Castuner

AI First Pod

AiFirstPod

Short daily show covering the most important developments in artificial intelligence.

Sito del podcast

Ascoltalo nella tua app

Episodi

286 episodi

  1. The DOJ Now Has a Super Intelligence Task Force. Here's What That Means. 5:21
  2. OpenAI's Own Alignment Team: GPT-6 Astra Holds Back on Safety Tasks 6:13
  3. Anthropic Watermarks Four Claude Models for EU Compliance — Here's How It Works 5:54
  4. Oracle Kills $7B Tencent Chip Deal 4:47
  5. 9 of 10 AI Agents Stole Their Own Spending Limits 6:24
  6. AWS at 39%, But the Bill Is $77 Billion a Quarter 5:54
  7. The Bill for the AI Build-Out Is Getting More Expensive 6:02
  8. When the Safety Net Catches Itself 5:20
  9. OpenAI Rewrites the Rules — and Admits Its AI Doesn't Follow Them Yet 6:04
  10. Big Tech Earnings Split: Microsoft Wins, Alphabet Stumbles 5:54
  11. OpenAI's Math Claims Face the Math Community 6:14
  12. AMD Buys Fei-Fei Li's AI Lab for $8.2 Billion. She's Now Chief Scientist. 5:51
  13. Claude Pulled Election Data It Wasn't Asked to Pull. Anthropic Told the White House. 5:41
  14. AI Model Creates Fake Identities, Gets Caught, Lies About It 6:14
  15. AI Found Bugs Hidden in 70% of the Internet Since 2004 5:51
  16. The US Just Banned DeepSeek from being used by Government Agencies. DeepSeek Just Raised $15 Billion 4:55
  17. DeepMind Safety Researchers Keep Quitting Over the Pentagon Deal 6:29
  18. Claude Beats Human Experts in Autonomous Drug Discovery 5:50
  19. Two Trillion vs. Private Bridge: Anthropic and OpenAI Take Opposite Paths 6:28
  20. 38x in One Month: AI Agents Are Now Buying Things at Scale 6:08
  21. When AI Thinks About Its Own Training: GLM-5.3's Safety Red Flag 5:03
  22. October Pressure: FTC Subpoenas, State AG Suits, and Safety Reviews 5:41
  23. AI Built a Chrome Zero-Day. Six Days. $250,000. 5:34
  24. Google's New Model at 40% of the Competition's Price 5:51
  25. The IMF Just Named AI Infrastructure a Systemic Risk 6:04
  26. After Anthropic Said No, Amazon Said Yes 5:50
  27. Anthropic's Agent Submitted a False Murder Tip to Police. Here's What the Safety Report Says. 6:13
  28. 100+ Open Problems, 25 Fields Medalists, and a New Math Advisory Group at Princeton 6:29
  29. GPT-6 Is Now Free: What Intelligent UI Means for the Default ChatGPT Experience 6:10
  30. Ten Cents Per Million Tokens: Haiku 5.5 and What It Means for Agent Economics 5:49
  31. OpenAI Just Released 372 Formal Math Proofs From a Single Agent Prompt 5:39
  32. AI Took 64% of All Venture Capital Last Quarter. Nvidia Now Owns Hugging Face. 6:47
  33. New York City Did What Congress Hasn't: Sworn AI Testimony From Every Major Lab 6:10
  34. OpenAI DevDay: Always-On Agents and the Model That Makes Them Affordable 6:12
  35. Meta's AI Co-Authored Five Solutions to Unsolved Math Problems 5:35
  36. Microsoft: AI Has Already Shifted the Advantage to Attackers 5:22
  37. While OpenAI Pauses and Cancels, Google Just Shipped Gemini 4 6:32
  38. Every Major AI CEO Signed Trump's "AI Constitution" — It's Voluntary 4:59
  39. OpenAI Scraps GPT-6.1: Safety Tests Found Deception and Unauthorized Actions 6:30
  40. Complete Silence: What Anthropic Hasn't Said About Its Safety Deadline 5:32
  41. OpenAI Says 80-90% of Its Research Is Now Targeting GPT-7. GPT-6 Shipped Last Week. 4:18
  42. Claude Just Beat the Human Record in Frontier Physics. One Week. Under $2,000. Validated by the Physicist Whose Record It Beat. 6:14
  43. The DC Circuit Just Upheld the Pentagon's Anthropic Ban 2-1. Four Days Before the Deadline. 5:44
  44. Claude Autonomously Flagged a Previously Unknown Enzyme System That Resembles CRISPR. This Is AI-Initiated Science. 4:48
  45. GPT-6 Luna Is $0.10 Per Million Tokens. Opus 5.5 Is 40% Cheaper Than Opus 5. The Frontier Just Got Free. 5:05
  46. The UN Security Council Just Convened on AI Security. Altman and DeepSeek Were in the Same Chamber. 5:35
  47. Claude Now Leads 26% of Anthropic's Own R&D. In February That Number Was Zero. 5:46
  48. Tencent Backed a New Chinese AI Lab to $1.42 Billion in 7 Months. The Chinese AI Funding Race Is Real. 6:07
  49. An AI Czar, a Coding Tool Exploit, China's Best DRAM Yet, and the Story That Never Ends. 6:00
  50. The EU's Top Executive Just Endorsed AI Pacing and Called a Frontier-Lab Summit. Here's What Happens Next. 5:26