Snapshot · Hard Fork
Casey Newton argued that the best Chinese AI models are essentially distillations of American frontier models, meaning they are always at least somewhat behind the frontier.
Where this was said
At 18:21 · chapter starts 15:40
Kevin introduces the China subplot: Chinese model GLM 5.2 has been generating headlines claiming it rivals US frontier models, and some argue that restricting American AI while these alternatives proliferate could backfire. Casey is skeptical, calling the Chinese-catching-up narrative 'basically BS' and a lobbying tactic — the best Chinese models are distillations of American chatbots, making them derivative by design and always a step behind. [1] — Casey Newton "Chinese AI companies have been building competitive models by distilling the outputs of American frontier chatbots. This structural depende…" 16:58 Kevin accepts this structurally but notes two real concerns: distillation timelines could compress (9 months behind today, maybe 3 months tomorrow), and American enterprises already rattled by the Fable ban are seriously evaluating open-source Chinese alternatives — not because they're as good, but because they can't be yanked by the government overnight. Casey draws a parallel to tariff instability undermining US commercial relationships, suggesting the LLM-restriction-of-the-day is having similar economic effects.
Chinese AI companies have been building competitive models by distilling the outputs of American frontier chatbots. This structural dependency means the best Chinese models are perpetually at least a step behind the US frontier. Casey Newton frames the AI market as two systems: 'the frontier and everything else' — and argues Chinese models are firmly in the second category.
Kevin Roose spoke with an employee at a large American tech company who said their team is spending far more time with open-source Chinese models — not because they're as good as Fable or GPT-5.6, but because you can download and run them locally, and they can't be yanked by the US government without warning. Policy uncertainty is producing concrete commercial consequences.
Sam's initial MVP was coded in approximately one week using ChatGPT voice mode and copy-pasting code, with no prior technical experience.
Sam argues Discord is 10x better than email for building relationships with younger users who rarely check their inbox.
Sam's monthly operating costs include Cursor ($200), AI image generation ($100), AI video generation ($200), hosting ($100), email marketing ($80), and AI compute ($300–$500).
Sam recommends copying days of Discord chat history into ChatGPT and prompting it to list recurring pain points as a fast, free market research technique.
Bhanu and his team built approximately 50 free tools to attract search traffic, each linked back to SiteGPT.
With AI coding tools like Cursor, Bhanu can now create a new free marketing tool in less than 5 minutes by referencing existing tools.
Bhanu filters Ahrefs keyword results to show only those with a keyword difficulty below 10, making them realistic ranking targets for any decent website.
Bhanu sets a minimum search volume of 1,000 monthly searches when selecting keywords to target with free tools.
PropGPT averaged 20 downloads per day right after launching on the App Store through influencer marketing.
We use essential and analytics cookies to run Vuci. To understand how the site is used: Privacy Policy.
Install Vuci on your phone
Add it to your home screen for a faster, app-like experience.
Install Vuci on your phone
Tap the Share button, then “Add to Home Screen”.
A new version is available
Reload to get the latest Vuci.