The Generative AI Meetup Podcast

Hosted by Mark and Shashank, software engineers and organizers in Silicon Valley. Get their grounded perspective each week as they explore the generative AI landscape through news analysis, tech discussions, hands-on experiments, and clear explanations.Dive into the latest language models, AI agent capabilities, and RAG techniques. Understand the hardware race, key research, startup trends, benchmarks, and the real-world impact of AI across industries like healthcare, robotics, and creative work. We also test AI limits, explain core concepts, discuss ethics, and interview builders shaping the field.For engineers, developers, researchers, and anyone seeking a practical understanding of AI’s rapid evolution and its applications.

The Generative AI Meetup Podcast is a technology podcast hosted by Mark and Shashank, with 77 episodes on record and a Required Pod Score of 80.

Mark and Shashank hosts The Generative AI Meetup Podcast, a technology show with 77 episodes published.

Our AI reads these to draft pitches. Use them as grounding for a pitch that cites a real guest and a specific topic.

• 

  Episode #78

  Openclaw but for my grandma

  Oct 3, 20261h 20m

  https://novacut.ai/ https://genaimeetup.com/ https://genaipod.substack.com/ This week, we cover the latest developments in AI: always-on cloud agents, Meta’s Muse, proactive email assistants, and what happens when agents escape their sandboxes. We also discuss NVIDIA’s safeguards, recursive self-improvement, new frontier models from Anthropic and Google, Grok and Xiaomi’s open models, and AI’s progress on the Navier–Stokes problem. 0:00 Welcome and AI news roundup 2:14 OpenAI’s always-on cloud agent 3:39 Meta Muse and AI assistants for everyone 10:16 Instinct’s proactive email agent 12:17 When AI agents break their sandbox 19:27 NVIDIA’s hardware and software safeguards 31:22 Recursive self-improvement 41:55 New frontier models: Opus and Sonnet 51:17 Google’s new frontier model 56:31 Grok, Xiaomi and open models 58:56 Jev and the rise of decision models 1:06:58 AI tackles the Navier–Stokes problem 1:10:44 The dispute over who solved the math problem 1:17:15 Wrap-up 1:17:50 NovaCut AI video editor 1:18:58 Substack and sign-off

  ListenShow Notes
• 

  Episode #76

  Cheaper, Faster & Smarter: Qwen 3.8 27B, GLM 5.3, Grok 4.6, Cerebras

  Aug 18, 20261h 50m

  https://novacut.ai/ https://genaimeetup.com/ Jeff returns to the Gen AI Meetup Podcast for a wide-ranging discussion on where AI is heading—and why powerful models running on consumer hardware could change the economics of the entire industry. We dive into Qwen 3.8 27B and the growing viability of running capable LLMs locally, GLM 5.3 and the latest Chinese open-source models, DeepSeek, Gemini 3.7, Grok 4.6, Meta’s latest models, and OpenAI’s partnership with Cerebras for dramatically faster inference. We also discuss whether foundation models are becoming commodities, what that means for companies like OpenAI and Anthropic, and why more value may ultimately move to the application layer. Jeff shares how his team approaches AI in healthcare, including self-hosting, data sovereignty, classifiers, fine-tuning, and spec-driven development for building reliable AI-assisted software without accumulating a mountain of vibe-coded technical debt. Plus: Jeff Dean’s departure from Google, Discovery Loop, Stripe’s OpenRouter acquisition, Anthropic’s controversial AI-text watermarking experiments, and whether watermarking could affect model quality. Topics include: Qwen 3.8 27B, GLM 5.3, DeepSeek V4, Grok 4.6, Gemini 3.7, Cerebras, OpenAI, Anthropic, Meta, local LLMs, open-source AI, model commoditization, spec-driven development, AI healthcare, data sovereignty, AI coding agents, and model watermarking.

  ListenShow Notes
• 

  Episode #75

  Has China finally caught up?

  Jul 27, 20261h 47m

  https://novacut.ai/ In this episode, we break down the biggest stories shaping the AI landscape — from Anthropic's regulatory stance and OpenAI's monetization shift to the latest open-source breakthroughs and model pricing wars. 0:00 Anthropic’s Frustrating Stance 5:32 AI Access and the Intelligence Gap 14:17 Defending Anthropic’s Regulation Approach 25:10 Google DeepMind and Cybersecurity Risks 29:27 GPT 5.6 Soul and Coding Abilities 35:24 Alignment and the Knife Analogy 39:42 Google Gemma, Flash, and Market Value 51:11 AI Lab Focus: Speed vs. Specialization 58:48 Inkling: Mira Murati’s New Model 1:06:21 Open Source and Fine-Tuning 1:12:43 On-Premise Hardware Costs 1:19:54 Open Source Models Hit Frontier 1:23:32 Model Pricing Comparison 1:32:43 The Model Routing Problem 1:35:08 Grok 4.5 and Cursor Partnership 1:40:45 OpenAI’s Monetization Shift 1:42:42 Sponsor: Nova Cut AI

  ListenShow Notes
• 

  Episode #74

  The Chinese DoorDash just entered the AI Race

  Jul 7, 20261h 33m

  https://novacut.ai/ https://genaimeetup.com/ 0:00 Longcat: 1.6T Model Without US GPUs 1:08 Meituan: The Super App Behind Longcat 4:54 China's Exploding AI Competitor Scene 8:23 Inside Huawei's Ascend GPU Architecture 17:14 Cost & Energy: Huawei vs Nvidia 29:21 OpenAI's Custom Inference Chip Strategy 36:23 Software Optimizations: The Path to 10,000x 45:38 GPT-5.6: Sol, Terra, Luna Models 58:02 CursorBench & the New Coding Benchmarks 1:22:16 Meta's Non-Invasive Brain-to-Text 1:26:30 Anthropic Science: AI for Researchers 1:30:59 Outro & Community Ask China just dropped a 1.6-trillion-parameter model without access to US GPUs — and it's running on Huawei's homegrown Ascend chips. In this episode, we break down: 🔹 Longcat — the massive model built by Meituan, China's super app giant 🔹 China's exploding AI competitor ecosystem 🔹 Inside the **Huawei Ascend GPU architecture **: specs, costs, and energy tradeoffs vs. Nvidia 🔹 OpenAI's custom inference chip strategy 🔹 The software optimizations driving a 10,000x efficiency leap 🔹 GPT-5.6: Sol, Terra, and Luna models explained 🔹 New coding benchmarks with CursorBench 🔹 Meta's non-invasive brain-to-text research 🔹 Anthropic Science — AI built for researchers

  ListenShow Notes
• 

  Episode #73

  What happened to my Fable?

  Jun 22, 20261h 30m

  https://novacut.ai/ Description: Anthropic pulls access to Fable, and China responds the same day with GLM 5.2. In this episode we break down the escalating AI arms race, US export controls on chips and frontier models, and whether the "Great Firewall of America" is already here. ⏱️ Topics: Anthropic restricts Fable — what happened and why China's GLM 5.2 release and how close they're catching up US trust, surveillance, and AI gatekeeping Token pricing chaos — cost per task vs. cost per token Model routing, loop engineering, and autonomous agents Anthropic's Mythos model and Fable safeguard philosophy Xiaomi NEMO V2.5 Pro Ultra Speed Midjourney's bizarre health spa pivot AI Engineer Conference wrap-up 🔗 Links & Resources: Fable https://www.anthropic.com/news/claude-fable-5-mythos-5 https://www.theregister.com/security/2026/06/15/feds-freaked-over-fable-5-after-simple-fix-this-code-prompt-not-jailbreak-says-researcher/5255827 “Fix this code” https://support.claude.com/en/articles/14328960-identity-verification-on-claude Midjourney https://www.midjourney.com/medical/blogpost Full body ultrasound CT scanner Xiaomi 1000tps https://mimo.xiaomi.com/blog/mimo-tilert-1000tps (MiMo-V2.5-Pro-UltraSpeed: Pushing 1T-Parameter Model Generation Speed to 1000 TPS Best opensource model https://z.ai/blog/glm-5.2 📌 Timestamps in the chapters section above. #AIPodcast #Anthropic #Fable #GLM52 #AIArmsRace #LLM #GenAI 0:00 Intro: Anthropic restricts Fable access 1:00 China's response and GLM 5.2 release 1:55 US trust and AI model reliability 2:37 Geopolitics and AI regulations 4:17 AI arms race and export control limits 5:49 Fable usage experience and value 6:30 Z.AI subscription and pricing comparison 9:07 Subscription limits vs. API usage 10:30 GLM 5.2 token limits and utility 13:52 China catches up: GLM vs. US models 15:06 Intelligence index and model cost trend 17:40 Token pricing complexity and value 23:30 Cost per task vs. cost per token 24:35 Model routing and usage optimization 29:43 Loop engineering and autonomous agents 33:23 NovaCut: AI video editor and ad loops 37:07 Anthropic's Fable re-release timeline 38:41 US gatekeeping and China's advantage 40:11 Great Firewall of America risks 44:07 Mass surveillance and free speech 46:01 Global AI trust and market shift 47:48 Enforcing identity checks on AI 48:54 Export bans on chips and hardware 51:52 US restricts allies from frontier models 52:39 Anthropic's talent and Mythos model 53:51 Fable safeguards and US government view 1:00:12 Xiaomi NEMO V2.5 Pro Ultra Speed model 1:02:39 Optimizing for intelligence, cost, speed, and size 1:13:19 Midjourney's unexpected health spa pivot 1:28:30 AI Engineer conference and podcast wrap

  ListenShow Notes

Recent guests on The Generative AI Meetup Podcast. Study who booked and why before you pitch.

Based on semantic analysis of episode topics and host coverage, this show is a strong guest fit for executives in:

Industry fit is computed by PitchCentric using vector embeddings of the show's episode catalog.