

1,145 episodes
by Nathaniel Whittemore · English
A daily news analysis show on all things artificial intelligence. NLW looks at AI from multiple angles, from the explosion of creativity brought on by new tools like Midjourney and ChatGPT to the potential disruptions to work and industries as we know them to the great philosophical, ethical and practical questions of advanced general intelligence, alignment and x-risk.

The episode examines real-world risks from current AI agents, arguing that harms may be less “existential” than the headlines suggest, but still cause significant disruption because cybersecurity and…
28 Sep 2026 · 29m · 10 chapters
The episode argues that US AI discourse has swung toward extremes (doom vs acceleration) and highlights “AI moderates” who seek a thoughtful middle—supporting useful AI while pushing for concrete…
27 Sep 2026 · 33m · 15 chapters
The episode explains Jev (TypeSafe’s “System 1” judgment model) and focuses on real-world ways people are using it beyond demos—fast, cheap yes/no, choice, and scoring judgments applied in parallel…
25 Sep 2026 · 26m · 12 chapters
AI agents moving from chat apps into real-world actions, plus related AI news on biology breakthroughs and global governance.
24 Sep 2026 · 28m · 10 chapters
A same-day roundup of four major model releases—Anthropic Claude Opus 5.5 and OpenAI GPT-6 Sol + GPT-6 Luna—focusing on cost-efficiency, benchmark results, safety/alignment, and early real-world…
23 Sep 2026 · 31m · 8 chapters
“Agent Wars” covers the rise of AI personal agents—especially Meta’s Muse—sparking platform battles over agentic shopping, plus a related discussion of AI safety/liability and a separate KPMG study…
22 Sep 2026 · 31m · 12 chapters
The “state of the AI debate” across the US and China, focusing on regulation (kill switches, data-center rules), US military/industrial moves, and whether China will join any AI slowdown/pacing…
21 Sep 2026 · 27m · 11 chapters
Seven shifts in how people use AI day-to-day: simpler integrated interfaces, “monothreads” (single persistent context), chatbots acting as agent fleet managers, voice-first control, goal/loop…
20 Sep 2026 · 26m · 10 chapters
How enterprises are responding to the recent AI safety debate, and what practical AI challenges they’re prioritizing now (governance, agent security/identity, evals, legacy integration, architecture…
18 Sep 2026 · 30m · 21 chapters
The episode argues that “personal AI agents” are finally moving from hype to consumer adoption.
17 Sep 2026 · 30m · 13 chapters
The episode introduces “AI judgment models” (TypeSafe’s JEV) that output calibrated probabilities for specific questions instead of generating text, and argues they could be cheaper/faster and better…
16 Sep 2026 · 25m · 8 chapters
The episode reviews AI governance and safety debates after Dario Amodei’s “Pacing the Frontier” proposal and lab leaders’ agreement, framed as a negotiation over who controls AI release speed.
15 Sep 2026 · 26m · 14 chapters
Anthropic CEO Dario Amodei’s essay “We Must Pace the Frontier” and a wave of public support from other frontier labs for slowing AI “frontier” progress to allow alignment, safety, and third-party…
14 Sep 2026 · 37m · 13 chapters
How to use “opportunity AI” (unlocking new capabilities) rather than only “efficiency AI” (doing existing tasks better), using GPT-6 Astra and related models as the example.
13 Sep 2026 · 28m · 12 chapters
The episode argues that the newest AI releases are mainly about two near-term shifts: (1) managing more computer interactions via voice (full-duplex, better noise/interrupt handling) and (2) using…
11 Sep 2026 · 32m · 11 chapters
The episode analyzes why two viral X posts by Anthropic/OpenAI researchers claiming AI could kill all humans (over 10% within a decade) suddenly drove major mainstream coverage and political calls…
10 Sep 2026 · 36m · 18 chapters
AI Model Month kickoff and the “Navier-Stokes” Millennium Prize controversy, plus rapid model releases (Google Gemini 3.8 Flash, Meta MuseSpark 1.3, OpenAI ChatGPT Images 2.5) and trust/agentic-work…
9 Sep 2026 · 34m · 12 chapters
GPT-6 Astra (OpenAI) is framed as a “significant and confounding” release: an “opportunity AI” focused on computer use and new interaction patterns (ambient, voice-driven, mouse-free), not just…
8 Sep 2026 · 29m · 9 chapters
The episode argues that AI “agents” are shifting from single-player use (individual agents producing private outputs) to “multiplayer AI” where teams share context, sessions, live steering/hand-offs,…
7 Sep 2026 · 26m · 8 chapters
What “AI-native” means for enterprises as agentic AI begins in 2026, and a 30-feature checklist for redesigning operations from the ground up (process blueprints, context/knowledge architecture,…
6 Sep 2026 · 28m · 19 chapters
Retrospective on how AI changed over the summer, covering frontier model releases and export controls, enterprise adoption shifts (agent costs, routing, open weights), agent management (harness…
4 Sep 2026 · 24m · 9 chapters
How to use “agentic loops” (repeatable, self-checking executions with a concrete, verifiable finish line) for knowledge work, and when to scale from single-agent loops to multi-agent “work graphs”…
3 Sep 2026 · 57m · 22 chapters
Anthropic’s Claude Fable 5.1 (and Mythos 5.1) launch—capability gains, cost/efficiency claims, guardrail and enterprise data-retention upgrades, plus early user impressions and debates about model…
2 Sep 2026 · 31m · 11 chapters
AI Daily Brief covers OpenClaw 2.0 and broader AI security/industry news.
1 Sep 2026 · 26m · 11 chapters