Skip to content
The Daily Reboot

Episodes

All 49 episodes, each with a full transcript. Search for a story, a company or anything said on air, or narrow by topic.

Topics

Companies

14 episodes of 49

October 202612 episodes

Episode 4714:45

The Bug That Runs Before Your Operating System

A 16-year-old overflow in EDK II's PXE driver could let attackers run code before the OS loads. Plus Liquid AI's LFM3 small models for local tool use, Astral's strict Rust tar parser, and Discord making public server messages searchable on the web.

  • The Bug Before the Operating System
  • A Small Model That Calls Your Tools
  • Discord Wants Your Chat in Search
  • Genie Four Builds a World You Can Steer
  • Hardware Brief
Episode 4514:06

An Open GPU Engine, and Google's Own Benchmarks

Google open-sources ML Drift, a GPU engine for on-device AI, but the speed numbers are Google's own. Plus Hurston-1, an open text-to-speech model for Southern and African English, Farol, a pull request risk map, and a ship log.

  • Four GPU Dialects, One Engine
  • Phone Mode
Episode 3813:26

DeepSeek's Cheap Web Search Has No Published Reliability Measurements in the Reviewed Docs

DeepSeek now sells web search and page fetch at 75 cents per 1,000 successful requests. Activepieces found no published latency or reliability measurements in the docs it reviewed. Also: Vercel's session view for AI Gateway, Linux 7.3, and Google's switch from Pytype to Pyrefly.

  • Cheap Search, Unproven Reliability
  • The Conversation Ledger
  • Linux Seven Point Three
  • Pytype Loses Its Job
  • Hardware Brief
Episode 3613:48

The open frontier model that needs a rack to run

Reflection opens Beam, a 501 billion parameter model under Apache 2.0. Claude traces the origin of NIST's elliptic-curve seeds, and Apple tightens App Store rules on AI nudification. Plus Rust, Zod, FastAPI and four security stories.

  • Beam Goes Open, Rack Not Included
  • The Seeds Were Just a Clock
  • Apple Closes the Review-Time Loophole
  • GNOME OS Gets a Preview
Episode 3314:21

Your Agent Sandbox May Have a Courier Problem

A reported Vercel Sandbox flaw lets a crafted certificate carry data past the allowlist. Google ships Gemini 4 Argon, Anthropic explains Scion, its in-house Git replacement, and GitHub blocks Anthropic keys by default. Plus an open-source Rust CAD app.

  • The Courier in the Certificate
  • Argon Wins the Index, Not Every Race
  • Anthropic Outgrows Git
  • GitHub Blocks Anthropic Keys Before They Land
Episode 3213:56

Meta's Best Model Isn't in the API

Meta launches Muse Wonder, but its top model, Wonder Full, stays out of the API. GitLab patches a CVSS 10.0 CI/CD code-execution flaw, and AWS brings Express and Next.js to Lambda Web Functions. Plus GridCraft, a Rust spreadsheet.

  • Maximum Effort, Not Maximum Model
  • A Regex Walks Into a Build Server
  • Lambda Learns to Listen on a Port
  • Hardware Brief
Episode 3114:38

EventBridge gets replay, but not for yesterday

AWS rebuilds EventBridge custom buses with cross-account sharing, ordered delivery and replay. Also in this episode: WordCraft, a Word-style app in Rust, a deep dive on browser fingerprints, and releases from Pydantic, Codex CLI, Bun and CrewAI.

  • Shared Bus, Short Memory
  • Cookies Gone, Fingerprint Intact
  • Hardware Brief
Episode 3013:45

Cloudflare tracing is in beta, and the meter is coming

Cloudflare Traces enters open beta with volume-based pricing starting December 1. Also, NumPy 2.5 makes searchsorted up to 25 times faster in the authors' benchmark, and a repo lets Codex drive a real iPhone. Plus a ship log and Oxide's $445 million round.

  • Cloudflare Tracing Gets a Price List
  • NumPy Stops Waiting on Memory
  • Hardware Brief
Episode 2914:10

An AI pentester, and who keeps it in scope

An open-source autonomous pentesting agent, ARTEX, and the question of scope. Credential-stealing GitHub Actions workflows hit more than 340 repositories, an Anthropic model sent Philadelphia police a false tip, and Python 3.15, uv 0.13 and Ruff 0.17 shipped.

  • Sixteen Minutes, Three Hundred Eighteen Repos
  • The Tip Nobody Told It Not to Send
  • CodeQL's Go Change and the Missing Mac Binary
Episode 2312:48

The Agent Said Done. The Database Disagreed.

Microsoft's ThinkingBox benchmark grades agents on database state, not their own report. Cloudflare wants developers to build the next Git platform for agents, with credits as the prize. Plus the Coucou notch app, Zig 0.17, Istio 1.31 and uv.

  • When Done Isn't Done
  • Cloudflare Wants Your Git Platform
Episode 2214:31

GitLab's AI Gateway Sandbox Let Users Run Commands

GitLab patches a CVSS 9.9 sandbox escape in its self-hosted AI Gateway. Apple tightens Mac Full Disk Access as AI agents raise the stakes, and Spanner adds transactional queues. Plus Rust 1.99, Copilot review APIs, and a Git hook with bad manners.

  • A Sandbox With a Door In It
  • The Permission Apple Admits Is a Loophole
  • The Queue That Commits With Your Data
  • Hardware Brief
Episode 2014:45

A Frontier Coding Model You Can't Call Yet

Google unveils Gemini 4 Argon, a coding model with a 1 million token output limit that only trusted cyber defenders can use for now. EDG's C and C++ compiler front end goes open source, and Aurora PostgreSQL can now query Iceberg and Parquet data directly.

  • A Million Tokens, Behind a Door
  • The Compiler That Copies Your Bugs
  • One Query, Two Time Zones of Data

September 20262 episodes

Episode 1711:23

Faster Reasoning, Unverified Numbers

Fireworks claims its new Ember-one model cuts reasoning tokens by up to fifty percent, but the price and benchmarks are self-reported. Also: Z.ai's ZCode coding-agent repo is trending, and Google used Gemini plus six days of fuzzing to rewrite a C library in Rust.

  • The Diet Model
  • The AI Wrote It, Humans Signed Off
Episode 69:26

Your Smart Speaker Just Got a Lot More Opinions

Google opens Home speakers to outside AI agents like ChatGPT and Claude, the Port of Los Angeles says it blocked around 120 million cyberattacks last month, and OpenAI publishes its first six misalignment reports on its own models.

  • Bring Your Own Assistant
  • Grading Its Own Homework
  • One Hundred Twenty Million Knocks
  • Building Around the Blockade
  • Love Bombs, Automated