AI and Software
Open-weight models
open models such as Llama, DeepSeek, Qwen and their impact
- Episodes
- 43
- Share
- 4.6%
- Shows
- 40
- Net sentiment
- +0.18
Down from 4.8% of episodes in the weeks before; net sentiment was +0.34.
Bullish 17Mixed 12Neutral 4Bearish 10
5 weeks
Share of each week's episodes, and net sentiment.
0%2%4%
Sep 21Sep 28Oct 5Oct 12Oct 19
Net sentiment
−10+1
Sep 21Sep 28Oct 5Oct 12Oct 19
The episodes behind it
Why each one was bullish, bearish or in between. Most time on the theme first.
Bullish 17
- Speakers argue open weights spread capability, aid sovereignty and access, and make the ecosystem more resilient.Mistral Is SO Back: Inside Le Chonk, the West’s Biggest Open Model (Pierre Stock, VP Science, Mistral) · The MAD Podcast: How AI Gets Built — with Matt Turck · Oct 22
- Open-weight models build community, accelerate research, and offer sovereign access that closed models cannot.Mistral Is SO Back: Inside Le Chonk, the West’s Biggest Open Model (Pierre Stock, VP Science, Mistral) · The MAD Podcast · Oct 22
- DeepSeek's MIT-licensed checkpoint and Apache-licensed inference stack are seen as major open-source advances.ThursdAI - Oct 22 - AI finds all the vulnerabilities, Claude Duet, DeepSeek drops Infer & more AI news · ThursdAI - The top AI news from the past week · Oct 23
- Open-weight models are far cheaper and gaining capability fast, so enterprises will shift toward them.Arvind Krishna & Nikesh Arora: Quantum Threat, Open vs Frontier AI, AI Cyber Warfare & SaaS Collapse · All-In Podcast · Oct 21
- Kurtz sees open-weight models as cheap, good enough for enterprises, and likely the future operating system of AI.The New Apex Predator: CrowdStrike CEO George Kurtz on Adversaries Armed With AI Agents · No Priors: Artificial Intelligence | Technology | Startups · Oct 22
- DeepSeek ban may push enterprises toward US open models like Nemotron, boosting demand for them.No One Likes Any of Sam Altman’s 3 AI Endgames | TWiAI E36 · This Week in AI · Oct 22
Show 11 moreShow fewer
- Open source models are racing ahead and seen as better for consumers and startups.The Frontier Ideas a16z Wants to Fund | Marc Andreessen, Ben Horowitz & Chris Dixon · The a16z Show · Oct 22
- Open-weight models offer lower cost, IP protection and auditability, so most token use may shift to them.Arvind Krishna & Nikesh Arora: Quantum Threat, Open vs Frontier AI, AI Cyber Warfare & SaaS Collapse · All-In with Chamath, Jason, Sacks & Friedberg · Oct 21
- Guest argues open models enable diverse checks, customization and safety, benefiting the ecosystem.Physical AI at NVIDIA · Software Engineering Daily · Oct 20
- Open models are seen as a big, monetizable market, with enterprises still wary of Chinese-origin models.20VC: You Need $1BN To Do Seed Today: The Untold Truths of Venture | Who Wins and Who Loses in the Next Decade of VC | Why Margin and Price Matter Less in a World of AI with Sameer Dholakia, Bessemer · The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch · Oct 19
- Cheaper Chinese models like GLM are described as nearly frontier-quality at a fraction of the cost."Stop Typing to AI": Building a 24/7 Autonomous Enterprise · The IDEAL Investor in the Age of Sustainable Abundance · Oct 22
- Host says cheaper Chinese open-weight models are catching frontier quality and gaining adoption.Is the AI Doom Talk a PSYOP?? · Digital Dopamine · Oct 21
- Local open-weight models let businesses run capable AI privately on their own hardware at lower cost.The Weekly Deep Dive: Smarter Software Agents And The End Of Chat Boxes · The RAIZOR Report: Your Weekly AI News Update · Oct 19
- Smaller open-weight and cheaper models are presented as a more diverse, efficient AI ecosystem.The Most Polite Financial Mugging? Max Wolff on the AI Bubble & Japan / 最後にババを引くのは誰？｜AIバブルと日本経済 · Malveaux: The Room · Oct 23
- Open-weight models are near frontier quality, making local inference increasingly attractive for some workloads.Microsoft AI, model routing, and the move toward portable workflows · The Intelligence Age · Oct 19
- Open-source models offer lower cost, sovereignty and control compared with frontier models for organizations.AI Isn’t Moving Too Fast. It’s Too Noisy - Nikolas Badminton · A Beginner's Guide to AI · Oct 19
- Hosts praise DeepSeek's cheap, strong open-weight models and their falling prices.Uneasy Money: Is the Internet Breaking Because Bots Never Get Lazy? · Unchained · Oct 22
Bearish 10
- Open-weight models offer roughly 80% of frontier performance at a fraction of the price, pressuring premium AI pricing.Can OpenAI and Anthropic Justify Their Valuations? + Flock, Privacy, and Plummeting Crime · The Prof G Pod with Scott Galloway · Oct 19
- Open weights with removed safeguards pose serious cyber and bioterrorism risk, requiring platform-level controls.Bill Gates says to prepare for 20 years of ‘tumultuous’ AI disruption · Decoder with Nilay Patel · Oct 22
- Unregulated Chinese open-source models are seen as a hard-to-control risk, distilled from frontier labs.The AI Labs are Bracing for a Catastrophe · Artificial Insanity · Oct 20
- Open source models could erode pricing power for cloud and AI providers, creating bust risk.Investors Shoot for the Moon · Motley Fool Hidden Gems Investing · Oct 19
- Chinese open models are driving token prices down and commoditizing frontier model economics.Nvidia, Memory & the Next Phase of the AI Trade | Protect the Pile Episode 29 w/ Felix Wang · Hedgeye Podcasts · Oct 24
- Open models lack oversight and alignment feedback loops, making them a likely source of rogue agents.35: Bear in the House: Rogue AI and the New Rules of Coexistence · Unseen Unknown · Oct 19
Show 4 moreShow fewer
- Cheaper open models and consumer models pressure frontier labs' token prices and revenue.SpaceX Could Overtake Nvidia, Microsoft & Google in the AI Race · Milk Road Stocks · Oct 19
- Chinese DeepSeek inference economics show strong margins and deflation pressure on token pricing.Citrini: Monkeys Are The Next Bottleneck | Citrini on AI Drug Discovery Experiment & Reflexive Financing Risk to AI Buildout & Hardware Trade · Monetary Matters with Jack Farley · Oct 25
- Freely downloadable model weights are hard to police, undermining regulatory control.Bill Gates on AI: Why a $40B Data Center Kills Moats | Stock Market · Deep Values · Oct 25
- Open models with no guardrails let low-skill attackers run offensive cyber operations at scale.The Real AI Threat Is Two Dudes and a GPU with Damien Lewke, CEO & Founder of Nebulock · RiskReversal Pod · Oct 21
Mixed 12
- Chinese open models now near frontier quality but are costly to run, weakening their cheap-alternative appeal.Ep 886: The U.S. vs China AI Cold War Is Starting: What It Means and How It Impacts You (Replay) · Everyday AI Podcast – An AI and ChatGPT Podcast · Oct 21
- Open-weight models enable attacks without guardrails but also cheaper enterprise defense and sovereignty.The New Apex Predator: CrowdStrike CEO George Kurtz on Adversaries Armed With AI Agents · No Priors · Oct 22
- Permissively licensed Chinese open models lead releases, giving builders strong options but raising provenance concerns.The DeepSeek Ban Meets a $50 Breach Nobody Can Cancel · The Context Report: Today in AI · Oct 20
- New open models from Reflection and Mistral are competitive, but hosts doubt open releases continue given frontier economics.#260 - Muse, Misalignment, Math · Last Week in AI · Oct 19
- NVIDIA's free open models protect its chip moat but add strategic tension with its own customers.Investor's Corner: Sell the Shovels, Keep the Cash · YPO Technology Network AI Brief · Oct 24
- Multi-model orchestration lowers inference costs but shifts value away from model providers.AI’s Next Winners: Signals From Private Markets · Thoughts on the Market · Oct 20
Show 6 moreShow fewer
- Open models will coexist with closed ones; monetization is possible but harder, and US enterprises hesitate over Chinese models.20VC: You Need $1BN To Do Seed Today: The Untold Truths of Venture | Who Wins and Who Loses in the Next Decade of VC | Why Margin and Price Matter Less in a World of AI with Sameer Dholakia, Bessemer · 20VC · Oct 19
- Cheaper open models may shift token revenue away from frontier labs, though they also boost compute demand.How To Invest In AI Now With Janus Henderson’s Jonathan Cofsky · Barron's Live · Oct 19
- Open models are cheaper and growing fast, but frontier labs stay months ahead and the US lacks dominant open models.Is AI a Bubble? I Asked the People Building It | WTF is AI? · WTF is with Nikhil Kamath · Oct 24
- Argument that security-capable models will eventually become open because everyone gets access anyway.Democratizing Security for the Agentic Web with CloudFlare CTO Dane Knecht · No Priors: Artificial Intelligence | Technology | Startups · Oct 23
- Argues defensive and offensive capabilities are intertwined, and open access to security models is inevitable.Democratizing Security for the Agentic Web with CloudFlare CTO Dane Knecht · No Priors · Oct 23
- Open-weight models are harder to safeguard against misuse than centrally served models.#163 Untangling AI risk - a simple breakdown of what we should worry about · plain.txt · Oct 19
Neutral 4
- Tests open-weight models like Qwen, Kimi, Llama and DeepSeek for resume writing quality and accuracy.Best AI for Resume and Cover Letter (Part 2): 24 More Models Tested · Biotech Career Coach · Oct 23
- Enterprises prefer open-weight models for explainability and auditability, shifting spend toward self-hosted compute.AI Cyber Panic, Ken Griffin's Bubble Warning, The Fall of France? · All-In Podcast · Oct 23
- Discussion of DeepSeek V4.1017 and Meta's open Muse model and Chinese small-model strengths.Meta Made A Frontier Model, Google Still Can't Even With Gemini 4 · Nerd Snipe with Theo and Ben · Oct 22
- Small open-weight models running on cheap IoT devices are discussed as a hypothetical security threat.This Dum Week 2026-10-18 · This Dum Week · Oct 19
Around it
Companies named
Shows that covered it most
- All-In Podcast2
- No Priors: Artificial Intelligence | Technology | Startups2
- No Priors2
- A Beginner's Guide to AI1
- 20VC1
- Barron's Live1
Travels with
Themes in the same episodes, against chance.
Themes, sentiment and reasons are AI-written paraphrases, checked so they never quote an episode. They may contain errors. Not investment advice. Episodes remain the property of their creators.
Briefs like this, for everything you follow.
Ethmos puts research agents on your coverage universe — every debate that touches your names, every executive appearance you'd have missed — and files briefs every morning.
Start free14-day free trial