The controversial practice of AI 'distillation' is not IP theft but a modern form of competitive benchmarking. It's akin to how early Google submitted queries to Yahoo to compare and improve its own search results. The focus is on learning from a competitor's public output, not stealing their underlying software or code.
A government ban on open-source AI models would create a duopoly for companies like Anthropic and OpenAI, effectively imposing a 'token tax' on all American enterprises. This forces them to use alternatives that are 50-100x more expensive, creating an irrational cost structure and making them globally uncompetitive.
The market's negative reaction to Google's huge CapEx spending is misguided. For a company that has historically compounded invested capital at over 30%, aggressively investing in the next wave of computing infrastructure is a strong positive signal. Investors should trust management's methodical plan to secure their long-term edge.
A critical disconnect exists between tech and policy circles regarding AI. Policymakers often confuse model 'weights' (the proprietary code, which is software) with model 'outputs' (the generated results). Learning from public outputs is standard practice, while stealing weights is theft. This confusion leads to flawed policy discussions.
Proposals to ban American developers from using Chinese open-source models would backfire badly. It would cut the US off from global innovation, as the rest of the world would continue to build upon these models. The correct approach to stop Chinese distillation is for US AI labs to block access at the source.
Anthropic's argument that Chinese AI model distillation is 'IP theft' is a potentially fatal legal mistake. This assertion can be used against them in lawsuits from content creators like the New York Times, as Anthropic's own models are built by 'distilling' public content, effectively confessing their product is based on stolen IP.
China may be promoting open-source AI to commoditize the knowledge and services economy, where the US holds a dominant position. This strategic move aims to shift global economic value towards the 'molecule economy'—manufacturing and energy production—where China has a massive and growing structural advantage over the West.
The economic value in AI is rapidly shifting away from foundational models, which are becoming commoditized far faster than anticipated. The real, sustainable business models are emerging at the infrastructure layer (cloud, chips) and the application layer, not in the foundational models themselves.
Leading AI labs like Anthropic are compared to basketball players 'flopping'—exaggerating a threat to draw a foul. They are accused of manufacturing a panic around competition from Chinese open-source models to lure the government into granting them a protected duopoly, despite their own record-breaking growth.
The progressive argument that 'evictions are violence' ignores the second-order effects on other residents. When landlords cannot remove delinquent or unruly tenants, it degrades the quality of life for everyone else in the building, including long-standing, law-abiding residents who cannot afford to move elsewhere.
Policies that ban landlords from performing credit checks or vetting income will have the opposite of their intended effect. To mitigate risk from potentially non-paying tenants, landlords will be forced to demand much higher upfront rents and multi-month prepayments, making housing even less affordable for responsible tenants.