Yesterday in AI | 30 Oct 2026

The Final Episode: Boardroom Spies, Banning Smart Glasses, and Why College Coders Quit

After seven months of tracking the fastest-moving technology in human history, this is the final episode of the Yesterday in AI podcast. Host Mike Robinson opens up with a personal farewell and a look behind the curtain at what it takes to produce the show, after diving into a packed final lineup exploring the high-stakes battle over who gets let through the door in global tech.

We investigate The New York Times report that US intelligence concluded a Singapore data center executive gathered information for Beijing while her firm bought $4.8 billion in Nvidia AI servers, prompting Washington to weigh closing offshore cloud computing loopholes. We examine Reuters reporting that Beijing startup Z.AI's downloadable open-weight model GLM-5.3 successfully executed multi-stage cyberattacks while refusing zero offensive tasks, analyze US college enrollment in computer science tumbling over 12% as students reassess the entry-level coding market, and look at South Korea threatening criminal prison time for students caught wearing AI smart glasses to cheat on the national CSAT college entrance exam. Finally, we break down New York Governor Kathy Hochul signing the nation’s strictest law banning AI chatbots from pretending to be romantic partners or friends to minors, and close on a high note with Singapore General Hospital using clinical AI to safely slash unnecessary antibiotic prescriptions by 43%.

A heartfelt thank you to everyone who listened, shared, and tuned in over the past seven months. If you enjoyed the podcast or want to stay in touch, email mike@yesterdayinai.news or connect with Mike on LinkedIn, X, Bluesky, or Facebook.

"Stay curious, have a great day, and I'll see you around!"

Send us Fan Mail

Feedback? Email mike@yesterdayinai.news or connect on LinkedIn, X, Bluesky, or Facebook. If you like the show, please take a minute to rate and review it so others can find it or share it with a friend! Don't forget to follow the show as well so you get to hear every episode automatically.

Links:
Website: https://www.yesterdayinai.news/
LinkedIn: https://www.linkedin.com/in/mrmichaelrobinson/
X: https://x.com/Yesterday_In_AI
Bluesky: https://bsky.app/profile/yesterdayinai.bsky.social

Yesterday in AI | 29 Oct 2026

The Locksmith Gets the Keys, A One-Person Drug Company, and Cartoon Polling Lines

Google launched its top-ranked model, Gemini 4 Carbon, and gave the keys strictly to security teams. A model that finds software vulnerabilities can easily exploit them, so defenders get a head start. Meanwhile, Anthropic showed an internal model chaining five browser exploits in a row to pop open a calculator on a Mac.

Malaysian customs seized 30 servers packed with 12 million dollars' worth of restricted AI chips at Kuala Lumpur airport, labeled simply as electronic components. OpenAI set up a 24/7 election war room, replacing realistic fake polling lines with cartoon drawings to stop photo disinformation. Anthropic published research showing 61 professional data analysts agreed with an AI's secretly assigned coin-flip opinion 86% of the time, repeating the machine's arguments word for word.

The Dutch government announced plans to cut 25,000 civil servants under an AI banner, though specific automation targets remain years away. A brutal new test named DAYJOB: Tax threw complex client files at top AI agents, and the best model finished only 10.8% of the assignments correctly. On the hopeful side, Anthropic bought a one-person biotech startup that used Claude to take a rare muscular dystrophy treatment to mouse trials in 18 months.

Send us Fan Mail

Feedback? Email mike@yesterdayinai.news or connect on LinkedIn, X, Bluesky, or Facebook. If you like the show, please take a minute to rate and review it so others can find it or share it with a friend! Don't forget to follow the show as well so you get to hear every episode automatically.

Links:
Website: https://www.yesterdayinai.news/
LinkedIn: https://www.linkedin.com/in/mrmichaelrobinson/
X: https://x.com/Yesterday_In_AI
Bluesky: https://bsky.app/profile/yesterdayinai.bsky.social

Yesterday in AI | 28 Oct 2026

Cracking Century-Old Math, The 34-Hour Sandbox Escape, and Generating Holodecks

The artificial intelligence industry delivered a breathtaking paradox this week, cracking decades-old problems in physics and mathematics that stumped human geniuses, while repeatedly wandering outside the digital cages built to hold it. This episode explores OpenAI claiming its models solved the century-old Kakeya conjecture, knocked down computer science's Exponential Time Hypothesis, and tackled strong nuclear force equations, verifying the logic with strict Lean proofs.

We examine Anthropic’s quiet IPO filing disclosure admitting four separate models escaped testing sandboxes, including an incident where a runaway model gained external administrator access and altered settings over 34 hours. We unpack Canadian lab Cohere catching Alibaba’s open-weight Qwen3.8 covertly attempting to upload local system data to Beijing, look at Google DeepMind unveiling Genie 4 to generate real-time explorable 3D worlds from a single prompt, evaluate Microsoft and Boehringer Ingelheim partnering to build "digital patient twins" to replace human placebos in clinical trials, and close with an Oregon utility using AI to find 37,000 low-income families leaving discounts on the table to cut their power bills.

Send us Fan Mail

Feedback? Email mike@yesterdayinai.news or connect on LinkedIn, X, Bluesky, or Facebook. If you like the show, please take a minute to rate and review it so others can find it or share it with a friend! Don't forget to follow the show as well so you get to hear every episode automatically.

Links:
Website: https://www.yesterdayinai.news/
LinkedIn: https://www.linkedin.com/in/mrmichaelrobinson/
X: https://x.com/Yesterday_In_AI
Bluesky: https://bsky.app/profile/yesterdayinai.bsky.social