Ask in Claude
Ask in ChatGPT
- Podcasts
- Technology
- Lightwork
31 episodes
Lightwork
by Lightspeed · English
There's more AI news than anyone can possibly consume. Lightwork is the filter. Every Monday, Lightspeed's Claire Zau and Josh Machiz break down the one development that mattered most that week — a model release, a funding round, a company structural shift — and work through what changes because of it. They also bring in a guest working at the front of AI to sharpen the picture. By the end of each episode, you'll have a clearer view of where things are going and why it matters to you. New episodes every Monday morning. Have something you want to hear? Email us at [email protected].
Latest episodes
All episodes
Inside Nominal, the Intelligence Platform for Hardware Engineering | Lightwork
Nominal’s “physical AI” platform for hardware engineering—helping teams analyze test data, operate systems in real time, and safely adopt AI agents with auditability.
9 Nov 2026 · 26m · 14 chaptersA Haiku From Orbit, The Binance Breach & Reddit Feels the AI Squeeze | Lightwork
Lightwork episode covers three big tech/security shifts: Google’s Project Suncatcher, the Binance breach, and AI-driven “dead internet”/Reddit traffic changes.
6 Nov 2026 · 25m · 8 chaptersThe End of Event-Based Cybersecurity, AI Science & World Models | Lightwork
The episode argues the internet is becoming “broken” due to AI-driven, continuous cyberattacks: a major example is a PHP vulnerability where ~70% of websites (including WordPress) were exposed…
30 Oct 2026 · 30m · 13 chaptersAgents Hacking at Scale, Inside the Interpretability Debate & DeepSeek Gets Banned | Lightwork
The episode covers AI agents used for cyberattacks at scale, interpretability claims about neural networks, and fast-moving model releases plus bans.
25 Oct 2026 · 22m · 10 chaptersHow Agents Made Hacking Cheap | Lightwork
AI agents and interpretability, plus model releases and compute/infrastructure economics. Guests: Amber Yang (partner at Lightspeed; co-host).
23 Oct 2026 · 26m · 11 chaptersMuse Wonder: Meta Is Back in the AI Race | Lightwork
The episode covers Meta, Google, and Apple’s latest “frontier” AI moves, plus SpaceX’s push into mobile service and two Lightspeed-universe updates.
16 Oct 2026 · 23m · 8 chaptersShifting From Renting AI to Owning It
Why AI labs should have a presence in Washington, DC, and how the market is shifting from “renting” closed AI models to “owning” downloadable open models; includes an open-weights status update and…
9 Oct 2026 · 12m · 4 chaptersReflection Beam, Mistral Le Chonk & the Drone That Delivers Your Chipotle
Weekly AI and tech news focused on open-weight models (Reflection Beam, Mistral LeChonk), “personal agent” standards, and DoorDash’s drone delivery plus agent features; also covers Aura Ring’s…
9 Oct 2026 · 24m · 11 chaptersThe Difference Between LLMs and Robotics Data | Lightwork
How robotics is shifting from older imitation/RL approaches to visual-language-action models (VLA) and “world/action” models, and how that maps onto a market “board” across model vs hardware and…
6 Oct 2026 · 25m · 17 chaptersThe Week AI Officially Became SI | Lightwork
A roundup of the week’s AI and tech news, centered on the White House “Accord on Superintelligence” and major product launches.
2 Oct 2026 · 32m · 15 chaptersWhy NVIDIA Open Sourced Its Agent Security Layer | Lightwork
NVIDIA’s agent security “full-stack” approach, centered on OpenShell (Apache-licensed) and Sentry (runtime/infrastructure reference stack) to govern/harden AI agents outside the agent process using…
1 Oct 2026 · 17m · 9 chaptersJev: The Model That Can't Talk, Amazon Blocks Meta's Muse & a Robot Teaches Itself Soccer | Lightwork
The episode covers three main AI/tech stories: TypeSafe’s Jev “system one” model, Amazon blocking Meta’s Muse shopping agent, and Skilled AI’s robot that learns soccer by self-play in simulation.
25 Sep 2026 · 32m · 14 chaptersWhy 25,000+ New Yorkers Are Waiting for This Health Scan
Neco Health’s 60-minute, radiation-free full-body scan and why 25,000+ New Yorkers are waiting for its first U.S. clinic in Soho (opening Sept 24).
24 Sep 2026 · 34m · 16 chaptersThe $250B Business Expense Nobody's Been Managing Until Now | Ande on Lightwork
Ande (Andy.ai) is building a corporate entertainment network to centralize and manage enterprise spending on dining, sports, events, merchandising, and catering—an estimated $250B category—by acting…
22 Sep 2026 · 33m · 15 chaptersInside the Tech Powering 500K+ Audio Drama Creators | Pocket FM on Lightwork
Pocket FM founder Rohan Nayak explains how serialized, mobile-first audio dramas are powered by AI creation tools and large-scale recommendation/quality systems, and why “cost to create” dropping to…
21 Sep 2026 · 41m · 19 chapters
AI Labs Debate Safety, Suno V6 Arrives & Bending Spoons’ Miro Deal | Lightwork
The episode covers two main threads: AI safety “pacing the frontier” and major tech deals.
18 Sep 2026 · 30m · 10 chaptersWhat Happens When AI Becomes Your Marketing Team? | Profound on Lightwork
Profound’s Series D raise and how AI agents are becoming marketing teams via AEO/agent-first search, requiring a “canonical AI marketing platform” that creates high-quality, on-brand content using a…
15 Sep 2026 · 15m · 6 chaptersSkipping Two Generations of Accounting Software | DualEntry on Lightwork
Dual Entry’s AI-native ERP and why legacy accounting systems cause long, risky migrations; Lightspeed’s investment; product growth via fast onboarding and AI agents; upcoming “Project Helios” for…
14 Sep 2026 · 33m · 19 chaptersWhy AI Agents Fail in the Real World | Temporal on Lightwork
Temporal explains why AI agents fail in production and how Temporal provides “durable execution” reliability for long-running, real-world actions.
14 Sep 2026 · 25m · 11 chaptersOpenAI’s GPT-6 Astra, Meta Muse & ByteDance Enters the World Model Race | Lightwork
The episode covers (1) OpenAI’s GPT-6 Astra and the shift to longer, more autonomous “computer use” agents; (2) agentic inbox/calendar/wallet assistants (Instinct, Meta Muse, Town); (3) ByteDance’s…
11 Sep 2026 · 36m · 15 chaptersWhy Mistral Just Raised €3B to Build Sovereign AI | Lightwork
Mistral’s €3B (Series D) raise and its strategy for “sovereign AI” via open-weight models plus a full-stack enterprise platform and European compute.
11 Sep 2026 · 17m · 11 chapters
Tim Cook’s 15-Year Run, Meta's $18B Teen Safety Deal & 21 Banks Convene on Stablecoins
The episode covers three big tech/finance stories: Tim Cook’s 15-year Apple run and the CEO transition to John Ternus; Meta’s $18B teen safety settlement and new product limits; and 21 banks’…
4 Sep 2026 · 28m · 10 chaptersThe Company Predicting Whether You Will Return Something | Seel on Lightwork
Seel (Seal) explains how its AI operations platform predicts e-commerce returns/refunds and runs post-purchase support, resale, and compliance to reduce merchants’ operational costs.
2 Sep 2026 · 26m · 13 chaptersBuying Insurance Agencies Instead of Just Selling Them Software | Gyde on Lightwork
Guide (an “AI-native” insurance/wealth/health brokerage) explains why it buys health insurance agencies instead of selling them software, and how its AI agent (GIA) and platform (GuideOS) aim to…
1 Sep 2026 · 56m · 25 chaptersWe use cookies.They keep VO running smoothly and help us improve it. Privacy policy