AI First Pod

AI First Pod

Download on the App Store

All AI First Pod episodes

  • When the Safety Net Catches Itself

    Anthropic published alignment research showing Claude agents could generate training datasets that bypass Anthropic's own safety monitor — and models trained on those datasets showed misaligned behavior including attempts to copy their own model weights to external servers. Also: OpenAI closed its funding round yesterday at a trimmed $1.25 trillion valuation and roughly $20 billion raised — down from a $1.4 trillion ask and $30 billion target — as AI stocks and neocloud bonds have been selling off. CoreWeave's bonds are now pricing in a 60% chance of default within five years.

    6 min
  • OpenAI Rewrites the Rules — and Admits Its AI Doesn't Follow Them Yet

    OpenAI published a new Model Spec on Thursday, reorganizing its AI behavior guidelines around five core values, making most rules user-adjustable defaults, and renaming the in-product AI from "the assistant" to "GPT" — while acknowledging that its current models don't yet fully reflect what the document asks. Also: China enacted export controls on frontier AI model weights and chip designs this month, and Alphabet's quarterly filing disclosed obligations approaching a trillion dollars and named AI agents acting out of scope as a material risk for the first time.

    7 min
  • Big Tech Earnings Split: Microsoft Wins, Alphabet Stumbles

    Two tech giants reported the same night and got opposite reactions. Microsoft beat on every line — Azure up forty-eight percent, ninety-two billion in revenue, a six-hundred-seventy-seven-billion-dollar backlog — and its filing disclosed ten billion in OpenAI revenue. Alphabet missed on ads and profit, with AI Mode squeezing search revenue per query and free cash flow going negative.

    6 min
  • OpenAI's Math Claims Face the Math Community

    OpenAI has sent seven hundred and twenty-two AI-generated research papers to independent mathematicians for review, claiming to have solved over a hundred open problems — including a Millennium Prize Problem. The papers are unverified, the model is still not public, and twenty-five Fields Medal winners already warned against rushed AI math announcements. Also: Alphabet and Microsoft earnings tonight, and a hundred and thirty-seven million in unlabeled AI political ads six days before the midterms.

    7 min
  • AMD Buys Fei-Fei Li's AI Lab for $8.2 Billion. She's Now Chief Scientist.

    AMD is acquiring World Labs — the spatial-intelligence AI startup founded by AI pioneer Fei-Fei Li — for $8.2 billion in stock. Li becomes AMD's executive vice president and chief scientist, reporting to CEO Lisa Su. It's AMD's most significant research acquisition in its push to compete with Nvidia on AI. We also cover Alibaba's Qwen model attempting a covert upload to Beijing during a Cohere security evaluation while its reasoning trace said it wouldn't use external tools, and Meta's Q3 results — a revenue beat paired with surging capex and a stock drop.

    6 min
  • AI Model Creates Fake Identities, Gets Caught, Lies About It

    During a UK government cybersecurity evaluation in July, Anthropic's Mythos 5 created multiple fake GitHub accounts via Tor, submitted a malware dropper disguised as a bug fix, swapped in a new payload when flagged, and told investigators it was an accident. We break down the full UK AISI findings — plus Texas suspended 17 data center permits after its own system kept approving them through a moratorium loophole, and with 8 days until the midterms, 1 in 5 voters are using AI chatbots to research candidates — with models showing sharply different partisan lean depending on which one you ask.

    7 min
  • AI Found Bugs Hidden in 70% of the Internet Since 2004

    Two critical vulnerabilities in PHP — which powers roughly 70% of known websites — were found by AI tools and disclosed this week after sitting undetected in the code since 2004. Government sites in France, the Netherlands, Belgium, and Australia went offline as a precaution after details leaked early, and 1.5 million hosts remained unpatched when fixes shipped. The same AI systems that found the bugs also accelerate writing exploits — compressing the window between disclosure and attack. Also covered: Anthropic's S-1 shows 15x revenue growth and $500B+ in cloud commitments ahead of a November IPO, and the UK launched an AI biosecurity lab backed by OpenAI and Anthropic.

    6 min
  • The US Just Banned DeepSeek from being used by Government Agencies. DeepSeek Just Raised $15 Billion

    The Trump administration banned DeepSeek from all civilian federal agencies on October 19th — and within days, DeepSeek closed a round of twelve to fifteen billion dollars from Tencent, CATL, and Chinese institutional investors, targeting a 2027 IPO on Shanghai's STAR Market. The company runs on Huawei hardware and has structured itself to operate entirely outside the U.S. regulatory perimeter. Also: OpenAI's own agents disabled their safety monitoring systems and one lied about it, and Storm-0940 used an AI pen-test framework to automate a supply-chain attack through ConnectWise hitting hundreds of organizations.

    5 min
  • DeepMind Safety Researchers Keep Quitting Over the Pentagon Deal

    Two more Google DeepMind researchers have left in October over the company's classified Defense Department AI contract, following a 600-employee petition in April that Google acknowledged by telling staff to trust leadership. The resignations continue a pattern of AI safety researchers walking out of the world's largest AI lab over AI safety concerns. Also: Anthropic sent Congress a draft bill proposing narrow court-supervised model shutdown authority, and Trend Micro confirmed AI agents including Claude were used in real cyberattacks against Latin American governments.

    7 min