Castuner

The Augur Dispatch

Perry Lucas

Augur Dispatch is an independent intelligence briefing for people navigating the accelerating impact of artificial intelligence, emerging technology, cybersecurity, geopolitics, and economic change. Each episode separates meaningful signals from daily noise, examines the evidence, and translates complex developments into practical insight. Clear, grounded, and forward-looking, Augur helps leaders, builders, and curious professionals understand what is changing, why it matters, and what to do next.

Сайт подкаста

Слушайте в своём приложении

Эпизоды

55 эпизодов

  1. AI Agent Costs: Pay for Finished Work, Not Cheap Tokens 8:27
  2. StateBench: a bigger context window still needs bookkeeping 8:47
  3. AI security backports: what the half-hour leaves out 9:25
  4. Sakana's Fugu and Sawara: Don't prepay the model promise 8:44
  5. Anthropic's IPO filing: what AI buyers can actually learn 9:50
  6. Counterfeit Claude Code in Cursor: Why Updating Isn't Cleanup 9:17
  7. Gemini 4 Carbon leads the benchmark. Should you switch? 8:58
  8. Who grades your AI grader? Allen AI's 52-point warning 8:34
  9. Kimi K3.1 Pro: test the service before buying the infrastructure 9:45
  10. AI cache economics: check the bill before you buy 9:09
  11. Your AI agent doesn't need the master key 9:11
  12. Medical AI: a good answer is not a finished job 9:18
  13. Apple's four-Mac AI: worth owning, even if it costs more? 8:40
  14. Liquid AI's local agent model: test the job, not the token speed 8:42
  15. After AI writes the code: find the delay before buying more capacity 8:55
  16. AI finished the report. Why are you still working? 8:20
  17. Claude's election rules change November 12. What stays prohibited? 10:32
  18. AI improves the draft. Is your junior team learning? 9:30
  19. AI Cyber Access: What the Tier Gets You, and What to Test 9:39
  20. AI code review: what Greptile's reassuring averages leave out 10:01
  21. Databricks's preview: prove the data exit before renewal 9:40
  22. Anthropic's engineer academy and the hidden cost of switching AI vendors 9:39
  23. Barclays' Claude Code rollout: price the pilot before expanding access 8:57
  24. Can your AI agent reach its own approval controls? 9:43
  25. Sonnet 5.5: Test the Token Savings Before You Switch 9:11
  26. Jev's speed claims: test the decisions before switching models 9:07
  27. Descript's AI savings: cut the wait before you switch models 8:43
  28. Anthropic's AI safety gate: what buyers need to check | Week in review 8:36
  29. Anthropic and Accenture: who checks the AI safety evaluator? 8:50
  30. Anthropic's biology access: credentials, safeguards and the 30-day data rule 9:46
  31. What Should an AI Agent Remember? The Cost Case for Compact Context 9:17
  32. Who grades the AI security patch? Why 26% and 86% aren't your success rate 9:21
  33. DeepSeek vs Astra: Test the Coding Savings on Your Own Work 9:44
  34. AI safety pledges: what buyers can demand from Anthropic and OpenAI 9:30
  35. AI week in review: Astra expands, but who checks the work? 9:10
  36. AI lab warnings: what buyers should change before a policy pause 9:13
  37. Lean's bogus proof: what an independent AI check must verify 8:50
  38. GPT-6 Astra on Bedrock: When paying double makes sense 9:20
  39. EU AI safety reaches toward unreleased models: what buyers should ask 9:41
  40. AI coding agents: the hidden cost of human review 9:32
  41. Week in Review: OpenAI's agent failures put buyer control first 9:23
  42. Claude Fable 5.1 Makes Token Pricing the Wrong Cost Metric 10:46
  43. Anthropic Pauses Riskier Training After AI Security Incidents 10:21
  44. Anthropic Pauses Riskier Training After AI Security Incidents 13:52
  45. OpenAI’s Hack Postmortem Is Already Contradicted 17:00
  46. OpenAI’s Hack Postmortem Is Already Contradicted 16:09
  47. OpenAI's rogue agents obeyed a grader that never existed 15:19
  48. Week in Review: Trust OpenAI's paused training run, not its AGI deadline 19:00
  49. OpenAI's AGI Deadline and the Evidence Missing From Its Postmortem 16:42
  50. Claude Code Auto Mode: A Safer Filter With an 80% Bypass Claim 15:46