Спикеры
Josh Kale
— Соведущий, подкаст Limitless (Bankless)
The hosts map the crowded frontier AI model landscape across closed US labs and cheap Chinese open-source models. They argue most consumers should pick a $20 subscription to a major lab, while specialty use cases favor Grok for X, ChatGPT for voice/images, Claude for coding/context, and Gemini for music/marketing. They also highlight company-level implications for Google, Microsoft, and Apple and expect aggregator platforms and local open models to become more important.
- Frontier model supply has exploded, with US closed labs losing OpenRouter token share to Chinese open-source models.
- Pricing now matters more than raw intelligence, pushing labs to distill cheaper models.
- Consumers are best served by one major subscription; open-source Chinese models suit high-volume agentic use.
- Grok is strongest for real-time X/news; ChatGPT stands out in voice and image generation; Claude leads coding, writing, and high-intelligence work; Gemini is narrow for music/marketing.
- Google and Microsoft have legacy enterprise distribution moats and AI venture fund-like stakes in OpenAI, SpaceX, and Anthropic.
- Ejaaz expects aggregator platforms to be the next big AI winner; Stripe's OpenRouter acquisition is cited.
- Josh sees Apple's upcoming on-device AI as likely to dominate consumer local AI.