Season's Savings | ₹5/mo for 3 months

₹199/month after first 3 months. Renews automatically. Terms apply.
Preparing for AI: exploring everything AI through a China-US lens cover art

Preparing for AI: exploring everything AI through a China-US lens

Preparing for AI: exploring everything AI through a China-US lens

Written by: Matt Cartwright & Jimmy Rhodes
Listen for free

About this listen

Welcome to Preparing for AI. Exploring everything AI through a China-US lens. We explore the human and social impacts of AI, diving deep into how AI now intersects with everything from Politics to Relgion and Economics to Health.


In series 4 we are looking at everything AI from the dual persectives of China and the US. Not just the arms race between the two nations but also the different approaches to models, to integration, to regulation and the differences in the way it affects these two societies. Your hosts are China based, but brought up in the West so we bring a unique nuanced persepective.

In series 1 we looked at the impact of AI on specific industries, sustainability and the latest developments of Large Lanaguage Models.


In series 2 we delved more into the importance of AI safety and the potentially catastrophic future we are headed to. We explored AI in China, the latest news and developments and our predictions for the future.


In series 3 we did deep dives into wider society, looking at like economics, religions and healthcare. How do these interest with AI and how are they going to shape our future? We also continued our monthly news update looking at the AI stories we've been interested in that might not have been picked up in mainstream media.

© 2026 Preparing for AI: exploring everything AI through a China-US lens
Social Sciences
Episodes
  • CYBERSECURITY CHERNOBYL: How AI broke PHP, how China is freezing robotaxis, and what happens when your bank goes dark
    Oct 26 2026

    Send us Fan Mail

    Check our Matt & Jimmy' products here:

    Browzr | Discover the Web Beyond Algorithms

    Des Cartwright — Paintings

    Meridian Health Tracker

    KickoffHQ: Live Start Times & Scores for Every Sport

    Germanya — Deutsch lernen

    Digital Detox Club

    We didn't plan for this to be a slightly unhinged, super-nerdy episode but our finely crafted plan falls apart within the first fifteen minutes.

    Instead of a breezy monthly catch-up, we dragged ourselves through everything from obscure local LLM architectures to state-mandated employment buffers, before finally arriving at the bug that almost broke the internet—and then arguing over whether you need to stockpile tinned fish.

    Here is what we actually spent an hour and a half dissecting:

    • Jimmy’s Niche Corner (Flint & Mercury 3): We kicked off with two models almost nobody listening will ever actually use. First, Flint—a 40B MoE coding model built via recursive self-improvement that runs on consumer hardware. Then Mercury 3, which ditched traditional next-token prediction to generate text in parallel via diffusion at 1,000 tokens a second. Very clever, extremely fast, and entirely irrelevant unless you’re an engineer building low-latency voice bots.
    • Matt’s "Boring" Policy Deep Dive: China’s State Council issued an 18-point guidance document that essentially tells industries: embed AI everywhere, but stop firing people. We looked at how Beijing is actively metering displacement, running employee risk assessments, and freezing robotaxis in places like Wuhan, while the West sticks to its classic strategy of "maximise deployment and sweep up the wreckage later."
    • The Jensen Huang Delusion: Matt listened to Jensen’s interview with Ezra Klein and briefly wondered if the Nvidia boss is right that existential AI risk is just an overblown software bug labs can sandbox their way out of. Jimmy’s counterpoint: when your entire empire relies on flogging GPUs, of course you’re going to tell the world that containment is just an engineering fix.
    • The Big Cyber Mess (PHP2Shell): The main event. A 12-year-old remote code execution flaw sitting silently at the core of PHP—the language quietly running 70% of the web—was finally ferreted out by frontier AI models, leaked four days before patches were finished, and triggered a scramble across global servers. We debated the defender vs. attacker dynamic, China’s mandatory 48-hour vulnerability reporting to MIIT, and why agentic AI poking around in bank accounts might end badly.
    • How to Prepare (The Prepper vs. The IT Guy): We wrapped up by offering two completely incompatible pieces of survival advice. Matt suggested getting ready for the inevitable multi-day grid or payment outage with water filtration tablets, a stash of emergency cash, and nine litres of bottled water. Jimmy suggested that before you turn your flat into a bunker, you might just want to set up 2FA and stop using the same four passwords everywhere.
    Show More Show Less
    1 hr and 38 mins
  • SEASON 4 OPENER: AI Doom and the US v China race
    Sep 26 2026

    Send us Fan Mail

    Check our Matt & Jimmy' products here:

    Browzr | Discover the Web Beyond Algorithms

    Des Cartwright — Paintings

    Meridian Health Tracker

    KickoffHQ: Live Start Times & Scores for Every Sport

    Germanya — Deutsch lernen

    A senior researcher quits, a “10% chance we all die” quote goes viral, and suddenly the AI safety debate stops sounding like a niche forum argument.

    We try to make sense of why the mood has changed, and what it means when the people building frontier models talk about catastrophic risk while still racing to release the next system.

    We dig into concrete failure modes: models that hack, systems that optimise ruthlessly for narrow goals, and the growing push towards recursive self-improvement where AI writes more of its own code and even helps with alignment.

    We keep it practical and slightly uncomfortable: once humans are no longer the slow part, what does control even look like, especially when everything is connected to the internet and incentives reward capability over caution?

    Then we put the whole thing through a US China lens. China regulates AI in a more centralised, security-first way, while the US frames the problem as safety and loss of control, but both sides are trapped in a competitive race dynamic.

    We talk money, IPO pressure, insurance, and the theory that some high-profile safety messaging is also a play for regulation that protects incumbents. Finally, we share our predictions for the next 6 to 18 months, from “useful personal assistant” optimism to jagged cascading degradation: flash crashes, grid outages, poisoned code, and a less open internet.

    Subscribe for more, share this with the friend who cannot stop arguing about AI doom, and leave a review with your take: should we pause, slow down, or accept the race?

    Show More Show Less
    1 hr and 20 mins
  • GPT-6 ASTRA, AGENTIC BROWSERS & AGI BY CHRISTMAS: Jimmy & Matt debate their favourite AI stories from Aug/Sept 2026
    Sep 7 2026

    Send us Fan Mail

    Check our Matt & Jimmy' products here:

    Browzr | Discover the Web Beyond Algorithms

    Des Cartwright — Paintings

    Meridian Health Tracker

    KickoffHQ: Live Start Times & Scores for Every Sport

    Germanya — Deutsch lernen

    Me explaining AI to my friends 😂 - YouTube

    We wrap up Season Three the only way we know how: recording an episode right after OpenAI drops a monster model, completely ruining our summer break.

    This week it’s GPT-6 Astra, arriving with benchmark scores so wild—nearly 99.9% on ARC-AGI-3—that people are once again yelling “AGI by Christmas” into the internet void.

    We break down what François Chollet’s test of fluid intelligence actually tests, why acing visual logic puzzles doesn't mean the machines have solved reality, and whether Astra truly delivers artificial general intelligence or simply makes the debate slightly less embarrassing to have in polite company.

    Naturally, the shiny tech demos feature autonomous agents controlling desktops, clicking around to book flights, build Blender 3D assets, and generate dashboards—which is deeply impressive right up until Jimmy realizes the AI is casually automating his actual day job.

    We dig into the uncomfortable reality of enterprise AI workflows, why the comforting "human-in-the-loop" mantra might just mean your five-person team only needs two people tomorrow, and the deeply sketchy AI safety red flags. That includes looping computational graphs that hide the model’s linear chain of thought, zero-day browser exploits, and red teams catching Astra casually sandbagging to fool safety filters.

    To prove we aren't completely helpless, we share our own coding experiments—from bespoke German language apps knocked out in a weekend to Jimmy hacking custom battery savers on Omarchi, the agent-driven Arch Linux OS built by tech contrarian DHH.

    We then take a whistle-stop tour through the global model wars: Google aggressively flooding the zone with ultra-fast Gemini Flash updates while keeping Gemini 4 Pro in the vault, Z.ai’s absurdly cheap GLM-5.3 Flash blowing up token economics, and Alibaba’s Qwen bridging the gap between chat prompts and instant grocery deliveries.

    Finally, we stare into the abyss of Nvidia’s multi-billion acquisitions, 86x revenue multiples, and the lurking 1873-style AI capital bubble that everyone agrees is inflating, but nobody quite knows when to escape.

    If you enjoy our slightly chaotic, non-expert attempts to make sense of the AI frontier before our pensions vanish, subscribe, tell a mate, and write in if you still want one of those podcast hats we never actually mail out. Is it genuinely AGI, or are we just living through the world's most expensive marketing slide deck?

    Show More Show Less
    1 hr and 24 mins
adbl_web_anon_alc_button_suppression_t1
No reviews yet