won't fit on a single GPU. Zach Mueller leads Accelerate at Hugging Face and spends his days helping people go from solo scripts to scalable systems. In this episode, he joins me to demystify distributed training and inference — not just for research labs, but for any ML engineer trying to ship real software. We talk through: • From Colab to clusters: why scaling isn't just about training massive models, but serving agents, handling load, and speeding up iteration • Zero-to-two GPUs: how to get started without Kubernetes, Slurm, or a PhD in networking • Scaling tradeoffs: when to care about interconnects, which infra bottlenecks actually matter, and how to avoid chasing performance ghosts • The GPU middle class: strategies for training and serving on a shoestring, with just a few cards or modest credits • Local experiments, global impact: why learning distributed systems—even just a little—can set you apart as an engineer If you've ever stared at a Hugging Face training script and wondered how to run it on something more than your laptop: this one's for you. LINKS Zach on LinkedIn (https://www.linkedin.com/in/zachary-mueller-135257118/) Hugo's blog post on Stop Buliding AI Agents (https://www.linkedin.com/posts/hugo-bowne-anderson-045939a5_yesterday-i-posted-about-stop-building-ai-activity-7346942036752613376-b8-t/) Upcoming Events on Luma (https://lu.ma/calendar/cal-8ImWFDQ3IEIxNWk) Hugo's recent newsletter about upcoming events and more! (https://hugobowne.substack.com/p/stop-building-agents)

Hey everyone, Alex here

This crossover episode from Inference by Turing Post features CEO Dev Rishi of Predibase discussing the shift from static to continuously learning AI systems that can adapt and improve from ongoing user feedback in production. Rishi provides grounded insights from deploying these dynamic models to real enterprise customers in healthcare and finance, exploring both the massive potential upside and significant safety challenges of reinforcement learning at scale. The conversation examines how "practical specialized intelligence" could reshape the AI landscape by filling economic niches efficiently, potentially offering a more stable alternative to AGI development. This discussion bridges theoretical concepts with real-world deployment experience, offering a practical preview of AI systems that "train once and learn forever." Turing Post channel:  @RealTuringPost  Turpin Post website: https://www.turingpost.com Sponsors: Google Gemini 2.5 Flash : Build faster, smarter apps with customizable reasoning controls that let you optimize for speed and cost. Start building at https://aistudio.google.com Labelbox: Labelbox pairs automation, expert judgment, and reinforcement learning to deliver high-quality training data for cutting-edge AI. Put its data factory to work for you, visit https://labelbox.com Oracle Cloud Infrastructure: Oracle Cloud Infrastructure (OCI) is the next-generation cloud that delivers better performance, faster speeds, and significantly lower costs, including up to 50% less for compute, 70% for storage, and 80% for networking. Run any workload, from infrastructure to AI, in a high-availability environment and try OCI for free with zero commitment at https://oracle.com/cognitive The AGNTCY: The AGNTCY is an open-source collective dedicated to building the Internet of Agents, enabling AI agents to communicate and collaborate seamlessly across frameworks. Join a community of engineers focused on high-quality multi-agent software and support the initiative at https://agntcy.org NetSuite by Oracle: NetSuite by Oracle is the AI-powered business management suite trusted by over 42,000 businesses, offering a unified platform for accounting, financial management, inventory, and HR. Gain total visibility and control to make quick decisions and automate everyday tasks—download the free ebook, Navigating Global Trade: Three Insights for Leaders, at https://netsuite.com/cognitive PRODUCED BY: https://aipodcast.ing CHAPTERS: (00:00) Sponsor: Google Gemini 2.5 Flash (00:31) About the Episode (03:46) Training Models Continuously (05:03) Reinforcement Fine-Tuning Revolution (09:31) Agentic Workflows Challenges (Part 1) (12:51) Sponsors: Labelbox | Oracle Cloud Infrastructure (15:28) Agentic Workflows Challenges (Part 2) (15:41) ChatGPT Pivot Moment (19:59) Planning AI Future (24:45) Open Source Gaps (Part 1) (28:35) Sponsors: The AGNTCY | NetSuite by Oracle (30:50) Open Source Gaps (Part 2) (30:54) AGI vs Specialized (35:26) Happiness and Success (37:04) Outro

Send us a textCyber Asset Assessment: Understanding the Importance of SamplingIn this episode, I dive into the crucial step of sampling in cyber asset assessment. Learn why sampling is essential, especially when dealing with large environments and limited resources. Discover the various types of sampling methods, including probability and non-probability sampling, and understand how to statistically correlate your sample size to the total population of your cyber assets. Perfect for anyone looking to efficiently and effectively assess their organization's cyber assets.00:00 Introduction to Cyber Asset Assessment00:26 Understanding Sampling in Large Environments01:23 Statistical Ties and Inference in Sampling02:30 Why Sampling is Essential03:12 Types of Sampling Methods04:25 Implementing Non-Probability Sampling05:32 Final Thoughts on Sampling

Can probability theory help explain how the mind works? In this episode of Deep Dives with Iman, host Iman Mossavat talks with Dr. Hadi Vafaii, a postdoctoral scholar at UC Berkeley's Redwood Center for Theoretical Neuroscience, working in Jacob Yates's lab. The conversation focuses on “perception as inference”—a century-old idea that continues to influence modern neuroscience, from predictive coding to the free energy principle. While rooted in probability theory and Bayesian statistics, Hadi builds intuition step by step, using simple, relatable examples. He explains how the brain interprets the world by guessing the hidden causes behind what we see and hear, combining prior knowledge with new evidence. The episode also explores how personal beliefs and life experiences shape interpretation and why uncertainty is a vital part of life rather than a weakness.

Proudly sponsored by PyMC Labs, the Bayesian Consultancy. Book a call, or get in touch!Intro to Bayes Course (first 2 lessons free)Advanced Regression Course (first 2 lessons free)Our theme music is « Good Bayesian », by Baba Brinkman (feat MC Lars and Mega Ran). Check out his awesome work!Visit our Patreon page to unlock exclusive Bayesian swag ;)Takeaways:INLA is a fast, deterministic method for Bayesian inference.INLA is particularly useful for large datasets and complex models.The R INLA package is widely used for implementing INLA methodology.INLA has been applied in various fields, including epidemiology and air quality control.Computational challenges in INLA are minimal compared to MCMC methods.The Smart Gradient method enhances the efficiency of INLA.INLA can handle various likelihoods, not just Gaussian.SPDs allow for more efficient computations in spatial modeling.The new INLA methodology scales better for large datasets, especially in medical imaging.Priors in Bayesian models can significantly impact the results and should be chosen carefully.Penalized complexity priors (PC priors) help prevent overfitting in models.Understanding the underlying mathematics of priors is crucial for effective modeling.The integration of GPUs in computational methods is a key future direction for INLA.The development of new sparse solvers is essential for handling larger models efficiently.Chapters:06:06 Understanding INLA: A Comparison with MCMC08:46 Applications of INLA in Real-World Scenarios11:58 Latent Gaussian Models and Their Importance15:12 Impactful Applications of INLA in Health and Environment18:09 Computational Challenges and Solutions in INLA21:06 Stochastic Partial Differential Equations in Spatial Modeling23:55 Future Directions and Innovations in INLA39:51 Exploring Stochastic Differential Equations43:02 Advancements in INLA Methodology50:40 Getting Started with INLA56:25 Understanding Priors in Bayesian ModelsThank you to my Patrons for making this episode possible!Yusuke Saito, Avi Bryant, Ero Carrera, Giuliano Cruz, James Wade, Tradd Salvo, William Benton, James Ahloy, Robin Taylor,, Chad

Kai Wang joins the MLOps Community podcast LIVE to share how Uber built and scaled its ML platform, Michelangelo. From mission-critical models to tools for both beginners and experts, he walks us through Uber's AI playbook—and teases plans to open-source parts of it.// BioKai Wang is the product lead of the AI platform team at Uber, overseeing Uber's internal end-to-end ML platform called Michelangelo that powers 100% Uber's business-critical ML use cases.// Related LinksUber GenAI: https://www.uber.com/blog/from-predictive-to-generative-ai/#uber #podcast #ai #machinelearning ~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExploreMLOps Swag/Merch: [https://shop.mlops.community/]Connect with Demetrios on LinkedIn: /dpbrinkmConnect with Kai on LinkedIn: /kai-wang-67457318/Timestamps:[00:00] Rethinking AI Beyond ChatGPT[04:01] How Devs Pick Their Tools[08:25] Measuring Dev Speed Smartly[10:14] Predictive Models at Uber[13:11] When ML Strategy Shifts[15:56] Smarter Uber Eats with AI[19:29] Summarizing Feedback with ML[23:27] GenAI That Users Notice[27:19] Inference at Scale: Michelangelo[32:26] Building Uber's AI Studio[33:50] Faster AI Agents, Less Pain[39:21] Evaluating Models at Uber[42:22] Why Uber Open-Sourced Machanjo[44:32] What Fuels Uber's AI Team

The era of making AI smarter just by making it bigger is ending. But that doesn't mean progress is slowing down — far from it. AI models continue to get much more powerful, just using very different methods, and those underlying technical changes force a big rethink of what coming years will look like.Toby Ord — Oxford philosopher and bestselling author of The Precipice — has been tracking these shifts and mapping out the implications both for governments and our lives.Links to learn more, video, highlights, and full transcript: https://80k.info/to25As he explains, until recently anyone can access the best AI in the world “for less than the price of a can of Coke.” But unfortunately, that's over.What changed? AI companies first made models smarter by throwing a million times as much computing power at them during training, to make them better at predicting the next word. But with high quality data drying up, that approach petered out in 2024.So they pivoted to something radically different: instead of training smarter models, they're giving existing models dramatically more time to think — leading to the rise in “reasoning models” that are at the frontier today.The results are impressive but this extra computing time comes at a cost: OpenAI's o3 reasoning model achieved stunning results on a famous AI test by writing an Encyclopedia Britannica's worth of reasoning to solve individual problems at a cost of over $1,000 per question.This isn't just technical trivia: if this improvement method sticks, it will change much about how the AI revolution plays out, starting with the fact that we can expect the rich and powerful to get access to the best AI models well before the rest of us.Toby and host Rob discuss the implications of all that, plus the return of reinforcement learning (and resulting increase in deception), and Toby's commitment to clarifying the misleading graphs coming out of AI companies — to separate the snake oil and fads from the reality of what's likely a "transformative moment in human history."Recorded on May 23, 2025.Chapters:Cold open (00:00:00)Toby Ord is back — for a 4th time! (00:01:20)Everything has changed (and changed again) since 2020 (00:01:37)Is x-risk up or down? (00:07:47)The new scaling era: compute at inference (00:09:12)Inference scaling means less concentration (00:31:21)Will rich people get access to AGI first? Will the rest of us even know? (00:35:11)The new regime makes 'compute governance' harder (00:41:08)How 'IDA' might let AI blast past human level — or not (00:50:14)Reinforcement learning brings back 'reward hacking' agents (01:04:56)Will we get warning shots? Will they even help? (01:14:41)The scaling paradox (01:22:09)Misleading charts from AI companies (01:30:55)Policy debates should dream much bigger (01:43:04)Scientific moratoriums have worked before (01:56:04)Might AI 'go rogue' early on? (02:13:16)Lamps are regulated much more than AI (02:20:55)Companies made a strategic error shooting down SB 1047 (02:29:57)Companies should build in emergency brakes for their AI (02:35:49)Toby's bottom lines (02:44:32)Tell us what you thought! https://forms.gle/enUSk8HXiCrqSA9J8Video editing: Simon MonsourAudio engineering: Ben Cordell, Milo McGuire, Simon Monsour, and Dominic ArmstrongMusic: Ben CordellCamera operator: Jeremy ChevillotteTranscriptions and web: Katy Moore

The Blueprint Show - Unlocking the Future of E-commerce with AI   Summary   In this episode of Seller Sessions, Danny McMillan and Andrew Joseph Bell explore the intersection of AI and e-commerce, with a focus on Amazon's technological advancements. They examine Amazon science papers versus patents, discuss challenges with large language models, and highlight the importance of semantic intent in product recommendations. The conversation explores the evolution from keyword optimization to understanding customer purchase intentions, showcasing how AI tools like Rufus are transforming the shopping experience. The hosts provide practical strategies for sellers to optimize listings and harness AI for improved product visibility and sales.   Key Takeaways     Amazon science papers predict future e-commerce trends.       AI integration is accelerating in Amazon's ecosystem.       Understanding semantic intent is crucial for product recommendations.       The shift from keywords to purchase intentions is significant.       Rufus enhances the shopping experience with AI planning capabilities.       Sellers should focus on customer motivations in their listings.       Creating compelling product content is essential for visibility.       Custom GPTs can optimize product listings effectively.       Inference pathways help align products with customer goals.       Asking the right questions is key to leveraging AI effectively.     Sound Bites     "Understanding semantic intent is crucial."       "You can bend AI to your will."       "Asking the right questions opens doors."     Chapters   00:00 Introduction to Seller Sessions and New Season   00:33 Exploring Amazon Science Papers vs. Patents   01:27 Understanding Rufus and AI in E-commerce   02:52 Challenges in Large Language Models and Product Recommendations   07:09 Research Contributions and Implications for Sellers   10:31 Strategies for Leveraging AI in Product Listings   12:42 The Future of Shopping with AI and Amazon's Innovations   16:14 Practical Examples: Using AI for Product Optimization   22:29 Building Tools for Enhanced E-commerce Experiences   25:38 Product Naming and Features Exploration   27:44 Understanding Inference Pathways in Product Descriptions   30:36 Building Tools for AI Prompting and Automation   38:58 Bending AI to Your Will: Creativity and Imagination   48:10 Practical Applications of AI in Business Automation

Anna Patterson is the cofounder of Ceramic, an AI infrastructure platform for large scale model training. They raised their seed round led by NEA along with amazing investors such as Lukas Biewald, Laszlo Bock, Sean Carey, Jeff Hammerbacher, Ankit Jain, Seval Oz, Joanna Rees, Gokul Rajaram, and Ram Sriram. She was previously the founder and managing partner at Gradient Ventures. She was the VP Engineering at Google for 14 years. Anna's favorite book: Books she reads with her daughters as part of their family book club(00:01) Introduction & AI Infra 101(01:11) Budget Breakdown: Training vs Inference(02:16) Mapping the AI Infra Landscape(04:18) Verticalized vs General-Purpose Infrastructure(06:22) Why Ceramic Was Built From Scratch(08:35) MVP Tradeoffs and Decision Framework(10:16) Achieving 2.5x Speedup in Long Context Training(11:50) Short vs Medium vs Long Context: A Primer(13:38) Long Context vs RAG (Retrieval-Augmented Generation)(15:24) Real-World Impact of Long Context Models(16:38) Bottlenecks at 96K Token Contexts(17:51) Data Pruning 101: What to Keep, What to Drop(21:01) What Is “Good Data” in Subjective Domains?(22:32) How to Grade Reasoning, Not Just Answers(24:15) Synthetic Data: Use Cases & Limits(26:19) Staying Current in Fast-Moving Domains(27:30) Will Every Company Have Its Own Model?(29:23) Unlocking the Next 10x in Infra(31:27) Favorite Recent AI Advancements(32:33) Rapid Fire Round--------Where to find Anna Patterson: LinkedIn: https://www.linkedin.com/in/anna-patterson-15921ba/--------Where to find Prateek Joshi: Newsletter: https://prateekjoshi.substack.com Website: https://prateekj.com LinkedIn: https://www.linkedin.com/in/prateek-joshi-91047b19 X: https://x.com/prateekvjoshi 

In this episode of the Data Center Frontier Show, we sit down with Kevin Cochrane, Chief Marketing Officer of Vultr, to explore how the company is positioning itself at the forefront of AI-native cloud infrastructure, and why they're all-in on AMD's GPUs, open-source software, and a globally distributed strategy for the future of inference. Cochrane begins by outlining the evolution of the GPU market, moving from a scarcity-driven, centralized training era to a new chapter focused on global inference workloads. With enterprises now seeking to embed AI across every application and workflow, Vultr is preparing for what Cochrane calls a “10-year rebuild cycle” of enterprise infrastructure—one that will layer GPUs alongside CPUs across every corner of the cloud. Vultr's recent partnership with AMD plays a critical role in that strategy. The company is deploying both the MI300X and MI325X GPUs across its 32 data center regions, offering customers optimized options for inference workloads. Cochrane explains the advantages of AMD's chips, such as higher VRAM and power efficiency, which allow large models to run with fewer GPUs—boosting both performance and cost-effectiveness. These deployments are backed by Vultr's close integration with Supermicro, which delivers the rack-scale servers needed to bring new GPU capacity online quickly and reliably. Another key focus of the episode is ROCm (Radeon Open Compute), AMD's open-source software ecosystem for AI and HPC workloads. Cochrane emphasizes that Vultr is not just deploying AMD hardware; it's fully aligned with the open-source movement underpinning it. He highlights Vultr's ongoing global ROCm hackathons and points to zero-day ROCm support on platforms like Hugging Face as proof of how open standards can catalyze rapid innovation and developer adoption. “Open source and open standards always win in the long run,” Cochrane says. “The future of AI infrastructure depends on a global, community-driven ecosystem, just like the early days of cloud.” The conversation wraps with a look at Vultr's growth strategy following its $3.5 billion valuation and recent funding round. Cochrane envisions a world where inference workloads become ubiquitous and deeply embedded into everyday life—from transportation to customer service to enterprise operations. That, he says, will require a global fabric of low-latency, GPU-powered infrastructure. “The world is going to become one giant inference engine,” Cochrane concludes. “And we're building the foundation for that today.” Tune in to hear how Vultr's bold moves in open-source AI infrastructure and its partnership with AMD may shape the next decade of cloud computing, one GPU cluster at a time.

Send us a textSupport the showWhatsApp: +66 (Thailand) 06 3359 0002Emails: Arseniobuck@icloud.com ////// arseniobuck2014@outlook.comInstagram: https://www.instagram.com/thearsenioseslpodcast/Second Instagram: https://www.instagram.com/arsenioseslpodcastt/ Facebook: https://www.facebook.com/ArseniosESLPodcast/ Youtube: https://www.youtube.com/channel/UCIzp4EdbJVMhhSnq_0u4ntA

Alex Edmans, a professor of finance at London Business School, tells us how to avoid the Ladder of Misinference by examining how narratives, statistics, and articles can mislead, especially when they align with our preconceived notions and confirm what we believe is true, assume is true, and wish were true.Alex Edmans May Contain LiesWhat to Test in a Post Trust WorldHow Minds ChangeDavid McRaney's TwitterDavid McRaney's BlueSkyYANSS TwitterYANSS FacebookNewsletterKittedPatreon

In this conversation, Jay Goldberg and Austin Lyons discuss Nvidia's recent earnings report, the future of AI and inference, and the dynamics of the AI market, including the impact of China on Nvidia's revenue. They explore the differences between consumer and enterprise workloads, the role of financing in AI server sales, and the challenges of realizing ROI from AI investments. The discussion also touches on real-world applications of AI in business and the future of AI integration in consumer products.

China may have been the big headline out of Nvidia's quarter, with 28 mentions on the earnings call, but right behind was inference at 27. It represents the next wave of AI, models that generate responses after getting trained, and could unlock a major new growth engine for Nvidia. 

Build and run your AI apps and agents at scale with Azure. Orchestrate multi-agent apps and high-scale inference solutions using open-source and proprietary models, no infrastructure management needed. With Azure, connect frameworks like Semantic Kernel to models from DeepSeek, Llama, OpenAI's GPT-4o, and Sora, without provisioning GPUs or writing complex scheduling logic. Just submit your prompt and assets, and the models do the rest. Using Azure's Model as a Service, access cutting-edge models, including brand-new releases like DeepSeek R1 and Sora, as managed APIs with autoscaling and built-in security. Whether you're handling bursts of demand, fine-tuning models, or provisioning compute, Azure provides the capacity, efficiency, and flexibility you need. With industry-leading AI silicon, including H100s, GB200s, and advanced cooling, your solutions can run with the same power and scale behind ChatGPT. Mark Russinovich, Azure CTO, Deputy CISO, and Microsoft Technical Fellow, joins Jeremy Chapman to share how Azure's latest AI advancements and orchestration capabilities unlock new possibilities for developers. ► QUICK LINKS: 00:00 - Build and run AI apps and agents in Azure 00:26 - Narrated video generation example with multi-agentic, multi-model app 03:17 - Model as a Service in Azure 04:02 - Scale and performance 04:55 - Enterprise grade security 05:17 - Latest AI silicon available on Azure 06:29 - Inference at scale 07:27 - Everyday AI and agentic solutions 08:36 - Provisioned Throughput 10:55 - Fractional GPU Allocation 12:13 - What's next for Azure? 12:44 - Wrap up ► Link References For more information, check out https://aka.ms/AzureAI ► Unfamiliar with Microsoft Mechanics? As Microsoft's official video series for IT, you can watch and share valuable content and demos of current and upcoming tech from the people who build it at Microsoft. • Subscribe to our YouTube: https://www.youtube.com/c/MicrosoftMechanicsSeries • Talk with other IT Pros, join us on the Microsoft Tech Community: https://techcommunity.microsoft.com/t5/microsoft-mechanics-blog/bg-p/MicrosoftMechanicsBlog • Watch or listen from anywhere, subscribe to our podcast: https://microsoftmechanics.libsyn.com/podcast ► Keep getting this insider knowledge, join us on social: • Follow us on Twitter: https://twitter.com/MSFTMechanics • Share knowledge on LinkedIn: https://www.linkedin.com/company/microsoft-mechanics/ • Enjoy us on Instagram: https://www.instagram.com/msftmechanics/ • Loosen up with us on TikTok: https://www.tiktok.com/@msftmechanics

In this episode of The Shoulder Physio Podcast, Dr Jared Powell sits down with Dr Mervyn Travers, physiotherapist, S&C coach, and researcher, to explore one of the most compelling frameworks in contemporary pain science: active inference. They discuss how this predictive brain model helps explain persistent musculoskeletal pain, why traditional exercise-based interventions might miss the mark, and how clinicians can use movement and context to shift a patient's pain experience. Merv blends philosophy, neuroscience, and clinical pragmatism in a way that's accessible, challenging, and highly relevant. Key talking points: What is active inference and how does it relate to predictive processing? The role of prior beliefs, culture, and clinical language in shaping pain Movement experimentation as a tool for model updating and recovery Why it's time to rethink how we prescribe exercise in pain rehab Clinical implications from landmark studies within the field that lend themselves to active inference A call for compassion, curiosity, and nuance in patient care Check out the Shoulder Physio Online Course here Connect with Jared and guests: Jared on Instagram: @‌shoulder_physio Jared on X: @‌jaredpowell12 Merv website: Home - Optimise Rehab Merv X: @‌mervtravers Merv Instagram: @‌optimise_rehab See our Disclaimer here: The Shoulder Physio - Disclaimer

New episode with my good friends Sholto Douglas & Trenton Bricken. Sholto focuses on scaling RL and Trenton researches mechanistic interpretability, both at Anthropic.We talk through what's changed in the last year of AI research; the new RL regime and how far it can scale; how to trace a model's thoughts; and how countries, workers, and students should prepare for AGI.See you next year for v3. Here's last year's episode, btw. Enjoy!Watch on YouTube; listen on Apple Podcasts or Spotify.----------SPONSORS* WorkOS ensures that AI companies like OpenAI and Anthropic don't have to spend engineering time building enterprise features like access controls or SSO. It's not that they don't need these features; it's just that WorkOS gives them battle-tested APIs that they can use for auth, provisioning, and more. Start building today at workos.com.* Scale is building the infrastructure for safer, smarter AI. Scale's Data Foundry gives major AI labs access to high-quality data to fuel post-training, while their public leaderboards help assess model capabilities. They also just released Scale Evaluation, a new tool that diagnoses model limitations. If you're an AI researcher or engineer, learn how Scale can help you push the frontier at scale.com/dwarkesh.* Lighthouse is THE fastest immigration solution for the technology industry. They specialize in expert visas like the O-1A and EB-1A, and they've already helped companies like Cursor, Notion, and Replit navigate U.S. immigration. Explore which visa is right for you at lighthousehq.com/ref/Dwarkesh.To sponsor a future episode, visit dwarkesh.com/advertise.----------TIMESTAMPS(00:00:00) – How far can RL scale?(00:16:27) – Is continual learning a key bottleneck?(00:31:59) – Model self-awareness(00:50:32) – Taste and slop(01:00:51) – How soon to fully autonomous agents?(01:15:17) – Neuralese(01:18:55) – Inference compute will bottleneck AGI(01:23:01) – DeepSeek algorithmic improvements(01:37:42) – Why are LLMs ‘baby AGI' but not AlphaZero?(01:45:38) – Mech interp(01:56:15) – How countries should prepare for AGI(02:10:26) – Automating white collar work(02:15:35) – Advice for students Get full access to Dwarkesh Podcast at www.dwarkesh.com/subscribe

 In this episode, Carrie explores whether Inference-based Cognitive Behavioral Therapy (ICBT) is a good fit for individuals struggling with OCD—especially those who haven't found success with exposure and response prevention (ERP). Episode Highlights:The key differences between ERP and ICBT, and why ICBT may be a better fit for certain individuals with OCD.How ICBT helps unpack the reasoning behind obsessions rather than just managing behaviors.Why ICBT can be especially valuable for Christians seeking faith-sensitive OCD treatment.The limitations and challenges of ERP, including dropout rates and religious exposure concerns.What it takes to succeed with ICBT, including a willingness to deeply engage with the learning and healing process. Join the waitlist for the Christians Learning ICBT training: https://carriebock.com/training/ Explore Carrie's services and courses: carriebock.com/services/ carriebock.com/resources/Follow us on Instagram: www.instagram.com/christianfaithandocd/and like our Facebook page: https://www.facebook.com/christianfaithandocd for the latest updates and sneak peeks.

Welcome to this week's episode of The Mixtape with Scott. Today's podcast guest is our 127th guest on the show—Vitor Possebom, Assistant Professor in the Department of Economics at the Fundação Getulio Vargas. Vitor's research sits at the intersection of two areas — econometrics and causal inference, and policy evaluation in Latin America, particularly Brazil. His contributions revolve around refining and extending tools for estimating causal effects in observational data, especially under common data imperfections like selection bias, measurement error, and treatment effect heterogeneity.* Sample selection and marginal treatment effects (e.g., “Identifying Marginal Treatment Effects in the Presence of Sample Selection” (Journal of Econometrics), “Crime and Mismeasured Punishment” (Review of Economics and Statistics))* Misclassification and measurement error (e.g., “Potato Potahto in the FAO-GAEZ Productivity Measures?”)* Inference and sensitivity in synthetic control methods (e.g., “Cherry Picking with Synthetic Controls”, “Synthetic Control Method: Inference, Sensitivity Analysis and Confidence Sets”)* Probability of causation in non-experimental settings (e.g., “Probability of Causation with Sample Selection”)I invited Vitor onto the podcast because of his creative contributions to causal inference, as he fits into a larger informal series I've been for the last several years on causal inference in general. In today's conversation, we talk about Vitor's path from Brazil to Yale University and then back. Vitor, thank you so much for joining us.Scott's Mixtape Substack is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber. Get full access to Scott's Mixtape Substack at causalinf.substack.com/subscribe

An airhacks.fm conversation with Juan Fumero (@snatverk) about: tornadovm as a Java parallel framework for accelerating data parallelization on GPUs and other hardware, first GPU experiences with ELSA Winner and Voodoo cards, explanation of TornadoVM as a plugin to existing JDKs that uses Graal as a library, TornadoVM's programming model with @parallel and @reduce annotations for parallelizable code, introduction of kernel API for lower-level GPU programming, TornadoVM's ability to dynamically reconfigure and select the best hardware for workloads, implementation of LLM inference acceleration with TornadoVM, challenges in accelerating Llama models on GPUs, introduction of tensor types in TornadoVM to support FP8 and FP16 operations, shared buffer capabilities for GPU memory management, comparison of Java Vector API performance versus GPU acceleration, discussion of model quantization as a potential use case for TornadoVM, exploration of Deep Java Library (DJL) and its ND array implementation, potential standardization of tensor types in Java, integration possibilities with Project Babylon and its Code Reflection capabilities, TornadoVM's execution plans and task graphs for defining accelerated workloads, ability to run on multiple GPUs with different backends simultaneously, potential enterprise applications for LLMs in Java including model distillation for domain-specific models, discussion of Foreign Function & Memory API integration in TornadoVM, performance comparison between different GPU backends like OpenCL and CUDA, collaboration with Intel Level Zero oneAPI and integrated graphics support, future plans for RISC-V support in TornadoVM Juan Fumero on twitter: @snatverk

Agentic AI is equally as daunting as it is dynamic. So…… how do you not screw it up? After all, the more robust and complex agentic AI becomes, the more room there is for error. Luckily, we've got Dr. Maryam Ashoori to guide our agentic ways. Maryam is the Senior Director of Product Management of watsonx at IBM. She joined us at IBM Think 2025 to break down agentic AI done right. Newsletter: Sign up for our free daily newsletterMore on this Episode: Episode PageJoin the discussion: Have a question? Join the convo here.Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineupWebsite: YourEverydayAI.comEmail The Show: info@youreverydayai.comConnect with Jordan on LinkedInTopics Covered in This Episode:Agentic AI Benefits for EnterprisesWatson X's New Features & AnnouncementsAI-Powered Enterprise Solutions at IBMResponsible Implementation of Agentic AILLMs in Enterprise Cost OptimizationDeployment and Scalability EnhancementsAI's Impact on Developer ProductivityProblem-Solving with Agentic AITimestamps:00:00 AI Agents: A Business Imperative06:14 "Optimizing Enterprise Agent Strategy"09:15 Enterprise Leaders' AI Mindset Shift09:58 Focus on Problem-Solving with Technology13:34 "Boost Business with LLMs"16:48 "Understanding and Managing AI Risks"Keywords:Agentic AI, AI agents, Agent lifecycle, LLMs taking actions, WatsonX.ai, Product management, IBM Think conference, Business leaders, Enterprise productivity, WatsonX platform, Custom AI solutions, Environmental Intelligence Suite, Granite Code models, AI-powered code assistant, Customer challenges, Responsible AI implementation, Transparency and traceability, Observability, Optimization, Larger compute, Cost performance optimization, Chain of thought reasoning, Inference time scaling, Deployment service, Scalability of enterprise, Access control, Security requirements, Non-technical users, AI-assisted coding, Developer time-saving, Function calling, Tool calling, Enterprise data integration, Solving enterprise problems, Responsible implementation, Human in the loop, Automation, IBM savings, Risk assessment, Empowering workforce.Send Everyday AI and Jordan a text message. (We can't reply back unless you leave contact info)

“We want every layer — chip, system, software — because when you own the stack you can outrun a GPU cluster by 40-70x,” Cerebras CEO Andrew Feldman says. In this episode of Tech Disruptors, Cerebras returns to the Bloomberg Intelligence podcast studios as Feldman joins Bloomberg Intelligence's Kunjan Sobhani and Mandeep Singh to explain the progress from “biggest chip” to “fastest inference cloud.” Feldman unpacks the WSE-3 upgrade, six new data-center builds and fresh Meta and IBM deals that aim to deliver sub-second answers at a fraction of GPU cost, plus Feldman's views on scaling laws, synthetic data and the looming power crunch. 

What if your LLM could think ahead—preparing answers before questions are even asked?In this week's paper read, we dive into a groundbreaking new paper from researchers at Letta, introducing sleep-time compute: a novel technique that lets models do their heavy lifting offline, well before the user query arrives. By predicting likely questions and precomputing key reasoning steps, sleep-time compute dramatically reduces test-time latency and cost—without sacrificing performance.​We explore new benchmarks—Stateful GSM-Symbolic, Stateful AIME, and the multi-query extension of GSM—that show up to 5x lower compute at inference, 2.5x lower cost per query, and up to 18% higher accuracy when scaled.​You'll also see how this method applies to realistic agent use cases and what makes it most effective.If you care about LLM efficiency, scalability, or cutting-edge research.Explore more AI research, or sign up to hear the next session live: arize.com/ai-research-papersLearn more about AI observability and evaluation, join the Arize AI Slack community or get the latest on LinkedIn and X.

In this episode of AI Basics, Jason sits down with Amin Vahdat, VP of ML at Google Cloud, to unpack the mind-blowing infrastructure behind modern AI. They dive into how Google's TPUs power massive queries, why 2025 is the “Year of Inference,” and how startups can now build what once felt impossible. From real-time agents to exponential speed gains, this is a look inside the AI engine that's rewriting the future.*Timestamps:(0:00) Jason introduces today's guest Amin Vahdat(3:18) Data movement implications for founders and historical bandwidth perspective(5:29) The shift to inference and AI infrastructure trends in startups and enterprises(8:40) Evolution of productivity and potential of low-code/no-code development(11:20) AI infrastructure pricing, cost efficiency, and historical innovation(17:53) Google's TPU technology and infrastructure scale(23:21) Building AI agents for startup evaluation and supervised associate agents(26:08) Documenting decisions for AI learning and early AI agent development*Uncover more valuable insights from AI leaders in Google Cloud's 'Future of AI: Perspectives for Startups' report. Discover what 23 AI industry leaders think about the future of AI—and how it impacts your business. Read their perspectives here: https://goo.gle/futureofai*Check out all of the Startup Basics episodes here: https://thisweekinstartups.com/basicsCheck out Google Cloud: https://cloud.google.com/*Follow Amin:LinkedIn: https://www.linkedin.com/in/vahdat/?trk=public_post_feed-actor-name*Follow Jason:X: https://twitter.com/JasonLinkedIn: https://www.linkedin.com/in/jasoncalacanis*Follow TWiST:Twitter: https://twitter.com/TWiStartupsYouTube: https://www.youtube.com/thisweekinInstagram: https://www.instagram.com/thisweekinstartupsTikTok: https://www.tiktok.com/@thisweekinstartupsSubstack: https://twistartups.substack.com

This episode is sponsored by the DFINITY Foundation.  DFINITY Foundation's mission is to develop and contribute technology that enables the Internet Computer (ICP) blockchain and its ecosystem, aiming to shift cloud computing into a fully decentralized state. Find out more at https://internetcomputer.org/     In this episode of Eye on AI, we sit down with Sid Sheth, CEO and Co-Founder of d-Matrix, to explore how his company is revolutionizing AI inference hardware and taking on industry giants like NVIDIA.   Sid shares his journey from building multi-billion-dollar businesses in semiconductors to founding d-Matrix—a startup focused on generative AI inference, chiplet-based architecture, and ultra-low latency AI acceleration.   We break down: Why the future of AI lies in inference, not training How d-Matrix's Corsair PCIe accelerator outperforms NVIDIA's H200 The role of in-memory compute and high bandwidth memory in next-gen AI chips How d-Matrix integrates seamlessly into hyperscaler and enterprise cloud environments Why AI infrastructure is becoming heterogeneous and what that means for developers The global outlook on inference chips—from the US to APAC and beyond How Sid plans to build the next NVIDIA-level company from the ground up.   Whether you're building in AI infrastructure, investing in semiconductors, or just curious about the future of generative AI at scale, this episode is packed with value.     Stay Updated: Craig Smith on X:https://x.com/craigss Eye on A.I. on X: https://x.com/EyeOn_AI   (00:00) Intro (02:46) Introducing Sid Sheth (05:27) Why He Started d-Matrix (07:28) Lessons from Building a $2.5B Chip Business (11:52) How d-Matrix Prototypes New Chips (15:06) Working with Hyperscalers Like Google & Amazon (17:27) What's Inside the Corsair AI Accelerator (21:12) How d-Matrix Beats NVIDIA on Chip Efficiency (24:10) The Memory Bandwidth Advantage Explained (26:27) Running Massive AI Models at High Speed (30:20) Why Inference Isn't One-Size-Fits-All (32:40) The Future of AI Hardware (36:28) Supporting Llama 3 and Other Open Models (40:16) Is the Inference Market Big Enough? (43:21) Why the US Is Still the Key Market (46:39) Can India Compete in the AI Chip Race? (49:09) Will China Catch Up on AI Hardware?  

Nick Kirby and Trace Fowler break down the Cincinnati Reds' Wednesday night loss to the Seattle Mariners. They analyze Nick Martinez's lackluster start, Elly De La Cruz's defensive miscues, and the contentious batter interference call on Austin Hays late in the game, among other key moments. Nick also recaps the Reds minor league action and preview Thursday's pitching matchup between Brady Singer and Emerson Hancock.     Today's Episode on YouTube: https://www.youtube.com/watch?v=t4ihlaIz8J8&t=3032s    DSC Commodities: https://deepsouthcommodities.com/    CALL OR TEXT 988 FOR HELP DAY OR NIGHT: https://mantherapy.org/get-help/national-resources/164/lifeline-crisis-chat    OTHER CHATTERBOX PROGRAMING:   Off The Bench: https://otbthombrennaman.podbean.com/  Chatterbox Bengals: https://podcasts.apple.com/us/podcast/chatterbox-bengals-a-cincinnati-bengals-nfl-podcast/id1652732141  Chatterbox Bearcats: https://chatterboxbearcats.podbean.com/  Dialed In with Thom Brennaman: https://www.youtube.com/playlist?list=PLjPJjEFaBD7VLxmcTTWV0ubHu_cSFdEDU  Chatterbox Man on the Street: https://www.youtube.com/watch?v=3Ye-HjJdmmQ&list=PLjPJjEFaBD7V0GOh595LyjumA0bZaqwh9&pp=iAQB 

In this riveting episode of the OCD Whisperer podcast, host Kristina Orlova sits down with Mike Parker, a licensed clinical social worker and the creator of the popular YouTube channel OCD Space.   Together, they embark on a deep dive into the world of OCD and the transformative power of Inference based cognitive-behavioral therapy (ICBT). But what happens when doubt becomes the driving force behind every thought? And how can someone trapped in the cycle of obsessional doubt ever learn to trust their own mind again?   Mike Parker pulls back the curtain on the insidious nature of "obsessional doubt," a phenomenon that leaves individuals questioning their every thought, memory, and perception. Why do those with OCD feel compelled to seek reassurance over and over, even when they know it offers only fleeting relief? And how does this relentless doubt keep them locked in a prison of their own mind?   As the conversation deepens, Kristina and Mike explore the critical differences between ICBT and exposure and response prevention (ERP). But here's the burning question: Can understanding the origin of obsessive thoughts be the key to breaking free from their grip? Mike sheds light on how inferential confusion and obsessional doubt drive OCD.   This episode is a masterclass in navigating the labyrinth of OCD treatment. Will listeners walk away with a newfound understanding of how to confront their doubts? Or will the complexities of the human mind leave them questioning everything they thought they knew? Tune in to uncover the answers—and perhaps, a path to freedom. In This Episode [00:02]  Introduction to the episode   [00:56]  Understanding ICBT   [02:00]  Obsessional doubt explained   [02:21]  Differentiating ICBT from ERP   [03:36]  The nature of obsessional doubt   [05:58]  Reassurance-seeking behavior   [09:25]  Understanding internal evidence   [11:27]  The role of self-knowledge   [13:31]  General facts vs. personal context   [14:49]  Handling real mistakes   [16:40]  Exploring early memories   [17:46]  Understanding obsessional doubt   [19:22]  Childhood influences on OCD   [20:28]  Clarifying ICBT vs. psychodynamic therapy   [21:44]  Focus of inference-based CBT   [22:41]  Cognitive distortions in OCD   [25:34]  Re-evaluating daily routines   [27:06]  Timeframe for progress in treatment   [29:22]  Complicating factors in OCD treatment   Notable Quotes [00:02:42] "Obsessional doubt is a core process identified in OCD when you're doing I-CBT. It's a thought process where someone with OCD knows something but doesn't trust themselves enough to stick with what they know, leading them to question, dismiss, and seek more information than they have." - Michael Parker   [00:18:26] "We can start to see how long the client has been telling themselves an obsessional story about themselves... It was all logged in there and then all put together, but if we go back, we can see this actually never meant you should be locked into never-ending doubt." - Michael Parker   [00:23:39]"I-CBT is primarily a cognitive therapy... The focus really is figuring out why you reject information, why you don't trust it... Let's figure out why you doubted." — Michael Parker Our Guest Mike Parker, LCSW, is a licensed clinical social worker and private practice therapist based in Pittsburgh, Pennsylvania. He specializes in treating obsessive-compulsive disorder (OCD) using cognitive-behavioral therapy (CBT) and inference-based cognitive therapy (I-CBT). As the host of the OCD Space YouTube channel, Mike is dedicated to educating individuals and mental health professionals on effective OCD treatment approaches. He is passionate about helping clients understand and overcome obsessional doubt while also training fellow therapists in evidence-based interventions. With a focus on empowering individuals to trust themselves and break free from the cycle of compulsions, Mike continues to be a leading voice in the OCD treatment community. Resources & Links   Kristina Orlova, LMFT Instagram YouTube OCD CBT Journal Tracker and Planner Website   Mike Parker Website  LinkedIn  YouTube Cognitive Therapy for OCD Disclaimer   Please note, while our host is a licensed marriage and family therapist specializing in OCD and anxiety disorders in the state of California, this podcast is for educational purposes only and should not be considered a substitute for therapy.   Stay tuned for bi-weekly episodes filled with valuable insights and tips for managing OCD and anxiety. And remember, keep going in the meantime. See you in the next episode!

Andrew Feldman is the Co-Founder and CEO @ Cerebras, the fastest AI inference + training platform in the world. In Sept 2024 the company filed to go public off the back of a rumoured $1BN deal with G42 in the UAE. Andrew is the leading expert for all things inference.  In Today's Episode We Discuss: 04:23 Where Was AI Landscape in 2015 When Cerebras Founded 05:57 NVIDIA's Biggest Strength Has Become Their Biggest Weakness 07:09 What Happens to the Cost of Inference? 08:55 Why Are AI Algorithms So Inefficient? 20:30 Why is it Total BS That We Have Hit Scaling Laws? 23:07 What Will Be the Ratio of Synthetic to Human Data Used in 5 Years? 31:37 What Specifically Was So Impressive About Deepseek? 31:51 Why is Distillation Not Wrong and OpenAI Need to Look in the Mirror? 32:34 Where Will Value Accrue in a World of AI? 34:08 How Will NVIDIA's Market Position Change Over the Next Five Years? 39:59 Why is the CUDA Lockin for NVIDIA BS? What is Their Weakness? 40:46 Why is Trump Better for Business than Biden? 49:41 Do We Underestimate China in a World of AI? 52:33 What is the Most Underappreciated Segment of AI? 54:00 Quickfire Round  

// GUEST //X: ⁠https://x.com/satmojoe⁠What's the Problem? X: ⁠https://x.com/SatsVsFiat⁠Website: ⁠https://www.satsvsfiat.com/⁠ // SPONSORS //The Farm at Okefenokee: ⁠https://okefarm.com/⁠iCoin: ⁠https://icointechnology.com/breedlove⁠Heart and Soil Supplements (use discount code BREEDLOVE): ⁠https://heartandsoil.co/⁠In Wolf's Clothing: ⁠https://wolfnyc.com/⁠Blockware Solutions: ⁠https://mining.blockwaresolutions.com/breedlove⁠On Ramp: ⁠https://onrampbitcoin.com/?grsf=breedlove⁠Mindlab Pro: ⁠https://www.mindlabpro.com/breedlove⁠Coinbits: ⁠https://coinbits.app/breedlove⁠ // PRODUCTS I ENDORSE //Protect your mobile phone from SIM swap attacks: ⁠https://www.efani.com/breedlove⁠Noble Protein (discount code BREEDLOVE for 15% off): ⁠https://nobleorigins.com/⁠Lineage Provisions (use discount code BREEDLOVE): ⁠https://lineageprovisions.com/?ref=breedlove_22⁠Colorado Craft Beef (use discount code BREEDLOVE): ⁠https://coloradocraftbeef.com/⁠ // SUBSCRIBE TO THE CLIPS CHANNEL //⁠https://www.youtube.com/@robertbreedloveclips2996/videos⁠ // OUTLINE //0:00 - WiM Episode Trailer1:12 - “What's the Problem?”10:04 - The Pernicious Cycle of Keynesian Systems26:15 - Breaking Out of the Fiat World28:41 - The Farm at Okefenokee30:00 - iCoin Bitcoin Wallet31:32 - Bitcoiner Openness, Disagreeability, and Humility32:46 - Many Problems are Downstream of Broken Money35:41 - Explaining Bitcoin to the Layperson40:07 - Money Printing Enables Theft to Fund War47:07 - Heart and Soil Supplements48:07 - Helping Lightning Startups with In Wolf's Clothing49:01 - All Government Spending is Capital Misallocation58:00 - Fight the System or Defund the System1:04:00 - Ikigai and What Individual Bitcoiners Can Do1:16:28 - Mine Bitcoin with Blockware Solutions1:17:47 - OnRamp Bitcoin Custody1:19:10 - Personality Dispositions of Bitcoiners and Broader Inclusion1:27:47 - Elon Musk and Business Bitcoin Adoption1:34:32 - Bitcoin Adoption is a Positive Feedback Loop1:42:05 - Mind Lab Pro Supplements1:43:13 - Buy Bitcoin with Coinbits1:44:41 - Physics, Inference, and Bitcoin1:52:05 - Joe Bryan's Orange-Pill Paradigm Shift1:58:41 - Financial and Linguistic Liberation2:00:51 - The Inevitability of Bitcoin?2:03:55 - Reactions to “What's the Problem?” “What's the Problem?”2:15:12 - Introduction
2:16:45 – The Problems We All Face
2:17:32 – The Island: A Story of Two Sides
2:21:06 – A Free Market with Perfect Money
2:24:18 – The Government Arrives…
2:26:37 – Manipulation of the Money Supply
2:35:12 – An Ever-Growing Crisis
2:38:48 – The Inevitable Collapse of Government Money
2:45:52 – The Real World Problems with our Money
2:51:09 – What's the Solution? Bitcoin
2:52:57 – How to Learn More about Bitcoin and Stay in Touch // PODCAST //Podcast Website: ⁠https://whatismoneypodcast.com/⁠Apple Podcast: ⁠https://podcasts.apple.com/us/podcast/the-what-is-money-show/id1541404400⁠Spotify: ⁠https://open.spotify.com/show/25LPvm8EewBGyfQQ1abIsE⁠RSS Feed: ⁠https://feeds.simplecast.com/MLdpYXYI⁠ // SUPPORT THIS CHANNEL //Bitcoin: 3D1gfxKZKMtfWaD1bkwiR6JsDzu6e9bZQ7Sats via Strike: ⁠https://strike.me/breedlove22⁠Dollars via Paypal: ⁠https://www.paypal.com/paypalme/RBreedlove⁠Dollars via Venmo: ⁠https://account.venmo.com/u/Robert-Breedlove-2⁠ // SOCIAL //Breedlove X: ⁠https://x.com/Breedlove22⁠WiM? X: ⁠https://x.com/WhatisMoneyShow⁠Linkedin: ⁠https://www.linkedin.com/in/breedlove22/⁠Instagram: ⁠https://www.instagram.com/breedlove_22/⁠TikTok: ⁠https://www.tiktok.com/@breedlove22⁠Substack: ⁠https://breedlove22.substack.com/⁠All My Current Work: ⁠https://linktr.ee/robertbreedlove

Support the show to get full episodes, full archive, and join the Discord community. The Transmitter is an online publication that aims to deliver useful information, insights and tools to build bridges across neuroscience and advance research. Visit thetransmitter.org to explore the latest neuroscience news and perspectives, written by journalists and scientists. Read more about our partnership. Sign up for the “Brain Inspired” email alerts to be notified every time a new “Brain Inspired” episode is released. To explore more neuroscience news and perspectives, visit thetransmitter.org. The concept of a schema goes back at least to the philosopher Immanuel Kant in the 1700s, who use the term to refer to a kind of built-in mental framework to organize sensory experience. But it was the psychologist Frederic Bartlett in the 1930s who used the term schema in a psychological sense, to explain how our memories are organized and how new information gets integrated into our memory. Fast forward another 100 years to today, and we have a podcast episode with my guest today, Alison Preston, who runs the Preston Lab at the University of Texas at Austin. On this episode, we discuss her neuroscience research explaining how our brains might carry out the processing that fits with our modern conception of schemas, and how our brains do that in different ways as we develop from childhood to adulthood. I just said, "our modern conception of schemas," but like everything else, there isn't complete consensus among scientists exactly how to define schema. Ali has her own definition. She shares that, and how it differs from other conceptions commonly used. I like Ali's version and think it should be adopted, in part because it helps distinguish schemas from a related term, cognitive maps, which we've discussed aplenty on brain inspired, and can sometimes be used interchangeably with schemas. So we discuss how to think about schemas versus cognitive maps, versus concepts, versus semantic information, and so on. Last episode Ciara Greene discussed schemas and how they underlie our memories, and learning, and predictions, and how they can lead to inaccurate memories and predictions. Today Ali explains how circuits in the brain might adaptively underlie this process as we develop, and how to go about measuring it in the first place. Preston Lab Twitter: @preston_lab Related papers: Concept formation as a computational cognitive process. Schema, Inference, and Memory. Developmental differences in memory reactivation relate to encoding and inference in the human brain. Read the transcript. 0:00 - Intro 6:51 - Schemas 20:37 - Schemas and the developing brain 35:03 - Information theory, dimensionality, and detail 41:17 - Geometry of schemas 47:26 - Schemas and creativity 50:29 - Brain connection pruning with development 1:02:46 - Information in brains 1:09:20 - Schemas and development in AI

Hagay Lupesko is the SVP for AI Inference at Cerebras Systems. Subscribe to the Gradient Flow Newsletter